JP3673558B2 - Image processing method and apparatus - Google Patents

Description

[0001]
[Industrial application fields]
The present invention relates to an image processing method and apparatus, and more particularly to an image processing method and apparatus for converting the pixel density when the output pixel density is higher than the pixel density of input image data.
[0002]
[Prior art]
In facsimile machines using thermal and thermal transfer recording, the main scanning resolution of the CCD (solid-state image sensor) or CS (contact sensor) as the image reading device and the main scanning resolution of the thermal head as the recording device are both ITU- It consists of 203.3dpi (equivalent to 8pel / mm), which is the T standard.
[0003]
[Problems to be solved by the invention]
However, the above-described technique has the following problems.
In recent years, the recording method has changed from the thermal method and the thermal transfer method to the LBP (laser beam printer) recording method and the ink jet recording method, and a low-cost and high-resolution recording device is a facsimile recording device. Came to be used as. For example, the mainstream of recording resolution is 300,400,600 dpi for LBP recording and 360 dpi for inkjet recording, but the reading resolution is low, so a sufficiently high recording resolution cannot be effectively utilized.
[0004]
In addition, the facsimile apparatus has been shifted from a dedicated facsimile apparatus to one having a full copy function. From this aspect, the demand for higher image quality is increasing.
[0005]
That is, it is necessary to convert the resolution of image data read in the main scanning direction 203.3 dpi in accordance with a high-resolution recording apparatus. However, this resolution conversion sometimes deteriorates the image quality. For example, if you want to send, select the halftone mode as the reading mode and select the fine mode as the reading resolution, then scan the original image at a resolution of 203.3 dpi in the main scanning direction and 195.58 dpi in the sub-scanning direction. A pseudo halftone process is performed by the average error minimum method or the like, and binarized and transmitted. In the copy mode, binarization is performed in the same way, but when resolution conversion to high resolution is performed according to the resolution of the recording device, the pattern due to halftone processing such as error diffusion method collapses, and stripes become conspicuous. Degradation of image quality will occur.
[0006]
In addition, there is a problem that if a high-resolution reading device is provided in accordance with the resolution of the recording device, it becomes expensive.
[0007]
SUMMARY An advantage of some aspects of the invention is to record a high-quality image utilizing the resolution of a recording apparatus.
[0008]
Another object of the present invention is to prevent image quality deterioration associated with resolution conversion.
[0009]
[Means for Solving the Problems]
The present invention has the following configuration as one means for achieving the above object.
[0010]
The image processing method of the present invention is a transmitting means for transmitting image data of a predetermined resolution, reads an original image, and outputs the image data representing the original image as the multi-valued data per pixel with the predetermined resolution. An image processing method for an image processing apparatus comprising: a reading unit; and a recording unit that records an image based on binarized image data output from the reading unit at a recording resolution different from the predetermined resolution. When a copy mode for recording an image based on the image data output by the reading unit is selected by the recording unit, the image data output by the reading unit is converted into the recording resolution and then binarized, and the reading is performed. When the transmission mode for transmitting the image data output by the means to the transmission means is selected, the resolution conversion is skipped and the reading means outputs Characterized by binarizing the image data.
[0011]
An image processing apparatus according to the present invention includes a selection unit that selects a mode, a transmission unit that transmits image data having a predetermined resolution, an original image that is read, and image data that represents the original image is stored at the predetermined resolution. Reading means for outputting as multivalued data per pixel, conversion means for converting the resolution of image data output by the reading means, and binarization means for converting image data output by the reading means into binary data , An image based on the image data binarized by the binarization unit, a recording unit that records the image at a second resolution different from the first resolution, and an image output by the reading unit by the selection unit When a copy mode in which an image based on data is recorded by the recording unit is selected, the image data output by the reading unit differs from the predetermined resolution by the converting unit. When the transmission mode for converting the recording resolution to binarization by the binarization unit and the selection unit transmitting the image data output by the reading unit to the transmission unit is selected, the resolution by the conversion unit Control means for skipping the conversion and binarizing the image data output from the reading means by the binarizing means.
[0016]
【Example】
Hereinafter, an image processing apparatus according to an embodiment of the present invention will be described in detail with reference to the drawings. In the following, a facsimile apparatus provided with an image processing apparatus according to an embodiment of the present invention will be described. However, the present invention is not limited to a facsimile apparatus, and can be applied to any image processing apparatus that handles image data. Needless to say, you can.
[0017]
First, the pixel density conversion will be briefly described.
[0018]
Pixel density conversion is to increase or decrease the number of pixels of input image data in accordance with the resolution of the output device, and is basically the same as image enlargement / reduction. For example, image data of A4 size with 1728 x 2287 pixels, resolution of main scanning 203.2dpi (equivalent to 8pel / mm), sub-scanning 195.58dpi (equivalent to 7.7line / mm), main scanning 406.4dpi (16pel / mm) Equivalent), and sub-scanning 391.16dpi (equivalent to 15.4line / mm) resolution to convert the pixel density to 4 pixels by converting one pixel to 4456 pixels. Is exactly the same as performing a double enlargement process in each scanning direction. In the following description, processes such as resolution conversion, enlargement / reduction, and pixel density conversion are all treated as digital enlargement / reduction processes.
[0019]
Next, it will be described at which stage of image data processing in the facsimile apparatus the pixel density conversion is performed.
[0020]
Normally, G3 and G4 facsimiles handle binary data in the storage and transmission / reception of image data. Therefore, since the image data already stored in the memory as an image file or the image data received through the line is a binary image, the binary image data is subjected to binary conversion to perform pixel density conversion or enlargement / reduction. The technique for images will be used. As these methods, the SPC method, the nine division method, the projection method, and the like have been proposed. In this embodiment, processing such as pixel density conversion is performed on multi-valued image data before being stored in an image file as facsimile image data, that is, before being binarized.
[0021]
Further, the binarization method is intended for a case where a halftone process such as an error diffusion method or an average error minimum method for expressing a pseudo gradation is performed.
[0022]
A pseudo gradation method using a binary image using error diffusion or the like expresses the density as a ratio of the number of black and white dots per unit area. For example, when expressing a density of 16 in 64 gradations, an average of 16 black dots are present in an 8 × 8 pixel block due to a pattern by halftone processing. The error diffusion method, which diffuses the error between the level of the input image data and the predetermined threshold two-dimensionally, is a binarization method that has both resolution and gradation compared to the dither method. When the pixel size representing one pixel is large, that is, when the recording resolution is relatively low like G3 facsimile, a unique stripe pattern due to a periodic dot pattern becomes conspicuous.
[0023]
Therefore, in order to take advantage of the error diffusion method, it is necessary to perform binarization after performing pixel density conversion before binarization processing by the error diffusion method to obtain high resolution data. However, conventionally, even in a facsimile machine equipped with a recording device with a resolution of 406.4 dpi, binarization processing is performed at a resolution of 203.2 dpi in all cases including copying, because of the relationship with transmission / reception data. Doubled and recorded.
[0024]
Therefore, in this embodiment, when it is known in advance that the recording apparatus has a high resolution, such as copying, an error diffusion method is performed after pixel density conversion is performed at the stage of multi-value data before binarization. By binarizing, the display resolution of error diffusion is increased and the size of one pixel for density display is reduced. This makes it possible to obtain a printed image in which the stripe pattern unique to the error diffusion method is not conspicuous. In addition, since the pattern caused by the error diffusion process is not destroyed, deterioration of the image quality can be prevented.
[0025]
Fig. 1 is a sample of a binarized image by error diffusion when the recording resolution is different between the main and the sub. Get a good impression.
[0026]
In addition, FIG. 2 is an enlarged view of a part of a line drawing having an edge portion, but the image shown in FIG. 2 (b) is error-diffused after pixel density conversion, compared to FIG. 2 (a). The notch in the edge portion is reduced, and an image having a sharp edge can be obtained.
[0027]
In this embodiment, pixel density conversion is performed after edge enhancement as resolution compensation is performed on image data read by a CCD. Here, the edge enhancement processing is to add the image data subjected to two-dimensional second-order differentiation to the image data of interest. The reason for performing pixel density conversion after edge enhancement is that if pixel density conversion is performed as a simple overlap process and then edge enhancement is performed, a high-frequency component is superimposed by Laplacian calculation in the image space that has been subjected to overlap processing. This is because a pseudo contour is generated at the edge portion.
[0028]
Fig. 3 shows an example of an image with pseudo contours. A square all-black image is doubled by double-enlarging both main scanning and sub-scanning by simple overlap processing and then edge-enhanced. There is a state in which a pseudo contour is generated.
[0029]
FIG. 4 is a diagram for explaining a process in which a pseudo contour is generated. In order to simplify the explanation, image data is considered here in one dimension. Also, the coefficient of the edge enhancement filter is [-1,2, -1] as shown in FIG.
[0030]
The upper part of FIG. 4 shows the result when edge enhancement processing is performed with the filter of FIG. 5 after double enlargement by simple overlap processing. The lower part shows the result when the simple overlap processing is performed after the edge enhancement processing by the same filter. When the binarization threshold is set to 30, in the process shown in the upper part of FIG. 4, the image data D2 at the edge portion becomes white by the edge enhancement process, and the image data D2 ′ generated by the overlap process becomes black. This result appears as a pseudowire. On the other hand, such a problem does not occur in the processing shown in the lower part.
[0031]
For the above reasons, in this embodiment, when it is known in advance that the output resolution is high, the input multi-valued image data is binarized in the next processing flow.
[0032]
(1) Perform edge enhancement for resolution compensation on input multi-valued image data
[0033]
(2) Perform pixel density conversion according to the output resolution for edge-enhanced image data
[0034]
(3) Binarize the pixel density converted image data using error diffusion method etc.
[0035]
In the following, an example in which luminance density conversion is performed after edge enhancement will be described. As is apparent from the above description, luminance density conversion processing is performed between edge enhancement processing and binarization processing. It may be performed after the enlargement process.
[0036]
Next, the amount of image data will be described.
[0037]
As described above, in order to improve image quality, it is desirable to perform pixel density conversion before binarization, but on the other hand, the image data after binarization is high-resolution image data. There is a drawback that the amount of data increases because the data is output. In particular, when image data is encoded and stored by an encoding method such as MH, MR, or MMR, the image data obtained by error diffusion processing increases conversely due to encoding, and therefore the amount of data becomes a problem. .
[0038]
Therefore, in this embodiment, in order to alleviate the problem of an increase in the amount of data compared to a general facsimile machine or the like, the output resolution is high, for example, when copying or by direct transmission that starts reading after recognizing the partner machine using a communication protocol. Only when the resolution is known in advance, pixel density conversion is performed before binarization.
[0039]
Hereinafter, the flow of image data and a specific configuration example of the binarization processing unit in the present embodiment will be described in detail.
[0040]
FIG. 6 is a block diagram illustrating a configuration example of a facsimile machine including the image processing apparatus according to the embodiment of the present invention.
[0041]
In the figure, reference numeral 100 denotes a CPU, which performs overall control of the present embodiment, image processing, and the like via a system bus 260 in accordance with a program stored in the ROM 110. In addition, information to be displayed on the operation unit 130 and font data and display data for report output are recorded in the ROM 110 in advance. A RAM 120 is used for storing work data of the CPU 100, storing image files such as encoded image data and received image data.
[0042]
Reference numeral 130 denotes an operation unit that functions as a user interface of the facsimile apparatus of the present embodiment. The operation unit 130 includes a display device such as an LCD, a keyboard, and a touch panel, and is sent from the CPU 100 to the operation state of the facsimile apparatus of the present embodiment. And setting conditions are displayed, and instructions from the user are input to the CPU 100.
[0043]
Various modes can be selected by the operation unit 130. That is, a copy mode and a transmission mode can be selected. In the copy mode, a single copy mode for performing one copy and a multi-copy mode for performing a plurality of copies can be selected for one original. it can. Further, in the transmission mode, a direct transmission mode in which a document is transmitted while reading and a memory transmission mode in which all transmission documents are temporarily stored in a memory and then transmitted can be selected.
[0044]
Further, the reading mode (image quality) can be selected by the operation unit 130, and a character mode for reading a document composed of characters and figures and a photo mode (halftone mode) for reading a halftone document such as a photograph are selected. be able to. In the character mode, a simple binarization method is used in which the resolution of the read image data is converted as necessary, and each pixel is compared with a predetermined threshold value, and binarization is performed according to the magnitude. In the photo mode, the resolution of the read image data is converted as necessary, compared with a predetermined threshold value, and the pixel of interest is binarized according to its magnitude, and the error generated here is converted into a predetermined weighting matrix. A well-known error diffusion method that diffuses according to the above is used.
[0045]
Further, the resolution (reading pitch) in the sub-scanning direction when reading the document image can be selected by the operation unit 130. In other words, it is possible to select 97.79dpi (equivalent to 3.85line / mm) which is the standard mode, 195.58dpi (equivalent to 7.7line / mm) which is the fine mode, and 391.16dpi (equivalent to 15.4line / mm) which is the superfine mode. . As a special case, transmission in the ultra fine mode 406.4 dpi × 391.16 dpi is also possible.
[0046]
In this embodiment, the case where the photo mode is selected will be described.
[0047]
Also, the resolution at the time of reading the document is 203.2 dpi in the main scanning direction and one of the above three types in the sub-scanning direction according to the operator's selection, but when the photo mode and the memory transmission mode are selected, the standard mode ( Only two types (203.2 dpi × 97.79 dpi) or fine mode (203.2 dpi × 195.58 dpi) can be selected.
[0048]
In general, a facsimile apparatus can receive and record an image transmitted in the standard mode and the fine mode, but cannot receive and record an image transmitted in the superfine mode. For this reason, if the original facsimile machine does not support the super fine mode when it is going to be scanned and stored in the memory in the super fine mode and stored in the memory, it will be thinned out on the transmission side. Resolution conversion is performed by processing or the like. However, if resolution conversion, such as thinning-out processing, is performed on an image binarized by halftone processing, the pattern by halftone processing is destroyed and image quality is degraded. Therefore, if only the standard mode and the fine mode can be selected, resolution conversion is not performed after binarization, and deterioration of image quality can be prevented. Furthermore, since halftone images have relatively poor encoding efficiency, the use of the superfine mode can reduce the amount of data and save memory usage.
[0049]
Note that the reading resolution in the copy mode may be set in advance or may be selectable by the operator.
[0050]
140 is a MODEM that modulates the encoded image data and demodulates the received data. An NCU 150 interfaces the line 280 with the facsimile apparatus of this embodiment.
[0051]
Reference numeral 170 denotes a motor controller, which drives and controls a motor 160 for moving a document in accordance with an instruction from the CPU 100. For example, the motor 160 is a stepping motor, and the motor controller 170 sends a phase switching control signal to control the motor 160. Thereby, the reading pitch (resolution) in the sub-scanning direction is controlled.
[0052]
Reference numeral 180 denotes a CCD (charge coupled device) composed of a plurality of light receiving elements, which reads an image of a document line by line at a resolution of 203.2 dpi at the time of transmission or copying, and outputs an image signal. Reference numeral 190 denotes an analog correction processing unit which performs shading correction and correction processing on the background density of the original on the image signal input from the CCD 180, and a 64-bit digital image of 6 bits / pixel, for example, by an A / D converter (not shown). Output a signal.
[0053]
A binarization processing unit 200 binarizes a 6-bit / pixel digital image signal having a resolution of 203.2 dpi input from the analog correction processing unit 190 and outputs the binarized image signal to the image bus 270. In addition, when the recording resolution such as copying is known in advance, the binarization processing unit 200 performs pixel density conversion in accordance with the recording resolution. For example, binarization is performed at a resolution of 203.2 dpi during normal memory transmission, and binarization is performed after pixel density conversion is performed at a resolution of 406.4 dpi according to the recording resolution during copying.
[0054]
A RAM 210 is connected to the image bus 270, and is used as a line buffer for temporarily storing binarized image data and decoded image data. An LBP (laser beam printer) 220 prints out the binary image data expanded in the RAM 210. Reference numeral 230 denotes an encoder, which inputs and encodes the image data expanded in the RAM 210, and then transfers the encoded data to the RAM 120 via the system bus 260. Reference numeral 240 denotes a decoder. The encoded data stored in the RAM 120 is input via the system bus 260 and decoded, and then the decoded image data is developed in the RAM 210. If the encoded data already stored in the RAM 120 is different from the encoding method to be transmitted, such as when trying to transmit the image encoded and stored by the MMR method using the MH method, Encoder 230, decoder 240, and RAM 210 are used to perform encoding conversion according to the encoding method.
[0055]
A bus arbiter 250 arbitrates between the image bus 270 and the system bus 260 when the CPU 100 expands the header image of the transmission image or the footer image to be added to the RAM 210.
[0056]
Next, a flow of processing related to reading of a document such as copying and transmission will be described with reference to flowcharts shown in FIGS. These processes are executed based on instructions from the CPU 100.
[0057]
First, it is determined whether or not the instruction given through the operation unit 130 in step S701 is a copy instruction. If it is a copy instruction, in step S702, it is determined whether or not a multi-copy instruction for copying a plurality of the same document is instructed. Here, if it is determined that the multi-copy is not instructed, it is a case where one copy is made, and the process proceeds to step S703, where reading by the CCD 180 is started. Here, it is assumed that the reading resolution is set to 203.2 dpi × 391.16 dpi, for example. The read image data is subjected to preprocessing such as shading correction by an analog correction processing unit 190. Next, in step S704, the binarization processing unit 200 performs binarization. In this case, since the output resolution is known, the binarization processing unit 200 converts the input image data into the recording resolution of the LBP 220. After converting the resolution to match, binarize and output. The high-resolution binary image data is temporarily stored for each line in the RAM 210, outputted in synchronization with the LBP 220, and printed on the recording paper in step S705.
[0058]
On the other hand, if it is determined in step S702 that the copy is a multi-copy, reading by the CCD 180 is started in step S706. The process up to the binarization processing of the image data read at a predetermined resolution in accordance with the recording resolution by the binarization processing unit 200 (step S707) is the same as that in the case of the single copy described above. Detailed description is omitted.
[0059]
In the case of multi-copy, since the same image data is output a plurality of times, it is necessary to store image data for one page in the apparatus. Therefore, the binary image data is encoded by the encoder 220 and an attempt is made to store one page of code data in the RAM 120. That is, when image data for one page is accumulated without causing memory overflow in step S708, the LBP 220 is activated and the decoded code data is stored in the RAM 210 while the accumulated code data is decoded by the decoder 240. To the LBP 220 and print out the first sheet in step S705. The second and subsequent sheets are output in the same manner as the first sheet using the code data stored in the RAM 120. When the desired number of copies is completed, the code data stored in the RAM 120 is cleared.
[0060]
Also, here, when the capacity of the memory for storing the code data is small or an image with poor encoding efficiency (for example, the information amount of the binarized image data increases after resolution conversion to high resolution) In this case, code data for one page may not be stored in the RAM 120. In this case, that is, when the memory is almost full (near full) in step S708, the process proceeds to step S709, and a message that multi-copy cannot be performed due to memory full is displayed on the operation unit 130. In step S710, the reading operation is temporarily stopped before the RAM 120 is in a memory full state, and the recording operation is started in step S711. Since this recording operation causes an empty space in the RAM 120, the reading operation is resumed in step S706, and these processes are repeated until one copy is completed, and one copy is performed by synchronous copy.
[0061]
If it is determined in step S701 that the copy operation is not performed, it is a transmission operation, and the process proceeds to step S801 to determine whether or not the memory transmission for transmitting the read image data once in the memory is instructed. to decide. Here, if it is determined that it is not memory transmission, this is a case where direct transmission for transmitting while reading a document image is instructed. In this case, the connection with the communication partner is first executed, and the recording resolution of the partner is confirmed by a facsimile protocol. That is, in step S802, it is determined whether or not the partner main scanning recording resolution is a high resolution such as 406.4 dpi and communication in the ultra fine mode is possible.
[0062]
If the destination main scanning recording resolution is not high resolution but 203.2 dpi (8 dots / mm), scanning is started in step S803 with the resolution (standard, fine, or superfine) selected by the operator. That is, the image data of the document read by the CCD 180 is input to the binarization processing unit 200 via the analog correction processing unit 190. The resolution in the sub-scanning direction such as standard, fine, and super fine is adjusted by the number of driving steps per line of the motor 160.
[0063]
In step S804, the binarization processing unit 200 binarizes the image data without converting the resolution. The binarized image data is temporarily stored in the RAM 210, sequentially encoded by the encoder 230, and the code data is stored in the RAM 120. Next, in step S805, the code data stored in the RAM 120 is modulated by the MODEM 140 and transmitted to the other party via the NCU 150 and the line 280.
[0064]
Then, the above process is repeated until it is determined in step S806 that the process is ended.
[0065]
On the other hand, if it is determined in step S802 that the recording resolution of the other party in the main scanning direction is high and communication in the ultra fine mode is possible, in step S807, the original image is read (main scanning direction 203.2 dpi, Sub-scanning direction 391.16 dpi) is started, and binarization is performed after the binarization processing unit 200 converts it to a high resolution in accordance with the recording resolution of the other party in step S808. Subsequently, in step S809, the binarized image data is transmitted according to the procedure described above.
[0066]
If it is determined in step S801 that the transmission is a memory, reading of the document is started in step S811. Here, since it is assumed that reading is performed in photo mode, the resolution in the sub-scanning direction that can be selected by the operator is either 97.79 dpi (standard) or 195.58 dpi (fine), and 391.16 dpi (super fine) is selected. Ban. This can be confirmed by the display of the operation unit 130. The image data read at the resolution selected by the operator is binarized without performing resolution conversion in the binarization processing unit 200 in step S812, and sequentially stored in the RAM 120. This image data is encoded by the encoder 230 and then stored in the RAM 120. The above processing is repeated until reading of all the transmitted originals is completed in step S813. When all the code data is stored in the RAM 120, in step S814, the code data is transmitted to the other party according to the procedure described above.
[0067]
Next, the reception operation will be described. Since this is a procedure similar to the conventional one, illustration is omitted.
[0068]
Code data received from the line 280 via the NCU 150 is demodulated by the MODEM 140 and temporarily stored in the RAM 120. When the code data to be output exists in the RAM 120, the CPU 100 activates the LBP 220, decodes the code data by the decoder 240, and once expands the decoded image data in the RAM 210, then the image data is transferred to the LBP 220. Send and print out.
[0069]
Here, when the resolution of the received image is different from the output resolution, pixel density conversion is performed by processing in the LBP 220 using the projection method or the SPC method.
[0070]
As described above, in the case of the copy mode in which the output resolution is known in advance or the direct transmission in which the output resolution of the communication partner is known, the binarization processing unit 200 performs binarization processing in order to maximize the output resolution. Then, pixel density conversion is performed to output image data binarized at a high resolution.
[0071]
Next, the binarization processing unit 200 will be described in detail.
[0072]
FIG. 7 is a block diagram showing a detailed configuration example of the binarization processing unit 200. Note that the binarization processing unit 200 of the present embodiment adds the conversion function by adding a small functional block to the binarization processing unit of the facsimile machine that does not have the function of converting to a high-density image. Considered to be able to add.
[0073]
In FIG. 7, 2001 is an edge enhancement unit, 2002 is a luminance density conversion unit, 2003 is a main scanning scaling unit, 2004 is an error diffusion unit, 2005 is a clock control unit including a selector and a mask processing unit, and 2006 is an OR gate. .
[0074]
FIG. 8 is a block diagram showing a detailed configuration example of the edge emphasizing unit 2001.
[0075]
In the figure, reference numeral 3001 denotes a line memory control unit, which switches the control of the line memories 3002 and 3003 for storing one line of image data in response to a start trigger issued by the CPU 100. Reference numeral 3004 denotes a latch group that stores image data in units of pixels, and 3005 denotes an edge enhancement calculation unit that performs edge enhancement filter calculation.
[0076]
Next, the operation of the binarization processing unit 200 will be described in detail.
[0077]
Image data read by the CCD 180 is subjected to analog correction processing and A / D conversion, and is input to the binarization processing unit 200 as digital image data of 6 bits / pixel 64 gradations, for example. The main scanning resolution of the input image data is determined by the CCD and the optical system and is 203.2 dpi. The input image data is input in synchronization with the line synchronization signal in units of lines and in synchronization with the falling edge of the pixel clock MCLK / 2 in units of pixels. Here, in consideration of handling with the analog correction processing unit 190, it is assumed that the image data is continuously input in synchronization with MCLK / 2.
[0078]
The image data output from the analog correction processing unit 190 is first input to the edge enhancement processing unit 2001. The edge emphasis processing unit 2001 sequentially transfers and accumulates the image data in line memories 3002 and 3003 in line units in synchronization with the line synchronization signal in accordance with the process activation triggers A and B issued from the CPU 100. Here, there are two types of start triggers A and B in order to take out the image data after overlapping edge emphasis by combining motor drive control and triggers A and B in the sub-scanning direction. The mechanism realizes line overlap processing after edge enhancement processing in the sub-scanning direction.
[0079]
When the binarization processing is activated by the trigger A, the image data output from the analog correction processing unit 190 is input to the latch E of the latch group 3004. At the same time, the image data is sequentially transferred to the line memory 3002 via the latch E.
[0080]
The line memory 3002 sequentially updates the stored contents with the input image data, and outputs the stored contents to the latch D before the stored contents are updated with new image data. Similarly, the line memory 3003 sequentially updates the stored contents with the image data input from the line memory 3002 via the latch D, and outputs the stored contents to the latch C before the update. As described above, the image data for three lines is simultaneously input to the latch group 3004 by the line buffers 3002 and 3003 of two lines.
[0081]
The latch group 3004 shifts the image data input from the analog correction processing unit 190, the image data input from the line memory 3002, and the image data input from the line memory 3003 in synchronization with the pixel clock MCLK / 2. As a result, 3 × 3 matrix data is output to the edge enhancement calculation unit 3005. Here, reference signs A to H and X given to the respective latches correspond to the reference of the edge enhancement filter shown in FIG. 9A and the pixel position of interest.
[0082]
The edge enhancement calculation unit 3005 performs edge enhancement calculation using the matrix data obtained from the latch group 3004, and outputs the result to the luminance density conversion unit 2002. The edge enhancement calculation unit 3005 performs resolution compensation by adding a result obtained by performing constant weighting on a calculation result obtained by Laplacian, which is a two-dimensional second-order differential calculation, to the target pixel X. FIG. 9B is a diagram showing an example of filter coefficients of the edge enhancement calculation unit 3005. The weighting is arbitrarily changed according to the magnification of pixel density conversion.
[0083]
On the other hand, when the binarization process is activated by the trigger B, unlike the case of the trigger A, the line memories 3002 and 3003 are not updated, and only the reading operation is performed on the line memory. Therefore, as shown in FIG. 10, for example, the same line data is read twice from the line buffer by alternately using triggers B and A for one-step document feeding by the motor 160. Thus, the data after the overlapping edge enhancement calculation can be taken out from the edge enhancement unit 2001.
[0084]
As a result, for one step of the motor 160, it is possible to output image data in which the pixel density in the sub-scanning direction is doubled after edge enhancement. In the case of a resolution higher than twice, the trigger B may be raised a plurality of times for one step of the motor 160 according to the magnification. Further, if one step in the sub-scanning direction is the same as the output resolution, by using only the trigger A, it is possible to obtain the same size image data in the sub-scanning direction. Further, in the case of reduction, the rate of start of binarization processing by the trigger A with respect to the feed of the motor 160 may be thinned according to the reduction rate.
[0085]
The luminance density conversion unit 2002 converts the image data whose resolution has been compensated by the edge enhancement unit 2001 from luminance data to density data using a conversion table. Here, the conversion table is determined based on the characteristics of the CCD 180 and the recordability of the binarized and recorded image data. Especially for recording devices such as LBP and inkjet, the recording characteristics of binary images differ depending on the resolution, so it is necessary to adjust the values to keep the recording density constant by changing the value of the luminance density conversion table according to the output resolution. is there. This conversion table is stored in advance in a ROM built in the brightness density conversion unit 2002, or a table read from the ROM 110, for example, by the CPU 100 is set in a RAM built in the brightness density conversion unit 2002. It is a thing.
[0086]
The 6-bit / pixel image data converted into density data by the brightness density converter 2002 is the main scan. Magnification In the unit 2003, the pixel density in the main scanning direction is converted at the magnification instructed by the CPU 100 by the clock control of the clock control unit 2005. Here, with reference to FIG. 11, a scaling process in which image data having a resolution of 203.2 dpi in the main scanning direction is reduced to 5/7 times at a resolution of 406.4 dpi, that is, a magnification process for enlarging to 10/7 times will be described.
[0087]
The pixel density conversion in the main scanning direction in this embodiment is realized by clock thinning, and is based on a simple pixel overlap process. Here, the magnification control is performed by controlling the input pixel clock MCLK / 2 and the output pixel clock CLK2.
[0088]
When 100 to 200% scaling processing is instructed by the CPU 100, the selector 2005a in the clock control unit 2005 selects MCLK that is twice the frequency of MCLK / 2 and sends MCLK to the mask processing unit 2005b. Supply. When a magnification less than 100% is instructed, MCLK / 2 which is the same as the input pixel rate of the binarization processing unit 200 is supplied to the mask processing unit 2005b. The mask processing unit 2005b performs a mask operation corresponding to the set magnification on the input clock to generate CLK2. Generated CLK2 is main scan Magnification Is supplied to the unit 2003 and the error diffusion unit 2004 as a drive clock.
[0089]
As shown in Figure 11, main scan Magnification In the unit 2003, the image data D0, D1, D2,... Transferred in synchronization with the falling edge of MCLK / 2 are sampled again at the rising edge of CLK2, so that the image data D00, D01, D10, D20 , Is converted to….
[0090]
The image data whose pixel density is converted in the main scanning direction is input to the error diffusion unit 2004 in synchronization with CLK2, and is output as binary image data binarized to 1 bit / pixel. A detailed description of the error diffusion process is omitted here.
[0091]
As described above, in the main scanning direction, duplication and thinning processing after luminance density conversion by clock control is performed, and in the sub-scanning direction, two types of triggers given to the binarization processing unit 200 are used to feed the document of the motor 160. By properly using the above, line overlap processing after edge enhancement can be performed, and the scaled image data can be binarized by the error diffusion unit and output.
[0092]
In the above-described embodiment, an example in which the edge enhancement filter coefficient shown in FIG. 9 (b) is used has been described, but if this edge enhancement filter coefficient is shown in FIGS. 12 (a) and (b), an edge It is possible to suppress the side effects of emphasis. In this case, the target pixel is F shown in FIG. 9A, and the pixel two lines before is referred to. For this reason, the data reference for the edge enhancement calculation is not performed on the data of the same line, and the side effect due to the edge enhancement after the overlap processing can be suppressed.
[0093]
As described above, according to this embodiment, when the read image data is converted into a binary image by the error diffusion method that expresses the density with the number of black dots per unit area and output, the binary image is converted into a binary image. By performing pixel density conversion before value conversion, it is possible to reduce the area of one dot of density expression, and it has the effect of making the texture unique to halftone processing inconspicuous in areas where the density gradient is relatively gentle. . Furthermore, by reducing the notch at the edge portion of the character or line drawing portion, it is possible to improve the image quality, such as improving the cutting of the edge portion of the character or line drawing portion.
[0094]
In addition, by implementing a pixel density conversion function with a simple clock control unit and two different triggers for line buffer control for edge enhancement, a line buffer for expanding in the sub-scanning direction is newly established. A binarization processing unit can be configured. Furthermore, a binarization processing unit that obtains a high-density binary image can be configured only by adding a simple control circuit and an error buffer for a necessary enlarged image.
[0095]
Of course, the enlargement process after edge enhancement prevents duplication of the image edge part (caused by superposition of high frequency components) as shown in Fig. 3, which is caused by duplicating the same data before edge enhancement. It is possible to obtain a good enlarged image.
[0096]
The line buffer necessary for edge enhancement may be a size that can store an image with a resolution of 203.2 dpi, and an increase in memory accompanying the addition of functions can be minimized.
[0097]
Furthermore, the image transfer clock of each calculation block in the enlargement process in the main scanning direction of the binarization processing unit uses a continuous clock in the edge enhancement and luminance density conversion processes, and the error diffusion process corrects the clock used for the edge enhancement. Arbitrary enlargement processing can be realized only by controlling the image transfer clock by thinning out several times the clock. In addition, since the image transfer clock for edge enhancement processing is a continuous clock, the connectivity with the CCD and the analog correction processing circuit is improved, and the line buffer of the edge enhancement portion can be reduced by one line with respect to the number of reference lines. .
[0098]
Further, the recording density can be kept constant by changing the luminance density conversion table in accordance with the pixel density to be converted in consideration of the density characteristics of the recording apparatus having different binary image recording characteristics depending on the resolution.
[0099]
Also, when transmitting in memory transmission mode, scanning in super fine mode (203.2dpi x 391.16dpi) is prohibited when scanning a document in photo mode, so do not perform resolution conversion after binarization. Can do.
[0100]
[Other embodiments]
In the above-described embodiment, in addition to the copy mode, when it is possible to communicate with the other party in the ultra-fine mode in the direct transmission mode, resolution conversion is performed before binarization. If the binarization setting is performed without converting the resolution, the reading of the original can be started immediately, which is efficient. In this case, selection of the ultra fine mode is prohibited. In addition, the super fine mode can be selected by registering the receiving capability of the other party corresponding to the one-touch key for each destination so that it can be selected when the recipient can receive in the super fine mode. In the transmission to other destinations, when the photo mode is further selected, the selection of the super fine mode is prohibited.
[0101]
In this way, the time from when the operator sets the document to when the operator removes the document can be shortened, which is efficient. At this time, since selection of the reading resolution is controlled, it is possible to prevent image quality deterioration due to resolution conversion.
[0102]
Further, in each of the above-described embodiments, binarization is performed by a process corresponding to each mode, so that copying of a document can be performed without deterioration in image quality in the copy mode, and resolution conversion can be performed in the transmission mode. It is possible to prevent degradation of image quality and reduction in processing efficiency due to.
[0103]
Further, by using a facsimile reading device (203.2 dpi), it is possible to construct the apparatus at a low cost, and transmit it while reading the resolution in the main scanning direction at the time of transmission (without performing resolution conversion). In addition, with the above processing, a high-quality copy image can be obtained even when a reading device having a resolution lower than the recording resolution is used, and a facsimile machine that maximizes the high resolution of the printer can be configured at low cost. In addition, by minimizing additional functions for processing in this case, no significant circuit correction is required compared to the conventional case.
[0104]
In the above-described embodiments, the error diffusion method has been described as an example of halftone processing, but it is needless to say that halftone processing with good density preservation, such as a minimum average error method, can be applied. Yes.
[0105]
Note that the present invention may be applied to a system composed of a plurality of devices or an apparatus composed of a single device.
[0106]
It goes without saying that the present invention can also be applied to a case where the present invention is achieved by supplying a system or apparatus with a program recorded in a storage medium. As a recording medium for supplying the program, for example, a floppy disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.
[0107]
【The invention's effect】
As described above, according to the present invention, it is possible to record a high quality image utilizing the resolution of the recording apparatus. In addition, it is possible to prevent deterioration in image quality due to resolution conversion.
[Brief description of the drawings]
FIG. 1 is a diagram showing an example of a binarized image by error diffusion when the recording resolution is twice different between main and sub;
FIG. 2 is an enlarged view of a part of a line drawing having an edge part.
FIG. 3 is a diagram showing an example of an image in which a pseudo contour is generated;
FIG. 4 is a diagram for explaining a process in which a pseudo contour is generated;
FIG. 5 is a diagram illustrating an example of an edge enhancement filter;
FIG. 6 is a block diagram illustrating a configuration example of a facsimile machine including an image processing apparatus according to an embodiment of the present invention;
FIG. 7 is a block diagram showing a detailed configuration example of the binarization processing unit in FIG.
FIG. 8 is a block diagram showing a detailed configuration example of the edge emphasizing unit in FIG.
FIG. 9 is a diagram showing edge enhancement filter reference, pixel position of interest, and filter coefficient executed by the edge enhancement calculation unit of FIG. 8;
FIG. 10 is a diagram for explaining a scaling process in the sub-scanning direction in the present embodiment;
FIG. 11 is a diagram illustrating a scaling process in the main scanning direction in the present embodiment;
FIG. 12 is a diagram showing another example of edge enhancement filter coefficients set in the edge enhancement calculation unit of FIG.
FIG. 13 is a flowchart showing a flow of processing related to reading a document such as copying and transmission;
FIG. 14 is a flowchart showing a flow of processing relating to reading of a document such as copying and transmission.
[Explanation of symbols]
100 CPU
110 ROM
120 RAM
130 Operation unit
140 MODEM
150 NCU
160 motor
170 Motor controller
180 CCD
190 Analog correction processor
200 Binarization processing section
210 RAM
220 LBP
230 Encoder
240 decoder
250 Bus Arbiter
260 System bus
270 image bus

Claims (6)

Transmitting means for transmitting image data of a predetermined resolution, reading means for reading a document image, and outputting image data representing the document image as multi-valued data per pixel with the predetermined resolution, and the reading means An image processing method of an image processing apparatus having a recording unit for recording an image based on binarized image data output at a recording resolution different from the predetermined resolution,
When a copy mode in which an image based on image data output by the reading unit is recorded on the recording unit is selected, the image data output by the reading unit is converted into the recording resolution and then binarized,
When the transmission mode for transmitting the image data output from the reading unit to the transmitting unit is selected, the resolution conversion is skipped and the image data output from the reading unit is binarized. Image processing method. A selection means for selecting a mode;
Transmitting means for transmitting image data of a predetermined resolution;
Reading means for reading a document image and outputting image data representing the document image as multivalued data per pixel with the predetermined resolution;
Conversion means for converting the resolution of the image data output by the reading means ;
Binarization means for converting image data output by the reading means into binary data ;
Recording means for recording an image based on the image data binarized by the binarization means at a second resolution different from the first resolution ;
By the selection unit, when the copy mode to record an image based on image data the reading unit outputs at the recording means is selected, the image data to which the reading means outputs a predetermined resolution by the converting means Is converted to a different recording resolution , then binarized by the binarizing means, and the selecting means selects the transmission mode for sending the image data output by the reading means to the transmitting means. skip the resolution conversion by the image processing apparatus according to claim image data to which the reading means outputs to a control means for binarizing by said binarizing means. 3. The image processing apparatus according to claim 2 , wherein the transmission unit transmits the image data by facsimile . The control means, the transmission mode is selected and the case the resolution conversion is required in accordance with the destination capability of the image data, according to claim 2 or claim, characterized in that performing the resolution conversion 3. The image processing apparatus described in 3. Further, the image processing apparatus according to any one of the preceding claims 2, characterized in that it comprises an edge enhancement means for performing edge enhancement processing to the image data before the resolution conversion. It said binarizing means is an image processing apparatus according to any one of the preceding claims 2, characterized in that performing the binarization by the error diffusion method.

Images