JP7797455B2 - Image forming device - Google Patents

Description

The present invention relates to an image forming apparatus that forms an image on paper.

Image forming devices change the maximum density and gradation characteristics of the images they form due to environmental fluctuations and changes in components over time. To address this, image forming devices perform calibration to maintain the maximum density at a target density and the gradation characteristics at target characteristics. Patent Document 1 discloses an image forming device that performs calibration by feeding back the results of reading a gradation pattern formed on paper to the image formation conditions. Patent Documents 2 and 3 propose a configuration that uses environmental conditions and the image formation conditions set in the image forming device as input values to predict the density of an image immediately after powering on or immediately after returning from power-saving mode. Based on the prediction results, the maximum density can be maintained at a target density.

Japanese Patent Application Laid-Open No. 2000-238341 Japanese Patent Application Laid-Open No. 2019-056760 Japanese Patent Application Laid-Open No. 2020-091427

Recovery from a neglected state (non-image forming state) is one factor that can cause changes in image quality, such as image density and gradation characteristics. Examples of such factors include environmental changes before and after neglect, recovery from power-on, and recovery from a power-saving state. Calibration methods for maintaining image quality include "actual measurement control," which updates image forming conditions based on the results of pattern image reading (measurement results), as described in Patent Document 1. Actual measurement control is suitable for maintaining quality, but it results in long downtime and affects productivity.

To reduce downtime, there is "predictive control," which does not perform actual measurement control but instead predicts image density based on the amount of time the printer has been left unused, as described in Patent Documents 2 and 3, and updates image formation conditions based on the predicted image density. Predictive control predicts image density using a predetermined model, which is vulnerable to disturbances and can result in large discrepancies between the predicted and actual measurement results.

In consideration of the above problems, the primary objective of the present invention is to provide an image forming apparatus that achieves both reduced downtime and maintained image quality.

The image forming apparatus of the present invention comprises an image forming means for forming an image on a sheet of paper, a reading means for reading a pattern image formed on the sheet of paper, and a control means capable of executing actual measurement control for correcting image forming conditions based on the result of reading the pattern image by the reading means, a first predictive control for correcting image densities of all gradations including a maximum density based on image densities predicted by environmental information of the image forming means, and a second predictive control for correcting image densities of all gradations excluding a maximum density based on image densities predicted by environmental information of the image forming means, wherein the control means is configured to When transitioning from a non-image forming state in which no image formation is performed to an image forming state in which image formation is performed, if the time period of the non-image forming state is equal to or longer than a predetermined time, the first predictive control is executed if the time period of the non-image forming state is less than the predetermined time and the temperature difference between the temperature at the end of the previous image formation and the temperature at the transition to the image forming state is equal to or longer than a predetermined temperature difference, and the second predictive control is executed if the time period of the non-image forming state is less than the predetermined time and the temperature difference between the temperature at the end of the previous image formation and the temperature at the transition to the image forming state is less than the predetermined temperature difference .
Another image forming apparatus of the present invention comprises image forming means for forming an image on paper, reading means for reading a pattern image formed on the paper, and control means capable of executing actual measurement control for correcting image forming conditions based on the result of reading the pattern image by the reading means, first predictive control for correcting image densities of all gradations including a maximum density based on image densities predicted by environmental information of the image forming means, and second predictive control for correcting image densities of all gradations excluding a maximum density based on image densities predicted by environmental information of the image forming means, wherein the control means controls the image forming conditions of the image forming means. When the image forming apparatus transitions from a non-image forming state in which no image formation is performed to an image forming state in which an image is formed, if the time period of the non-image forming state is equal to or longer than a predetermined time period, the actual measurement control is executed, if the time period of the non-image forming state is less than the predetermined time period and the humidity difference between the humidity at the end of the previous image formation and the humidity at the time of transition to the image forming state is equal to or longer than a predetermined humidity difference, the first predictive control is executed, and if the time period of the non-image forming state is less than the predetermined time period and the humidity difference between the humidity at the end of the previous image formation and the humidity at the time of transition to the image forming state is less than the predetermined humidity difference, the second predictive control is executed.

This invention makes it possible to reduce downtime while maintaining image quality.

System configuration diagram. FIG. 2 is a diagram illustrating the hardware configuration of the image forming apparatus. Diagram of the machine learning server configuration. FIG. Functional block diagram for machine learning of the system. 1A and 1B are explanatory diagrams of a learning model. 4A to 4C are explanatory diagrams of a printer control unit. 10 is a flowchart showing image density correction control. FIG. 10 is an explanatory diagram of a process for creating a predicted density characteristic. FIG. 10 is a flowchart showing a creation mode selection process. 10 is a table showing the confirmation results for each mode under each condition. 10 is a flowchart showing a creation mode selection process. 10 is a table showing the confirmation results for each mode under each condition. FIG. FIG. 10 is a combination table for selecting whether or not to enable image density correction control. 10 is a flowchart showing a selection process for image density correction control.

A preferred embodiment of the present invention will now be described with reference to the accompanying drawings. In this embodiment, an electrophotographic image forming apparatus will be described, but the present invention is also applicable to inkjet printers, dye-sublimation printers, and the like. In other words, the present invention is applicable to image forming apparatuses in which image density fluctuates in correlation with fluctuations in environmental conditions, etc.

(System configuration)
1 is a configuration diagram of a system including an image forming apparatus according to this embodiment. This system is configured by connecting an image forming apparatus 100, a machine learning server 102, a data server 105, a general-purpose computer 103, and the like so that they can communicate with each other via a network 104. The image forming apparatus 100 is a printer, a copier, a multifunction peripheral, a facsimile machine, or the like. The general-purpose computer 103 transmits image data to the image forming apparatus 100, for example. A plurality of image forming apparatuses 100 and a plurality of general-purpose computers 103 may be connected to the network 104. The network 104 is, for example, a wired local area network (LAN), a wireless LAN, a public communication line, or the like.

The image forming device 100 is equipped with an AI (Artificial Intelligence) function. The machine learning server 102 plays a central role in generating the trained model used to implement this AI function. The data server 105 collects training data used for machine learning in the machine learning server 102 from external devices and provides it to the machine learning server.

The image forming device 100 is capable of realizing specific AI functions by obtaining generated trained models from the machine learning server 102 at any time. The machine learning server 102 obtains the training data required to train the trained models to realize specific AI functions from external devices such as the data server 105, the image forming device 100, and the general-purpose computer 103. The machine learning server 102 is capable of performing machine learning using at least a portion of the training data obtained from the external devices.

In this system, data representing the device status of the image forming device 100 is collected by the data server 105, and this data is learned by the machine learning server 102 to generate a learning model. The image forming device 100 acquires from the machine learning server 102 a learning model that estimates the image density setting value when outputting an image. The image forming device 100 is equipped with an AI function that utilizes the acquired learning model. The image forming device 100 is able to suppress fluctuations in image density by performing calibration using the estimated image density setting value.

(Image forming apparatus)
2 is a hardware configuration diagram of the image forming apparatus 100. The image forming apparatus 100 includes an operation unit 140, a controller 1200, a reader 250, and a printer 20. The operation unit 140, the reader 250, and the printer 20 are connected to the controller 1200. The controller 1200 controls the operations of the operation unit 140, the reader 250, and the printer 20, and communicates with the machine learning server 102, the data server 105, and the general-purpose computer 103 via the network 104.

The operation unit 140 is a user interface and includes an input interface for accepting user instructions and input of setting values, and an output interface for outputting various information to the user. The input interface is, for example, a key button or touch panel. The output interface is, for example, a display or speaker.

The reader 250 is an image reading device that reads images in response to instructions from the operation unit 140. The reader 250 has a processor that controls the reader 250, as well as a light source and scanning mirror for reading images. The printer 20 prints the image on paper. The configurations of the reader 250 and printer 20 will be described in detail below.

The controller 1200 includes a system bus 1207 and an image bus 2008. The system bus 1207 and image bus 2008 are communicatively connected via a bus interface (I/F) 1205. The bus I/F 1205 is a bus bridge that performs processes such as data structure conversion between the system bus 1207 and the image bus 2008.

The system bus 1207 is connected to a CPU (Central Processing Unit) 1201, RAM (Random Access Memory) 1202, ROM (Read Only Memory) 1203, and storage 1204. The CPU 1201 controls the operation of the image forming apparatus 100 by executing computer programs stored in the ROM 1203 and storage 1204. A boot program is stored in the ROM 1203. The storage 1204 stores system software, image data, software counter values, etc.

RAM 1202 provides a work area when CPU 1201 executes processing and stores temporary data, etc. RAM 1202 stores image formation conditions, control tables, conversion tables, etc. Storage 1204 is a large-capacity storage device such as an HDD (Hard Disk Drive) or SSD (Solid State Drive). RAM 1202 or storage 1204 records output attribute information, including user name, number of copies, color printing, etc., when a print job or copy job is executed, as well as the job execution history, as job log information.

The system bus 1207 is connected to the following interfaces: an operation unit I/F 1206, a wired communication I/F 1210, a modem 1211, a wireless communication I/F 1270, and a communication I/F 1208. The operation unit I/F 1206 is connected to the operation unit 140, acquires instructions from the operation unit 140, transmits them to the CPU 1201, and outputs various information from the operation unit 140 in response to instructions from the CPU 1201. The wired communication I/F 1210 and the wireless communication I/F 1270 are communication interfaces for communicating via the network 104. The wireless communication I/F 1270 can control communication with the network 104 via the wireless line 106. The modem 1211 is connected to the public line 3001, and communicates (sends and receives) data with an external facsimile machine (not shown). The communication I/F 1208 controls communication between the reader 250 and the printer 20.

A GPU (Graphics Processing Unit) 1291 and a timer 1209 are connected to the system bus 1207. The GPU 1291 is capable of performing efficient calculations by processing large amounts of data in parallel, making it effective when performing multiple learning rounds using a learning model such as deep learning. In this embodiment, when performing machine learning, the CPU 1201 and GPU 1291 work together to perform the processing. Note that if the processing capabilities of the CPU 1201 and GPU 1291 are high, machine learning may be performed independently by the CPU 1201 or GPU 1291.

The image bus 2008 is connected to the RIP (Raster Image Processor) unit 1260, reader image processing unit 1280, printer image processing unit 1290, image rotation unit 1230, image compression unit 1240, and device I/F 1220.

The RIP unit 1260 converts PDL (Page Description Language) records included in a print job acquired from the general-purpose computer 103 into bitmap images. The reader image processing unit 1280 performs image processing such as correction, processing, and editing on image data acquired from the reader 250. The image data acquired from the reader 250 is read data representing an image read from an original by the reader 250. The printer image processing unit 1290 performs image processing such as correction and resolution conversion on image data representing an image to be output (printed) by the printer 20. The image rotation unit 1230 performs image rotation on image data. The image compression unit 1240 compresses and decompresses images. For example, the image compression unit 1240 compresses and decompresses multi-value image data in accordance with the JPEG standard and binary image data in accordance with the JBIG, MMR, or MH standard. The device I/F 1220 converts image data between synchronous and asynchronous systems between the reader 250, printer 20, and controller 1200.

(machine learning server)
3 is a configuration diagram of the machine learning server 102. The machine learning server 102 includes a CPU 301, a RAM 302, a ROM 303, a storage 304, an IO unit 305, a GPU 306, and a communication I/F 310. These components are connected to a system bus 307.

The CPU 301 controls the operation of the machine learning server 102 by executing computer programs stored in the ROM 303 and storage 304. The RAM 302 provides a work area when the CPU 301 executes processing and stores temporary data, etc. The ROM 303 stores the BIOS (Basic Input Output System), the OS (Operating System) startup program, configuration files, etc. The storage 1204 is a large-capacity storage device such as an HDD or SSD. The storage 1204 stores system software, etc. The communication I/F 310 is connected to the network 104 and controls communications with other devices connected to the network 104, such as the image forming apparatus 100.

The IO unit 305 is an interface with an operation unit (not shown) that is composed of a display such as a touch panel and an input device. Predetermined information is displayed on this operation unit at a predetermined resolution, number of colors, etc. For example, a GUI (Graphical User Interface) screen is formed on the operation unit, and various windows and data required for operation are displayed.

GPU 306 is capable of performing efficient calculations by processing large amounts of data in parallel, making it effective when performing multiple rounds of learning using a learning model such as deep learning. In this embodiment, when performing machine learning, CPU 301 and GPU 306 work together to execute the process. Specifically, when executing a learning program including a learning model, CPU 301 and GPU 306 work together to execute the process, thereby performing learning. If the processing capabilities of CPU 301 and GPU 306 are high, machine learning may be performed independently by CPU 301 or GPU 306.

This section explains how to use the GPU 1291 of the image forming apparatus 100 and the GPU 306 of the machine learning server 102. The computational resources of the GPUs 1291 and 306 are effectively utilized depending on the communication load of the network 104, the processing load of each GPU 1291 and 306, the power saving mode of the image forming apparatus 100, etc. For example, when the image forming apparatus 100 transitions to power saving mode, the GPU 306 on the machine learning server 102 side is actively utilized. When the communication load is heavy, the GPU 1291 on the image forming apparatus 100 side is utilized.

(Configuration of image forming apparatus)
4 is a cross-sectional view of the image forming apparatus 100. A full-color printer is illustrated in FIG. 4. As described above, the image forming apparatus 100 includes the reader 250 and the printer 20.

The reader 250 reads images from the original 21 or test chart. The test chart is a sheet of paper on which multiple pattern images for adjusting image formation conditions are printed. The reader 250 includes a platen glass 22, a light source 23, an optical system 24, and a reading unit 25. The light source 23 irradiates light onto the original 21 placed on the platen glass 22. The optical system 24 guides the light reflected from the original 21 to the reading unit 25, where it forms an image. The reading unit 25 is composed of an array of multiple photoelectric conversion elements, such as CCDs (Charge-Coupled Devices). The reading unit 25 generates red, green, and blue color component signals based on the formed reflected light. The reader image processing unit 1280 performs image processing (e.g., shading correction) on the color component signals obtained from the reading unit 25 to generate image data representing the read image. The reader image processing unit 1280 transmits the image data to the printer image processing unit 1290.

Printer 20 forms (prints) a toner image on paper S based on image data. Printer 20 includes image forming units 10a, 10b, 10c, and 10d that form toner images of the colors Y (yellow), M (magenta), C (cyan), and K (black). Image forming units 10a, 10b, 10c, and 10d are arranged along intermediate transfer belt 205. Image forming unit 10a is used to form yellow images. Image forming unit 10b is used to form magenta images. Image forming unit 10c is used to form cyan images. Image forming unit 10d is used to form black images. Image forming units 10a, 10b, 10c, and 10d have the same configuration. In the following description, the suffixes a, b, c, and d are omitted when it is not necessary to distinguish between colors. Note that printer 20 of this embodiment is not limited to a color printer that forms full-color images, and may be, for example, a monochrome printer that forms single-color images.

The image forming unit 10 includes a photosensitive drum 201. The photosensitive drum 201 is a drum-shaped photosensitive body with a photosensitive layer on its surface. A charger 202, an exposure unit 200, a developing unit 203, a drum cleaner 4, and a primary transfer unit 204 are provided around the photosensitive drum 201. The charger 202 applies a charging bias to charge the surface of the photosensitive drum 201, which rotates in the direction of arrow R1. The exposure unit 200 scans the charged surface of the photosensitive drum 201 with a laser beam (light beam), forming an electrostatic latent image on the surface of the photosensitive drum 201. The exposure unit 200 outputs laser light modulated based on image data processed by the printer image processing unit 1290. The developing unit 203 applies a development bias to develop the electrostatic latent image with developer (toner), forming a toner image on the surface of the photosensitive drum 201.

When a primary transfer bias is applied, the primary transfer unit 204 transfers the toner image formed on the surface of the photosensitive drum 201 onto the intermediate transfer belt 205, which is an image carrier. The intermediate transfer belt 205 is sandwiched between the photosensitive drum 201 and the primary transfer unit 204 and is looped around rollers such as the inner secondary transfer roller 221. The intermediate transfer belt 205 is an endless belt-like intermediate transfer body that rotates in the direction of arrow R2.

A yellow toner image is formed on photosensitive drum 201a. A magenta toner image is formed on photosensitive drum 201b. A cyan toner image is formed on photosensitive drum 201c. A black toner image is formed on photosensitive drum 201d. The toner images of each color are transferred sequentially, superimposed, onto the rotating intermediate transfer belt 205. Any toner remaining on photosensitive drum 201 that is not transferred to intermediate transfer belt 205 is removed by drum cleaner 4.

As the intermediate transfer belt 205 rotates, it transports the transferred toner images of each color to the inner secondary transfer roller 221. A secondary transfer roller 222 is provided opposite the inner secondary transfer roller 221, with the intermediate transfer belt 205 sandwiched between them. The inner secondary transfer roller 221 and the secondary transfer roller 222 form a secondary transfer unit. The toner images of each color transferred to the intermediate transfer belt 205 are transferred together onto the paper S between the inner secondary transfer roller 221 and the secondary transfer roller 222. The toner images are transferred by applying a secondary transfer bias to the secondary transfer roller 222. The paper S onto which the toner images have been transferred is heated and pressed by the fixing unit 40, thereby undergoing a fixing process.

Paper S is fed from paper feed cassette 209 or manual feed tray 210. For example, paper S stored in paper feed cassette 209 is fed by paper feed roller 218 and transported to registration roller 211 by transport roller 214. Registration roller 211 corrects any skew in paper S and transports paper S to the secondary transfer unit in time with the timing at which the toner image carried on intermediate transfer belt 205 is transported to the secondary transfer unit. This causes the toner image to be transferred to a predetermined position on paper S.

When printing on one side of the paper S, or when printing on both sides of the paper S is complete during double-sided printing, the paper S after the fixing process is discharged outside the device by discharge rollers 208. When printing on both sides of the paper S with an image printed on one side (first side), the paper S passes through the fixing unit 40 and is then transported toward the double-sided reversing path 212. The paper S is reversed in the transport direction by the double-sided reversing path 212 and transported to the double-sided path 213. By reversing the transport direction by the double-sided reversing path 212, the printed side is reversed from the first side to the other side (second side). After passing through the double-sided path 213, the paper S is transported by the transport rollers 214 to the registration rollers 211, where an image is printed in the same way as on the first side. In this way, a finished product with an image printed on the paper S is obtained.

An in-line sensor 215 is provided between the fixing unit 40 and the paper discharge roller 208. The in-line sensor 215 is an optical sensor used to detect image defects, image density, image position, color misalignment, etc.

An image density sensor 408 is provided near the intermediate transfer belt 205, downstream of the image forming unit 10d in the rotation direction of the intermediate transfer belt 205. The image density sensor 408 measures the unfixed image (toner image) carried on the intermediate transfer belt 205. The image density sensor 408 is, for example, an optical sensor having a light-emitting element and a light-receiving element, and measures reflected light from the unfixed image on the intermediate transfer belt 205. The intensity or amount of light reflected from the unfixed image changes depending on the amount of toner attached to the unfixed image. The image forming apparatus 100 can detect the image density of the unfixed image, for example, based on a conversion table between the intensity of reflected light from the unfixed image and the image density of the unfixed image.

The image forming apparatus 100 is equipped with an environmental sensor 409 within the printer 20. The environmental sensor 409 is capable of acquiring environmental information within the image forming apparatus 100 before, during, and after image formation. In this embodiment, the environmental sensor 409 is configured as a detection unit capable of detecting the temperature, humidity, absolute moisture content, etc. of the environment in which the image forming apparatus 100 is installed. The detected values (environmental information) detected by the environmental sensor 409 are used to determine image formation conditions and to select the image correction control to be executed from multiple image correction controls.

The environmental sensor 409 is not limited to measuring temperature and humidity, but may measure any environmental condition that may affect the image being printed. The environmental sensor 409 may also be provided at multiple locations within the image forming apparatus 100. The location of the environmental sensor 409 is not limited as long as it can measure environmental information that is sensitive to the quality of image density and gradation characteristics within the image forming apparatus 100.

(Function block)
Fig. 5 is a functional block diagram for performing machine learning in the system of this embodiment. Each function of the image forming apparatus 100 is realized, for example, by the CPU 1201 executing a computer program in the hardware configuration of Fig. 2. Each function of the machine learning server 102 is realized, for example, by the CPU 301 executing a computer program in the hardware configuration of Fig. 3. The same is true for each function of the data server 105, which is realized by a CPU (not shown) of the data server 105 executing a computer program.

The system of this embodiment learns information for adjusting the image density of images formed by the image forming device 100 and performs processing to estimate the image density setting value. The functional block diagram in Figure 5 shows the functional blocks for performing this processing.

The image forming apparatus 100 functions as a data storage unit 501, a job control unit 503, an image density adjustment unit 511, an image reading unit 504, an estimation processing unit 505, a counter unit 506, and an apparatus status detection unit 507. The image forming apparatus 100 also functions as an image density detection unit 508, an environmental condition detection unit 509, and an image formation condition control unit 530. The machine learning server 102 functions as a learning data generation unit 513, a machine learning unit 514, and a data storage unit 515. The data server 105 functions as a data collection and provision unit 510 and a data storage unit 512.

Each function of the image forming apparatus 100 will be described.
The data storage unit 501 records data input to and output from the image forming apparatus 100, such as image data, learning data, and learning models, in the RAM 1202 or storage 1204. The job control unit 503 executes basic functions of the image forming apparatus 100, such as copying, faxing, and printing, based on user instructions, and primarily transmits and receives instructions and data between other function blocks associated with the execution of these basic functions. The image density adjustment unit 511 sets an image density setting value that optimally adjusts the image density based on the results of image density adjustment. The image density adjustment unit 511 also reflects the estimation results by the estimation processing unit 505 in the image density setting value.

The image density detection unit 508 detects the image density of the toner image on the intermediate transfer belt 205 based on the detection results from the image density sensor 408. The environmental condition detection unit 509 detects the environment in which the image forming apparatus 100 is installed and the internal environmental conditions (environmental conditions) of the image forming apparatus 100 based on the detection results from the environmental sensor 409. The environmental condition detection unit 509 generates environmental information representing the detected environmental conditions. The image formation condition control unit 530 sets optimal image formation conditions in the image forming apparatus 100 to satisfy the specification requirements transmitted and received by the job control unit 503. For example, the image formation condition control unit 530 generates image formation conditions according to the image density setting value and sets them in the image forming apparatus 100.

The counter unit 506 stores data containing counter information for image formation and non-image formation states. The counter information is, for example, the number of times toner has been replenished to the developing device 203. The image forming device 100 has a replenishment mechanism (not shown). The replenishment mechanism replenishes the target developing device 203 with a predetermined amount of toner per replenishment operation. Therefore, the image forming device 100 can predict the amount of toner to be replenished to the developing device 203 based on the count value of the counter unit 506.

When performing the copy function or scan function based on instructions from the job control unit 503, the image reading unit 504 controls the optical reading operation of the original 21 by the reader 250 and the optical reading operation of the paper S by the inline sensor 215.

The estimation processing unit 505 is implemented by the CPU 1201 and GPU 1291, and performs estimation processing, classification processing, and the like on data input and output by the image forming apparatus 100 in order to realize AI functions. The estimation processing unit 505 performs processing based on instructions from the job control unit 503. The processing results of the estimation processing unit 505 are sent to the job control unit 503 and fed back to the user. Feedback to the user is provided, for example, by displaying a notification message on the operation unit 140.

The functions of the data server 105 will be described below. The data collection and provision unit 510 collects and provides learning data for the machine learning server 102 to learn from. In this embodiment, the data collection and provision unit 510 acquires learning data from the image forming device 100, including device information such as the device status, usage history, environmental conditions, and image density setting value of the image forming device 100, and provides the learning data to the machine learning server 102. The data collection and provision unit 510 stores and manages the collected learning data in the data storage unit 512.

The functions of the machine learning server 102 are explained below. The learning data generation unit 513 acquires learning data from the data server 105. The learning data generation unit 513 performs pre-processing on the acquired learning data to effectively obtain learning effects, such as removing data that becomes noise, and optimizes the learning data. Pre-processing is performed, for example, by filtering device information acquired from the image forming device 100 immediately after adjusting the image formation conditions. This processing results in learning data that can be used to effectively learn image density setting values, for example.

The data storage unit 515 stores data received from the data server 105, learning data generated by the learning data generation unit 513, and a learned model in the machine learning unit 514. The data storage unit 515 temporarily records these data and models in the RAM 302 or the storage 304.
The machine learning unit 514 performs machine learning using as input the learning data generated by the learning data generation unit 513. The machine learning unit 514 is realized by the GPU 306 and the CPU 301. The machine learning unit 514 performs machine learning based on a learning method using a learning model, which will be described later with reference to FIG. 6 .

(Learning model)
Fig. 6 is an explanatory diagram of a learning model of machine learning performed by the machine learning unit 514 of the machine learning server 102. Fig. 6 shows the input/output structure using the learning model by the machine learning unit 514, and illustrates a learning model using a neural network.

In the neural network shown in Figure 6(a), the set values or measured values that can be obtained in the image formation state and non-image formation state are used as input data X. Here, examples of input data X include X1 to X8, which are related to the generation of a learning model for predicting image density set values.

The elements of the learning data are not limited to data related to image density fluctuations, but can also be data obtainable from sensors installed in the image forming device 100. Furthermore, when data expressed as categorical variables, such as paper settings, double-sided printing, and continuous or intermittent operation, are handled as numerical values in machine learning, data preprocessing involves converting the numerical representation using known techniques such as one-hot encoding.

Specific machine learning algorithms include neural networks, nearest neighbor methods, naive Bayes methods, decision trees, and support vector machines. Also included is deep learning, which uses neural networks to generate features and connection weighting coefficients for learning. Any of the above algorithms that are available can be used as appropriate and applied to this embodiment.

As shown in FIG. 6(b), the learning model (W) may include an error detection unit and an update unit. The error detection unit, for example, uses a loss function to derive a loss (L) representing the error between the training data T and the output data Y (4) output from the output layer of the neural network in response to the input data X (2) input to the input layer. The update unit updates the connection weighting coefficients between the nodes of the neural network so as to reduce the loss (L) obtained by the error detection unit. The update unit updates the connection weighting coefficients, for example, using the backpropagation method. The backpropagation method is a technique for adjusting the connection weighting coefficients between the nodes of each neural network so as to reduce the error between the output data Y and the training data T.

The learning model (W) is created by preparing a large amount of training data consisting of sets of "input data with known correct values" and "correct values," and adjusting the weighting coefficients within the learning model (W) so that when input data corresponding to these correct values is input, the output comes as close as possible to the correct value. By performing this type of processing, the learning model (W) becomes a highly accurate learning model (W). This type of processing is called the "learning process," and a learning model that has been adjusted through the learning process is called the "trained model."

The training data T to be prepared is a set of "input data with known correct values" and "correct values." The learning model in the machine learning unit 514 is not limited to the deep learning system described above, and models based on linear regression or non-linear regression may also be applied. There is no problem with the learning model calculating the correct value by multiplying the input data by a known coefficient.

(Controller 1200 of image forming apparatus 100)
7 is an explanatory diagram of a printer control unit including the controller 1200 of the image forming apparatus 100 in FIG. 2 and the engine controller 711 provided in the printer 20. In the controller 1200, the same components as those in FIG. 2 are denoted by the same reference numerals. A description of the same components as those in FIG. 2 will be omitted.

Figure 7(a) shows the configuration of the printer control unit 700. The engine controller 711 mainly controls the operation of the printer 20. The controller 1200 includes a CPU 1201, RAM 1202, ROM 1203, host I/F 704, reader I/F 705, RIP unit 1260, color processing unit 707, tone correction unit 708, halftone unit 709, and engine I/F 710. The operation unit 140 and engine controller 711 are connected to the controller 1200.

The host I/F 704 is realized by at least one of the wired communication I/F 1210 and the wireless communication I/F 1270 shown in Fig. 2. The host I/F 704 acquires print job instructions and image data from the general-purpose computer 103, for example.
The reader I/F 705 and the engine I/F 710 are realized by the communication I/F 1208 and the device I/F 1220 in Fig. 2. The reader I/F 705 acquires image data representing the image of the original 21 read by the reader 250. When the original 21 is a test chart on which a pattern image is formed, the reader I/F 705 acquires image data (read data) that is the result of reading the test chart.
The engine I/F 710 communicates with the engine controller 711. For example, the engine I/F 710 transmits image data to the engine controller 711 and acquires various measurement values from the engine controller 711.

The color processing unit 707, tone correction unit 708, and halftone unit 709 are implemented by the printer image processing unit 1290 in Figure 2. The color processing unit 707 converts the color space of a bitmap image using a color management profile or the like. For example, the color processing unit 707 converts RGB format image data into YMCK format image data. The tone correction unit 708 corrects the image data based on a tone correction table (γLUT) so that the tone characteristics of the image formed by the printer 20 become ideal. The halftone unit 709 performs pseudo-halftone processing such as dither matrix or error diffusion on the tone-corrected image data. The image data output from the halftone unit 709 is sent to the engine controller 711 via the engine IF 710. The color processing unit 707, tone correction unit 708, and halftone unit 709 support multiple image processing methods.

Note that the image forming apparatus 100 may execute at least some of these image processes, including the processing of the RIP unit 1260, using an image processor such as the GPU 1291. Furthermore, these image processes may be performed by the GPU 1291 in cooperation with the CPU 1201. There may be multiple image processors.

The engine controller 711 includes a CPU 712, RAM 713, ROM 714, a high-voltage power supply 717, a timer 719, and a counter 720. The engine controller 711 is connected to the environment sensor 409, density sensor 716, exposure unit 200, and in-line sensor 215. The engine controller 711 is built into the printer 20.

The CPU 712 controls the operation of the printer by executing computer programs stored in the ROM 714. The RAM 713 provides a work area for the CPU 712 when it executes processing.

High-voltage power supply 717 generates high voltages such as a charging bias, a developing bias, a primary transfer bias, and a secondary transfer bias. Timer 719 starts timing when the image formation process is completed, thereby measuring the non-image formation time (idle time) during which image forming apparatus 100 is not forming an image. Timer 719 can also measure operating time (image formation time). Counter 720 counts the number of times toner has been replenished to developer 203d. Image forming apparatus 100 has a replenishment mechanism (not shown). The replenishment mechanism replenishes the amount of toner to developer 203d in each replenishment operation, which is predetermined. Therefore, controller 1200 can predict the amount of toner to be replenished to developer 203d based on the count value of counter 720.

The density sensor 716 detects the toner density (black toner density) of the developing unit 203d. The toner density is, for example, a parameter that indicates the ratio of toner to carrier. The density sensor 716 is, for example, a magnetic permeability sensor. The in-line sensor 215 includes a light source and a line sensor having multiple light-receiving elements. The in-line sensor 215 is positioned so that the longitudinal direction of the line sensor is perpendicular to the transport direction of the paper S. The in-line sensor 215 reads a test chart on which a pattern image is printed on the paper S.

Figure 7(b) shows the functions realized by the CPU 1201 executing the control program. The CPU 1201 functions as a potential control unit 721, a loading amount adjustment unit 722, a table creation unit 723, and a prediction unit 724.

The potential control unit 721 determines voltage values such as the charging bias and developing bias based on environmental information acquired by the environmental sensor 409. The potential control unit 721 sets the determined voltage value in the high-voltage power supply 717. The high-voltage power supply 717 outputs a voltage of the set voltage value. The potential control unit 721 and high-voltage power supply 717 function as a voltage control unit.

The loading amount adjustment unit 722 adjusts the maximum amount of toner that can be loaded on the sheet S (maximum toner loading amount). The maximum toner loading amount is adjusted based on the charging bias and the developing bias. The loading amount adjustment unit 722 may also adjust the maximum toner loading amount by controlling the intensity of the laser light (laser power LPW) output from the exposure unit 200d. The loading amount adjustment unit 722 and the exposure unit 200d function as an exposure control unit.

The table creation unit 723 creates a gradation correction table (γLUT) used by the gradation correction unit 708. The table creation unit 723 can create gradation correction tables in three creation modes.

The first mode is a mode in which a pattern image is formed as in the conventional method, and a tone correction table is created based on the measurement results of the pattern image without performing predictive control using a learning model. In this embodiment, the tone correction table created in the first mode is called the "basic table."

The second mode is a mode specific to this embodiment, in which image density is predicted based on environmental information, image forming conditions, etc., and a gradation correction table is created based on the predicted image density. The gradation correction table created here is called a "correction table."
The prediction of image density is performed by the prediction unit 724. In this embodiment, a correction table is created based on the predicted image density (predicted density, predicted value), and a composite table is created each time the basic table and the correction table are combined. The composite table is set as a gradation correction table in the gradation correction unit 708. The gradation correction unit 708 corrects the gradation of the image data using the set composite table (gradation correction table). In the second mode, pattern image formation and measurement are not performed, thereby significantly reducing downtime.

The third mode, like the second mode, predicts image density based on environmental information, image formation conditions, etc., and creates a tone correction table based on the predicted image density (predicted value). The difference from the second mode is that it does not predict the maximum density that can be output during image formation. In other words, the third mode creates the remaining tone correction tables, excluding the predicted maximum density, based on the predicted image density.

Image density prediction is performed by the prediction unit 724. In this embodiment, a correction table is created based on the predicted density, and a composite table is created by combining the basic table and the correction table. The composite table is set as a gradation correction table in the gradation correction unit 708. The gradation correction unit 708 corrects the gradation of the image data using the set composite table (gradation correction table). In the third mode, as in the second mode, pattern image formation and measurement are not performed, significantly reducing downtime.

Figure 7(c) shows some of the information stored in RAM 1202. RAM 1202 stores a basic table 725, a correction table 726, and image forming conditions 727. The basic table 725 is a tone correction table created in the first mode. The correction table 726 is a table created in the second or third mode, and is a table for correcting the basic table 725 to obtain a composite table. The image forming conditions 727 include, for example, the voltage value of the charging bias (charging setting), the voltage value of the developing bias (development setting), the laser power LPW (exposure setting), the fixing temperature (fixing setting), etc.

The prediction of image density by the prediction unit 724 will now be described. The prediction unit 724 sends multiple predicted densities to the table creation unit 723. These multiple predicted densities (predicted density group) form image density characteristics and are used to create a gradation correction table. The table creation unit 723 creates a gradation correction table based on the predicted density group. As described above, the table creation unit 723 creates a correction table 726 based on the predicted density group, combines it with the basic table 725 to create a gradation correction table (composite table), and sets it in the gradation correction unit 708.

(Image density correction control)
Image density correction control is performed based on the above three creation modes. Figure 8 is a flowchart showing image density correction control (calibration). Here, the image density correction control performed in the first mode is called main calibration, and the image density correction control performed in the second and third modes is called predictive calibration. The main calibration is actual measurement control (first calibration) in which image formation conditions are generated based on image density obtained from a pattern image formed on paper S. The predictive calibration is predictive control (second calibration) in which image formation conditions are generated based on predicted density.

The CPU 1201 determines whether the conditions for executing main calibration (first mode) are met (S800). If the conditions are met (S800: Y), main calibration is executed. If the conditions are not met (S800: N), predictive calibration is executed. The conditions for executing main calibration include, for example, the elapsed time since the previous main calibration, the number of pages printed, and changes in the installation environment. It is determined that the conditions for executing main calibration are met if a predetermined amount of time has passed since the previous main calibration, if a predetermined number of pages have been printed since the previous main calibration, or if the image forming device 100 has been moved.

Main Calibration When main calibration is performed, the CPU 1201 determines potentials such as the charging bias (VdT) and developing bias (Vdc) from the potential control unit 721 in order to perform potential control on the engine controller 711 (S801). The CPU 1201 determines the charging bias (VdT) and developing bias (Vdc) in accordance with the environmental conditions (temperature, humidity, absolute moisture content, etc.) acquired by the environmental sensor 409. Since potential control is known, a detailed description thereof will be omitted.

The CPU 1201 adjusts the image formation conditions for obtaining an image of maximum density using the toner loading amount adjustment unit 722 (S802). The maximum density may also be referred to as the maximum toner loading amount. For example, the toner loading amount adjustment unit 722 sets the voltage values of the charging bias (VdT) and the developing bias (Vdc) determined by potential control in the engine controller 711 and controls the printer 20 to form a pattern image on the paper S for adjusting the maximum toner loading amount. The CPU 1201 causes the inline sensor 215 to read the paper S (test chart) on which the pattern image has been formed. The CPU 1201 acquires the test chart reading results (read data) from the inline sensor 215. The toner loading amount adjustment unit 722 determines the relationship between the toner loading amount and the laser power LPW based on the read data. Based on this relationship, the toner loading amount adjustment unit 722 determines the laser power LPW as the image formation condition for obtaining the maximum toner loading amount. Since the method for adjusting the maximum toner loading amount is also known, detailed description thereof will be omitted.

The CPU 1201 controls the printer 20 via the engine controller 711 to form a pattern image for gradation correction on the paper S based on the laser power LPW that allows the maximum toner load (S803). The pattern image for gradation correction includes, for example, a pattern image of 64 gradations for each color. The paper S on which the pattern image for gradation correction is printed is different from the paper S on which the pattern image for adjusting the maximum toner load is printed. The CPU 1201 causes the inline sensor 215 to read the paper S (test chart) on which the pattern image has been formed. The CPU 1201 obtains the reading results (read data) of the test chart from the inline sensor 215.

The CPU 1201 uses the table creation unit 723 to detect the image density of each gradation based on the scanned data of the pattern image for gradation correction (S804). The CPU 1201 then acquires the detected image density as a reference density using the table creation unit 723, and acquires the reference signal values of each sensor at this time (S805). The table creation unit 723 acquires the reference values and reference signal values of the image formation conditions set in the engine controller 711 to form the pattern image. The reference values of the image formation conditions are, for example, the charging bias, developing bias, and laser power LPW. The reference density is the image density of each gradation. The reference signal value is, for example, the toner density, count value, and timer value described above. The reference values, reference signal values, and reference density are stored in RAM 1202.

The CPU 1201 uses the table creation unit 723 to create a basic table 725 based on the measured image density so that the gradation characteristics of the image formed on the paper S match the ideal gradation characteristics (gradation target) (S806). The table creation unit 723, for example, performs interpolation and smoothing processes on the measured image density to obtain the gradation characteristics of the printer 20. The table creation unit 723 creates the basic table 725 based on the gradation characteristics of the entire density range and the gradation target. The table creation unit 723 sets the basic table 725 in the gradation correction unit 708.

Predictive Calibration After a predetermined time has passed since the basic table 725 was created, or after a predetermined number of images have been formed, the environmental conditions and the state of the image forming apparatus 100 change successively. Therefore, the basic table 725 needs to be corrected in response to these changes.

Creating the base table 725 during main calibration requires the creation of a pattern image, resulting in downtime. Predictive calibration (second and third modes) is a process that updates the gradation correction table without creating a pattern image. By employing predictive calibration, downtime is significantly reduced. Note that predictive calibration generates a correction table 726, which is combined with the base table 725 created during main calibration. This corrects (modifies) the gradation correction table.

When performing predictive calibration, the CPU 1201 first determines whether the conditions for performing predictive calibration are met (S807). Predictive calibration execution conditions include, for example, power-on, returning from sleep mode (power-saving mode), environmental changes, and reaching a preset timing. Predictive calibration is executed more frequently than main calibration. If the execution conditions are not met (S807: N), the CPU 1201 returns to processing at S800.

If the execution conditions are met (S807: Y), the CPU 1201 calculates the predicted density for each gradation (D prediction) using the prediction unit 724 (S808). Here, the prediction unit 724 calculates 10 predicted densities (D prediction) corresponding to 10 gradations. The CPU 1201 then uses the table creation unit 723 to create predicted density characteristics (predicted gradation characteristics) based on the calculated 10 predicted densities (D prediction) (S809).

FIG. 9 is an explanatory diagram of the process of creating the predicted density characteristics.
In the second mode, the prediction unit 724 determines the image densities of all gradations by performing interpolation calculations using ten predicted densities (D predictions) D_tgt1 to D_tgt10 shown on the vertical axis of Fig. 9. Note that the table creation unit 723 may determine the image densities of all gradations by an approximation formula that represents predicted density characteristics using the ten predicted densities (D predictions).
In the third mode, the prediction unit 724 determines the image density of all gradations except D_tgt10 (maximum density) by performing an interpolation calculation using nine predicted densities (D prediction) D_tgt1 to D_tgt9. The table creation unit 723 may determine the image density of all gradations except D_tgt10 using an approximation formula that represents predicted density characteristics using the nine predicted densities (D prediction). In order to distinguish between the second mode and the third mode, the predictive calibration of the second mode may be called the "second calibration," and the predictive calibration of the third mode may be called the "third calibration."

Figure 9(a) shows a gradation target 901, a base table 725, and a reference density characteristic 903. The horizontal axis represents the input signal corresponding to the gradation level. The vertical axis represents the image density. The reference density characteristic 903 is the reference density obtained in the processing of S805. The base table 725 is generated by inverting (inversely converting) the reference density characteristic 903 relative to the gradation target 901.

Figure 9(b) shows a gradation target 901, a reference density characteristic 903, and a predicted density characteristic 904. The predicted density characteristic 904 is the predicted density (D prediction) obtained in the processes of S808 and S809.

Due to changes over time, etc., the image density characteristics of the image forming device 100 change from the reference density characteristics 903 to the predicted density characteristics 904. Therefore, if the gradation correction unit 708 uses the basic table 725 created based on the reference density characteristics 903, it will no longer be able to correct the gradation characteristics with high precision.

The CPU 1201 causes the table creation unit 723 to create the correction table 726 based on the predicted density characteristics 904 (S810).
In the second mode, the table creation unit 723 performs correction control of the maximum density targeted by the basic table for the predicted density of D_tgt10, which is the maximum density. The table creation unit 723 corrects the maximum density by changing at least one of the image formation condition settings of the exposure setting, the charging setting, and the development setting based on the correction amount. Next, the table creation unit 723 creates a new predicted density characteristic 904 based on the changed image formation condition settings, and corrects the predicted density characteristic 904 after the correction control of the maximum density for the characteristics of all gradations in the basic table 725. The table creation unit 723 creates a correction table 726 by inversely converting the predicted density characteristic 904 to the characteristics of the basic table 725.
In the third mode, the table creation unit 723 does not perform the correction control for the maximum density that is performed in the second mode. The table creation unit 723 performs an interpolation calculation using nine predicted densities (D predictions) D_tgt1 to D_tgt9 while retaining the information of D_tgt10 in the previously set basic table 725. The table creation unit 723 corrects the interpolated predicted density characteristics 904 with respect to the characteristics of all gradations in the basic table 725. The table creation unit 723 creates a correction table 726 by performing an inverse conversion of the predicted density characteristics 904 with respect to the characteristics of the basic table 725.

The CPU 1201 creates a modified gradation correction table (composite table) by compositing the base table 725 and the modification table 726 using the table creation unit 723 (S811). Figure 10 is an explanatory diagram of the composite table. Figure 10 shows the gradation target 901, base table 725, modification table 726, and modified gradation correction table 1001 (composite table). The table creation unit 723 sets the created gradation correction table 1001 in the gradation correction unit 708. The gradation correction unit 708 converts the input image signal into image data using the gradation correction table 1001.

(First Example)
In the first embodiment, image adjustment control (image density correction control) is performed when restarting image formation after completion. Here, selection of a creation mode for a gradation correction table as image density correction control will be described. Figure 11 is a flowchart showing the selection process for the creation mode (image density correction control).

The printer control unit 700 acquires the current printer 20 setting information from the engine controller 711 (S1100). In this embodiment, the printer control unit 700 acquires the idle time t during which the printer 20 has not been operating since the end of the previous image formation to the present, and the current temperature dnow, as setting information. The printer control unit 700 acquires the previous setting information to store in RAM 713 (S1101). The printer control unit 700 compares the current setting information with the previous setting information. In this embodiment, the printer control unit 700 acquires the threshold value T for the idle time and the temperature d at the end of the previous image formation as the comparison results.

The printer control unit 700 compares the idle time t contained in the current setting information with the threshold value T (S1102). The printer control unit 700 also compares the temperature dnow with the temperature d (S1103). The comparison of the temperature dnow with the temperature d is a comparison of whether the temperature difference (|d - dnow|) between the time when the previous image formation ended and the time when the current image formation ends is equal to or greater than a predetermined threshold value D. In this embodiment, the temperature that constitutes the threshold value D is 5°C. The creation mode is determined based on the comparison results of the processes in S1102 and S1103. Note that in the conditional branching of the processes in S1102 and S1103, the setting range for the threshold value T is preferably 6 hours or more, and the setting range for the threshold value D is preferably 3°C or more.

If the unused time t is equal to or greater than the threshold value T (S1102: t≧T), the printer control unit 700 selects the first mode as the creation mode (S1104). In other words, if the unused time is equal to or greater than a predetermined time, the image density correction control performs a first calibration involving the formation of a pattern image and the detection of the pattern image.
If the standing time t is less than the threshold value T (S1102: t<T) and the temperature difference |d-dnow| is equal to or greater than the threshold value D (S1103: |d-dnow|≧D), the printer control unit 700 selects the second mode as the production mode (S1105). In other words, if the standing time is less than a predetermined time and the temperature difference is equal to or greater than a predetermined temperature difference, the image density correction control performs the second calibration using 10 predicted densities without forming a pattern image.
If the left-standing time t is less than the threshold value T (S1102: t<T) and the temperature difference |d-dnow| is less than the threshold value D (S1103: |d-dnow|<D), the printer control unit 700 selects the third mode as the production mode (S1106). In other words, if the left-standing time is less than the predetermined time and the temperature difference is less than the predetermined temperature difference, the image density correction control performs the third calibration using nine predicted densities without forming a pattern image.

In this embodiment, the standing time and temperature are used for conditional branching to select the creation mode. By selecting the creation mode based on the standing time and temperature, the following effects can be achieved.

When the image forming apparatus 100 is left unused for an extended period of time, the state of each component changes significantly from the end of the previous image formation. As a result, the correction accuracy of predictive control decreases, and control that grasps the current state through actual measurement control becomes necessary. In other words, first mode main calibration (first calibration) is required.

Even if the device has not been left unused for a long period of time, if the environmental conditions, particularly temperature, have changed significantly since the end of the previous image formation, the charging characteristics of the photosensitive drum 201 and chemical materials such as toner will change significantly. In this case, the gradation, including the maximum density during image formation, will also be affected. For this reason, second-mode predictive calibration (second calibration) is selected, which incorporates a model that can predict characteristic changes due to environmental fluctuations. By performing second-mode predictive calibration, it is possible to correct the gradation, including the maximum density.

If the printer has not been left unused for a long period of time and there are no significant changes in environmental conditions, the charging characteristics described above will not change significantly. Therefore, image quality can be maintained by performing third-mode predictive calibration (third calibration) and correcting only the gradation.

The effects of the first embodiment were verified under the following conditions. As a preliminary test, 2,000 full-color A4 images with an image coverage rate of 7% were printed, and after the printing was complete, an image in which the gradation could be confirmed was output. Then, a full image correction process was performed under the following conditions 1 to 3, an image in which the gradation could be confirmed was output, and the color difference ΔE76 from the image output in the preliminary test, the control time (downtime), and the toner consumption were calculated. Note that, for comparative verification, the full image correction process was applied under each of conditions 1 to 3, and the threshold T was set to 8 hours and the threshold D to 5°C for the condition branching settings of the process in FIG. 11 .
・Condition 1
Time left: 10 hours Temperature inside the device after preparation: 25°C Temperature inside the device before correction: 25°C
・Condition 2
Time left: 2 hours Temperature inside the device after preparation: 25°C Temperature inside the device before correction: 15°C
・Condition 3
Time left: 2 hours Temperature inside the device after preparation: 25°C Temperature inside the device before correction: 28°C

Condition 1 is the first mode, condition 2 is the second mode, and condition 3 is the third mode, which are the optimal creation modes for suppressing density changes. Figure 12 is a table showing the confirmation results for each mode under each condition.

It has been confirmed that under condition 1, the first mode, which performs main calibration, is able to suppress the occurrence of color differences more effectively than the second and third modes, which perform predictive calibration. This confirms the validity of the selection made through condition branching.
Under condition 2, it is confirmed that the second mode, which performs predictive calibration, can ensure color difference accuracy equivalent to that of the first mode, while also reducing control time and toner consumption compared to the first mode. This confirms the validity of the selection based on conditional branching.
Under condition 3, it is confirmed that the third mode, which performs predictive calibration, can ensure color difference accuracy equivalent to that of the first and second modes, and can reduce the control time more than the first mode. Therefore, the validity of the selection by condition branching can be confirmed.
As described above, it can be seen that the effect of applying the selection of the correction process according to the conditional branch shown in FIG. 11 is manifested by the three conditions.

As described above, when transitioning from a non-image forming state to an image forming state, the image forming device 100 can select the image correction control to be performed immediately after the transition based on the state of the printer 20 immediately after the transition and the state of the printer 20 from before the transition to the image forming state. Non-image forming states include, for example, a power-off state, a standby state, a sleep state, and a degenerate state. An image forming device 100 with this configuration can shorten the time until image formation after a state transition and maintain an optimal printer 20 state.

In this embodiment, temperature change is used as the basis for the second conditional branch, but the basis for the branch is not limited to this, and environmental information such as humidity or absolute moisture content may also be set for the conditional branch. By using factors that are sensitive to fluctuations in image density as the basis for the conditional branch, it is possible to ensure the accuracy of image density prediction.

(Second Example)
In the second embodiment, image adjustment control (image density correction control) is performed when restarting image formation after completion. Here, the selection of a gradation correction table creation mode as image density correction control will be described. In the second embodiment, the decision criteria for the second conditional branch are different from those in the first embodiment. In the second embodiment, humidity information is used as the decision criteria for the second conditional branch. In this case, too, a factor sensitive to fluctuations in image density is set in the conditional branch, allowing for optimization of the selection of the correction process. Figure 13 is a flowchart showing the selection process for the creation mode (image density correction control).

The printer control unit 700 acquires the current printer 20 setting information from the engine controller 711 (S1200). In this embodiment, the printer control unit 700 acquires the setting information including the idle time t from the end of the previous image formation to the present and the current humidity hnow. The printer control unit 700 acquires the previous setting information to be stored in RAM 713 (S1201). The printer control unit 700 compares the current setting information with the previous setting information. In this embodiment, the printer control unit 700 acquires the threshold value T for the idle time and the humidity h at the end of the previous image formation as the comparison results.

The printer control unit 700 compares the standing time t contained in the current setting information with the threshold value T (S1202). The printer control unit 700 also compares the humidity dnow with the humidity h (S1203). The comparison of humidity hnow with humidity h determines whether the humidity difference (|h-hnow|) between the time of completion of the previous image formation and the time of completion of the current image formation is equal to or greater than a predetermined threshold value H. In this embodiment, the humidity that constitutes threshold value H is 10%. The creation mode is determined based on the comparison results of the processes in S1202 and S1203. In the conditional branching of processes in S1202 and S1203, the setting range for threshold value T is preferably 6 hours or more, and the setting range for threshold value H is preferably 10% or more.

If the unused time t is equal to or greater than the threshold value T (S1202: t≧T), the printer control unit 700 selects the first mode as the creation mode (S1204). In other words, if the unused time is equal to or greater than a predetermined time, the image density correction control performs a first calibration involving the formation of a pattern image and the detection of the pattern image.
If the standing time t is less than the threshold value T (S1202: t<T) and the humidity difference |h-hnow| is equal to or greater than the threshold value H (S1203: |h-hnow|≧H), the printer control unit 700 selects the second mode as the production mode (S1205). In other words, if the standing time is less than a predetermined time and the temperature difference is equal to or greater than a predetermined humidity difference, the image density correction control performs the second calibration using 10 predicted densities without forming a pattern image.
If the standing time t is less than the threshold value T (S1202: t<T) and the humidity difference |h-hnow| is less than the threshold value H (S1203: |h-hnow|<H), the printer control unit 700 selects the third mode as the production mode (S1206). In other words, if the standing time is less than the predetermined time and the temperature difference is less than the predetermined humidity difference, the image density correction control performs the third calibration using nine predicted densities without forming a pattern image.

In this embodiment, the standing time and humidity are used for conditional branching to select the creation mode. By selecting the creation mode based on the standing time and humidity, the following effects can be achieved.

When the image forming apparatus 100 is left unused for an extended period of time, the state of each component changes significantly from the end of the previous image formation. As a result, the correction accuracy of predictive control decreases, and control that grasps the current state through actual measurement control becomes necessary. In other words, first mode main calibration (first calibration) is required.

Even if the device has not been left unused for a long period of time, if the environmental conditions, particularly humidity, have changed significantly since the end of the previous image formation, the charging characteristics of the photosensitive drum 201 and chemical materials such as toner will change significantly. In this case, the gradation, including the maximum density during image formation, will also be affected. For this reason, second-mode predictive calibration (second calibration) is selected, which incorporates a model that can predict characteristic changes due to environmental fluctuations. By performing second-mode predictive calibration, it is possible to correct the gradation, including the maximum density.

If the printer has not been left unused for a long period of time and there are no significant changes in environmental conditions, the charging characteristics described above will not change significantly. Therefore, image quality can be maintained by performing third-mode predictive calibration (third calibration) and correcting only the gradation.

The effects of the second embodiment were verified under the following conditions. As a preliminary test, 2,000 full-color A4 images with an image coverage rate of 7% were printed, and after the printing was complete, an image in which the gradation could be confirmed was output. Then, a full image correction process was performed under the following conditions 1 to 3, an image in which the gradation could be confirmed was output, and the color difference ΔE76 from the image output in the preliminary test, the control time (downtime), and the toner consumption were calculated. Note that, for comparative verification, the full image correction process was applied under each of conditions 1 to 3, and the threshold T was set to 8 hours and the threshold H to 15% for the condition branching settings of the process in FIG. 13 .
・Condition 1
Leaving time: 10 hours Humidity inside the unit after preparation: 50% Humidity inside the unit before correction: 50%
・Condition 2
Time left: 2 hours Humidity inside the unit after preparation: 50% Humidity inside the unit before correction: 80%
・Condition 3
Time left: 2 hours Humidity inside the unit after preparation: 50% Humidity inside the unit before correction: 45%

Condition 1 is the first mode, condition 2 is the second mode, and condition 3 is the third mode, which are the creation modes that are optimal for suppressing density changes. Figure 14 is a table showing the confirmation results for each mode under each condition.

It has been confirmed that under condition 1, the first mode, which performs main calibration, is able to suppress the occurrence of color differences more effectively than the second and third modes, which perform predictive calibration. This confirms the validity of the selection made through condition branching.
Under condition 2, it is confirmed that the second mode, which performs predictive calibration, can ensure color difference accuracy equivalent to that of the first mode, while also reducing control time and toner consumption compared to the first mode. This confirms the validity of the selection based on conditional branching.
Under condition 3, it is confirmed that the third mode, which performs predictive calibration, can ensure color difference accuracy equivalent to that of the first and second modes, and can reduce the control time more than the first mode. Therefore, the validity of the selection by condition branching can be confirmed.
As described above, it can be seen that the effect of applying the selection of the correction process according to the conditional branch shown in FIG. 13 is manifested by the three conditions.

As described above, when the image forming device 100 transitions from a non-image formation state to an image formation state, it can select the image correction control to be performed immediately after the transition based on the state of the printer 20 immediately after the transition and the state of the printer 20 from before the transition to the transition. An image forming device 100 configured in this way can shorten the time until image formation after the transition and maintain the optimal state of the printer 20.

In this embodiment, humidity change is used as the basis for the second conditional branch, but the basis for the decision is not limited to this, and environmental information such as absolute moisture content may also be set in the conditional branch. By using a factor that is sensitive to fluctuations in image density as the basis for the conditional branch, it is possible to ensure the accuracy of image density prediction.

(Third Example)
In the processes of the first and second embodiments, a conditional branch may be selectable or unselectable depending on the usage environment and usage conditions of the image forming apparatus 100. Depending on the user's usage environment, the variety of jobs due to various print settings, and sensitivity to color design preferences, the selected process may not optimize the correction process. For this reason, it is preferable that the image density correction control be selectable on the image forming apparatus 100. In other words, by being able to enable/disable the correction process (image density correction control), more suitable image density control is possible. Furthermore, it is preferable that the image density correction control be disabled if the user determines that predictive calibration is not operating normally under unexpected circumstances, including a malfunction of the image forming apparatus 100.

In other words, by allowing the user to select whether to enable or disable image density correction control depending on the usage situation and the end product, it is possible to provide a higher quality end product. In this embodiment, the selection of whether to enable or disable image density correction control is performed by the operation unit 140. The selection information input by the operation unit 140 is sent to the estimation processing unit 505 (see Figure 5).

Figure 15 is an explanatory diagram of the operation unit 140. In this embodiment, the operation unit 140 is configured by combining a display 1301, a touch panel 1302, and key buttons. The key buttons include a setting key 1303, a power saving key 1304, a group of hard keys 1305, a reset key 1306, a stop key 1307, and a start key 1308. An operation screen is displayed on the display 1301 under the control of the CPU 1201. In this embodiment, a setting screen for selecting whether to enable or disable image density correction control is displayed on the display 1301.

By operating the key buttons or software keys displayed on the operation screen, information corresponding to the operated key is sent to the CPU 1201 via the operation unit I/F 1206. The start key 1308 is used to issue instructions to start processes such as copying and printing. The start key 1308 is equipped with two LEDs (light-emitting diodes) in green and red (not shown). The LED indicates that the process can be started when lit green, and that the process cannot be started when lit red. The stop key 1307 is used to stop an ongoing operation. The hard key group 1305 includes a numeric keypad, a clear key, and an authentication key.

The power saving key 1304 is used to transition the image forming apparatus 100 to sleep mode or to return from sleep mode. The image forming apparatus 100 transitions to sleep mode when the power saving key 1304 is pressed in normal mode, and transitions to normal mode when the power saving key 1304 is pressed in sleep mode. The setting key 1303 is used to configure settings such as AI function settings. The operation unit 140 is also used to input information required for creating job information, such as the user name, number of copies to be printed, and output attribute information.

Figure 16 is an example of a settings screen displayed on the display 1301. The settings screen displays a selection screen for image density correction control (here, gradation correction processing). The selection screen is displayed, for example, by pressing a software key to start selecting image density correction control from the menu screen, which is the initial screen.

The selection screen displays buttons that allow the user to select whether to enable or disable each of the multiple image density correction controls (first to third gradation correction processes). Selection buttons 1401 are displayed corresponding to the first to third gradation correction processes. The user selects a setting for each trained model (a button that allows the user to select whether to enable or disable) and presses the confirm button 1402, thereby selecting whether to enable or disable each of the first to third gradation correction processes. The selection results are sent to CPU 1201. CPU 1201 saves the selection results of whether to enable or disable the multiple gradation correction processes in RAM 1202 depending on the selection results. CPU 1201 executes control based on the contents saved in RAM 1202.

Figure 17 is a combination table for selecting whether to enable image density correction control (first to third gradation correction processes). As shown in the combination table, there are eight combinations for enabling or disabling the first to third gradation correction processes (combination 1 to combination 7, unselectable). In this embodiment, the combination that disables all processes is unselectable.

This combination table may be displayed on the display 1301 so that the user can view it. By selecting from the combination table, the user can enable or disable image density correction control. This makes it possible to select the optimal image density correction control depending on the situation, thereby producing higher quality output.

The settings made on the selection screen in Figure 16 can be limited to those specific to the user, or can be selected to be applicable to other users. Furthermore, image density correction control settings on the selection screen may be restricted to administrators only. In this case, for example, it is possible to prevent incorrect settings from being selected.

Figure 18 is a flowchart showing the selection process for image density correction control. The CPU 1201 displays the setting screen shown in Figure 16 on the display 1301 and acquires the selection results. The CPU 1201 performs the following processing based on the selection results.

The CPU 1201 determines whether the selection results indicate that all image density correction controls (tone correction processes) are to be deactivated (S1500). If all image density correction controls (tone correction processes) are to be deactivated (S1500: Y), the CPU 1201 displays an alarm screen on the display 1301 prompting the user to select activation (S1501).

If at least one image density correction control (tone correction process) is selected to be enabled (S1500: N), the CPU 1201 checks the combination pattern of enabled/disabled for each image density correction control (first to third tone correction process) (S1502). The CPU 1201 performs exclusive processing for the image correction control that has been disabled based on the image correction control selection results from the end of the previous image formation until the restart (S1503). The CPU 1201 executes the image density correction control that is not excluded (enabled tone correction process) (S1504).

As described above, by switching the image density correction control between enabled and disabled depending on the user's usage conditions, etc., it is possible to provide higher quality deliverables. Note that while there are three image density correction controls in this embodiment, this is not limited to this and there may be four or more. Furthermore, the enable/disable setting can be set by the general-purpose computer 103 in addition to the operation unit 140. In this case, the setting screen in Figure 16 is displayed on a display provided on the general-purpose computer 103. This makes it possible to remotely change whether the trained model is enabled or disabled from a location other than the installation location of the image forming device.

The system of this embodiment has been described above as adjusting image density, but it can also be used for general control of image quality. For example, the system of this embodiment is also effective for adjusting the geometric characteristics of an image, such as the print position of the image on paper and the inclination of the image, as well as adjusting color misalignment.

Claims (6)

an image forming means for forming an image on a sheet;
a reading means for reading a pattern image formed on the paper;
a control means for executing an actual measurement control for correcting image forming conditions based on the result of reading the pattern image by the reading means , a first predictive control for correcting image densities of all gradations including the maximum density based on image densities predicted by environmental information of the image forming means, and a second predictive control for correcting image densities of all gradations excluding the maximum density based on image densities predicted by environmental information of the image forming means ,
When the image forming means shifts from a non-image forming state in which no image is formed to an image forming state in which an image is formed , if the time period of the non-image forming state is equal to or longer than a predetermined time period, if the time period of the non-image forming state is less than the predetermined time period and the temperature difference between the temperature at the end of the previous image formation and the temperature at the time of shifting to the image forming state is equal to or longer than a predetermined temperature difference, the control means executes the actual measurement control, if the time period of the non-image forming state is less than the predetermined time period and the temperature difference between the temperature at the end of the previous image formation and the temperature at the time of shifting to the image forming state is less than the predetermined temperature difference, the control means executes the second predictive control, if the time period of the non-image forming state is less than the predetermined time period and the temperature difference between the temperature at the end of the previous image formation and the temperature at the time of shifting to the image forming state is less than the predetermined temperature difference .
Image forming device. further comprising an operation means for accepting an input from a user;
The actual measurement control, the first predictive control, and the second predictive control can be enabled and disabled by the operation means.
2. The image forming apparatus according to claim 1. The actual measurement control, the first predictive control, and the second predictive control cannot all be disabled.
3. The image forming apparatus according to claim 2. an image forming means for forming an image on a sheet;
a reading means for reading a pattern image formed on the paper;
a control means for executing an actual measurement control for correcting image forming conditions based on the result of reading the pattern image by the reading means, a first predictive control for correcting image densities of all gradations including the maximum density based on image densities predicted by environmental information of the image forming means, and a second predictive control for correcting image densities of all gradations excluding the maximum density based on image densities predicted by environmental information of the image forming means,
When the image forming means shifts from a non-image forming state in which no image is formed to an image forming state in which an image is formed, if the time period of the non-image forming state is equal to or longer than a predetermined time period, the control means executes the actual measurement control if the time period of the non-image forming state is less than the predetermined time period and the humidity difference between the humidity at the end of the previous image formation and the humidity at the time of shifting to the image forming state is equal to or longer than a predetermined humidity difference, and executes the second predictive control if the time period of the non-image forming state is less than the predetermined time period and the humidity difference between the humidity at the end of the previous image formation and the humidity at the time of shifting to the image forming state is less than the predetermined humidity difference.
Image forming device. further comprising an operation means for accepting an input from a user;
The actual measurement control, the first predictive control, and the second predictive control can be enabled and disabled by the operation means.
5. The image forming apparatus according to claim 4. The actual measurement control, the first predictive control, and the second predictive control cannot all be disabled.
6. The image forming apparatus according to claim 5 .