JP7710928B2 - Information processing device, information processing system, information processing method, and program - Google Patents

Description

The present invention relates to object identification.

Japan's fish catch is decreasing year by year. One of the reasons is the decline in fishery resources. To stop the decline in fish catch, the government is trying to manage the catch and strengthen fishery resource management. To grasp the catch, the government requires fishermen to report detailed catch records, but the reality is that fishermen do not have the time to devote to investigating catches. Therefore, an image analysis system for monitoring catches is required. To realize an image analysis system for monitoring catches, a model for classifying fish species is necessary. Patent Document 1 discloses a method for identifying fish species using deep learning. Patent Document 2 discloses a method for recognition by applying an appropriate recognition method according to the shooting conditions.

JP 2019-135986 A JP 2009-239871 A

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. Renra. NIPS2015 Yann LeCun, Leon Bottou, Joshua Bengio, Partick Haffner. “Gradient-Based Learning Applied to Document Recognition”. In Proceedings of the IEEE, 86, pp. 2278-2324, 1998.

However, because the brightness and color of fish bodies vary from species to species, it can be difficult to identify fish species from images taken under fixed shooting conditions. For example, blue fish such as horse mackerel and pacific saury tend to reflect light, and if the light source is from the front, the fish body will shine and cannot be correctly identified. In addition, if the color of the stand on which the fish is placed is similar to that of the fish body, it is difficult to distinguish the fish body, making it difficult to identify the fish species.

The present invention was developed in consideration of these problems, and aims to identify the type of object with high accuracy.

The information processing device of the present invention has a first identification means for identifying shooting conditions suitable for identifying the type of object by inputting a first image of the object captured under first shooting conditions into a first model, and a second identification means for identifying the type by inputting a second image of the object captured under second shooting conditions identified by the first identification means into a second model, wherein the second models are prepared for each shooting condition, and the second identification means uses the second model associated with the second shooting condition from among a plurality of the second models .

The present invention makes it possible to identify object types with high accuracy.

FIG. 1 is a diagram illustrating an example of an overall configuration of an information processing system. FIG. 2 is a diagram illustrating an example of a hardware configuration of an information processing system. FIG. 2 is a diagram showing a flow of data used in the information processing system. 13 is a flowchart showing a discrimination process. FIG. 4 is a diagram showing an example of shooting conditions. 13 is a flowchart showing a learning process. FIG. 11 is a diagram showing data generated by a learning process. FIG. 1 illustrates an example of a learning system.

The following describes in detail an embodiment of the present invention with reference to the attached drawings. Note that the configurations shown in the following embodiments are merely examples, and the present invention is not limited to the configurations shown in the drawings.

1 shows an example of the overall configuration of an information processing system according to this embodiment. The information processing system includes a belt conveyor 104 that transports objects 105 and 106, imaging devices 101 and 102 that image the objects 105 and 106, an information processing device 100 that performs identification, and an automatic sorting machine 107 that sorts the objects 105 and 106. The imaging devices 101 and 102 image the objects 105 and 106 on the belt surface of the belt conveyor 104 and transmit the images to the information processing device 100. The imaging device 102 is disposed downstream of the imaging device 101 in the conveying direction of the belt conveyor 104. In addition, a lighting device 103 is provided as a lighting facility for the imaging device 102. The lighting device 103 includes a light source that irradiates the object on the belt surface from the front at the shooting position of the imaging device 102, and a backlight incorporated in the belt conveyor 104 to irradiate the object on the belt surface from the back. The belt surface is translucent. The automatic sorter 107 is disposed downstream of the imaging device 102 in the conveying direction of the belt conveyer 104, and sorts the objects 105 and 106 traveling on the belt conveyer 104 into one of a number of sorting boxes 108-110 according to the identification result. The information processing device 100, the imaging devices 101 and 102, and the lighting device 103 are connected to each other so as to be able to communicate with each other via a network 211 (FIG. 2). The information processing device 100 outputs the identification result to the automatic sorter 107 via an output I/F 206 (FIG. 2). In the following description, the objects are described as fish, but fish is merely one example of an object as the object.

The information processing device 100 receives an image captured under a predetermined shooting condition at the shooting position of the imaging device 101 from the imaging device 101, and detects the area of the target fish from the received image. Then, the image of the fish area is input to a first model that has been learned in advance, thereby identifying the shooting conditions suitable for identifying the type of the target fish. In addition, a second model associated with the identified shooting conditions is specified. The information processing device 100 controls the camera settings of the imaging device 102 and the lighting settings of the lighting device 103 to be changed according to the identified shooting conditions. After that, the information processing device 100 receives from the imaging device 102 an image captured at the timing when the target fish flows on the belt conveyor 104 and reaches the shooting position of the imaging device 102. Note that the information processing device 100 predicts the timing when the same fish as the fish detected from the image of the imaging device 101 will arrive at the shooting position of the imaging device 102 from the conveying speed of the belt conveyor 104 and the distance between the imaging devices 101 and 102. The information processing device 100 may receive video from the imaging device 102 and obtain frame images at predicted timing from the received video. The information processing device 100 detects a fish area from the image received from the imaging device 102 and estimates the species of the target fish by inputting the image of the fish area into a second model that has been trained in advance. The information processing device 100 controls an automatic sorting machine 107 to sort the target fish into sorting boxes 108-110 according to the estimated fish species.

Here, the characteristics of the fish as the object in this embodiment, such as the degree of surface brilliance, surface color, surface smoothness, shape, size, etc., differ for each fish species. Therefore, it is difficult to accurately identify all fish species from images captured under fixed shooting conditions. For example, blue fish such as horse mackerel and pacific saury are prone to reflect light, so if they are photographed from the front with a light source from the front, the fish body will shine. In such a case, the fish species can be easily identified by adjusting the installation angle of the lighting device and the imaging device so that the reflection of the illumination light does not directly enter the imaging device. Also, if the color of the fish body surface is similar to the color of the belt surface of the belt conveyor, it is difficult to distinguish the fish body from the color of the belt conveyor. In such a case, the fish species can be easily identified by using the backlight of the belt conveyor and changing the color of the backlight according to the color of the fish body surface. Also, if the surface of the fish body is not smooth like a rockfish, it is easier to identify the fish species by applying light from an angle to cast a shadow, or by applying light from the front to illuminate the entire fish body in the case of a flat fish. As described above, it is believed that the ability to identify fish species can be improved by using images captured under shooting conditions suitable for identifying fish species. Therefore, in this embodiment, the first model is used to identify shooting conditions suitable for identifying the target fish species.

Next, an overview of the shooting conditions of the object to be identified in the first model will be described with reference to FIG. 5. In this embodiment, as shown in FIG. 5, various shooting parameter settings are determined for each of a plurality of shooting conditions. Note that FIG. 5 shows shooting conditions A to D, but shooting conditions A to D are examples, and there may be more shooting conditions. The shooting parameter settings include the camera settings of the imaging device 102 and the lighting settings of the lighting device 103. The camera settings include, for example, settings related to the shooting position, such as whether to shoot from the front or from a 30-degree angle, and settings related to camera parameters such as shutter speed. The lighting settings include, for example, settings related to the lighting position, such as whether to irradiate from the front or from a 45-degree angle, settings related to the amount of light, and settings related to the backlight (ON/OFF, amount of light, color). Note that the shooting parameter settings are not limited to these. For example, in addition to these, there may be settings such as the type of light source and color temperature. For example, there may be settings for the aperture and focus. There may also be settings for the conveying speed of the belt conveyor 104. There may also be combinations of settings that are not used between the shooting parameters. For example, a combination of low light intensity and a fast shutter speed is not used because it results in insufficient exposure. The camera settings also include a zoom magnification setting, such as zooming in on the target fish when the range of the fish area in the image is less than a specified size. Which of the multiple shooting conditions A to D shown in Figure 5 is suitable for identifying fish species is acquired by learning. The learning method will be described later.

In this embodiment, as shown below, a two-stage classification is performed. In the first stage classification, a photographing condition suitable for identifying a fish species is identified, and in the second stage classification, a fish species is identified using an image captured under the photographing condition identified in the first stage. In the first stage classification, a first model, which is a classification model for performing the first stage classification, is used to identify a photographing condition suitable for identifying a fish species appearing in an image (first image) captured under a fixed photographing condition (first photographing condition). The second model is a classification model for performing the second stage classification, which is learned for each photographing condition and is stored in association with the photographing condition. In this embodiment, since there are a plurality of photographing conditions, there are also a plurality of second models. In the second stage classification, a second model associated with the photographing condition identified in the first stage classification is used to identify a fish species appearing in an image (second image) captured under the photographing condition (second photographing condition) determined in the first stage classification.
As described above, in the first stage of classification, classification is performed under shooting conditions suitable for classifying fish species, and in the second stage of classification, classification of fish species is performed using images captured under shooting conditions suitable for each fish species. This improves the accuracy of fish species classification. Furthermore, in the first stage of classification, classification of fish species groups is performed under shooting conditions suitable for classifying fish species, and in the second stage of classification, classification is performed within groups of fish species having the same suitable shooting conditions. This makes it possible to reduce the number of classes to be classified, and shorten the time required for fish species classification. The learning methods of the first model and the second model will be described later.

Next, the hardware configuration of the information processing system according to this embodiment will be described. FIG. 2 shows an example of the hardware configuration of the information processing system according to this embodiment. The information processing device 100 is composed of a CPU 201, a ROM 202, a RAM 203, an external storage device 204, an input I/F 205, an output I/F 206, and a communication I/F 207. Each of these components is connected to be able to communicate with each other via a system bus 208.

The CPU (Central Processing Unit) 201 controls the entire information processing device 100. The ROM (Read Only Memory) 202 stores programs and parameters that do not require modification. The RAM (Random Access Memory) 203 temporarily stores data supplied from an external device or the like. The external storage device 204 is a storage device such as a hard disk or a memory card that is fixedly installed in the information processing device 100. The external storage device 204 may include an optical disk such as a flexible disk (FD) or a compact disk (CD) that is detachable from the information processing device 100, a magnetic or optical card, an IC card, a memory card, or the like. The functions and processes of the information processing device 100 described below are realized by the CPU 201 reading out a program stored in the ROM 202 or the like and executing this program.

The input I/F 205 is an interface with an input device 209 such as a pointing device or a keyboard that receives a user's operation and inputs data. The output device I/F 206 is an interface with a monitor 210 for displaying data held by the information processing device 100 and data supplied thereto, and with an automatic sorting machine 107 that automatically sorts fish based on the identification results. The communication I/F 207 is connected to a network 211. The CPU 201 transmits and receives data to and from the imaging devices 101 and 102 and the lighting device 103 via the network 211.
The imaging devices 101 and 102 and the lighting device 103 have a communication function and are connected to a network 211. The belt conveyor 104 may also be connected to the network 211. In this case, the information processing device 100 may control the start/end of conveyance of the belt conveyor 104 and the conveyance speed.

Figure 3 shows an example of the flow of data used in the information processing system according to this embodiment. In Figure 3, the information processing device 100 functions as a detection unit 301 that detects fish regions from an image, and a recognition unit 302 that performs recognition using an image of the fish region. A first model and a second model used by the recognition unit 302 are stored in the external storage device 204 of the information processing device 100. The second model is prepared for each of a number of shooting conditions. The data flow will be described below.

First, the information processing apparatus 100 acquires the captured image 303 from the imaging apparatuses 101 and 102 via the network 211 and the communication I/F 207 .
The detection unit 301 analyzes the captured image 303 and detects the area of the fish in the image. The detection unit 301 uses, for example, an object detection method using CNN (Convolutional Neural Networks), Faster R-CNN described in Non-Patent Document 1. Faster R-CNN can detect multiple classes of objects such as people and cars. When using Faster R-CNN, learning is performed using a large number of fish images collected in advance so that fish can be detected.

The identification unit 302 identifies the fish area extracted by the detection unit 301. If the target captured image 303 is acquired from the imaging device 101, the identification unit 302 reads out the first model 304 from the external storage device 204 and identifies the shooting conditions using the first model 304. Furthermore, the identification unit 302 specifies the second model 307 associated with the identified shooting conditions 306 from among the multiple second models stored in the external storage device 204. After that, the information processing device 100 transmits the identified shooting conditions 306 to the imaging device 102 and the lighting device 103 via the communication I/F 207. The imaging device 102 and the lighting device 103 adjust various settings based on the received shooting conditions 306 and capture the image.

On the other hand, if the target captured image 303 has been acquired from the imaging device 102, the identification unit 302 reads the second model 307 specified above from the external storage device 204, identifies the fish species using the second model 307, and acquires the identification result 305. The information processing device 100 then notifies the automatic sorting machine 107 of the identification result 305 via the output I/F 206. The automatic sorting machine 107 performs sorting according to the identification result 305. The identification unit 302 may be configured to acquire the shooting conditions used when capturing the image from the imaging device 102 together with the captured image 303, and use the second model associated with the acquired shooting conditions.

Next, the identification process executed by the information processing device 100 according to this embodiment will be described with reference to FIG. 4. FIG. 4 is a flowchart showing the flow of the identification process. The flowchart in FIG. 4 is realized by the CPU 201 expanding a program stored in the ROM 202 etc. into the RAM 203 and executing it. The flowchart in FIG. 4 starts when the input device 209 is operated to operate the conveyor belt 104 and a fish is placed on the conveyor belt 104. Each process (step) in the flowchart will be described below with an S (step) added to the beginning of each reference number.

First, in step S<b>401 , the CPU 201 acquires a first image from the imaging device 101 via the communication I/F 207 .
Next, in S402, the CPU 201 detects a fish region from the first image acquired in S401. Note that the flow when one fish region is detected will be described here. When multiple fish regions are detected, the subsequent processes from S403 to S409 can be repeated, and the description thereof will be omitted here.
Next, in S403, the CPU 201 inputs the image of the fish region detected in S402 to the first model 304, and specifies the photographing conditions 306 and the second model 307 suitable for identifying the fish species. The first model 304 is a model for identifying which photographing conditions are suitable for identifying the type of fish detected in S402. The CPU 201 uses the first model 304 to identify a class corresponding to the photographing conditions suitable for identifying the fish species of the input image. Specifically, classifying the photographing conditions A to D, etc. shown in FIG. 5 is performed. The second model 307 is a model for identifying the fish species identified under each photographing condition, and one second model is linked to each photographing condition.

Next, in S404, the CPU 201 compares the shooting conditions 306 specified in S403 with the shooting conditions used when the first image was captured, and determines whether they are the same. If the CPU 201 determines that the shooting conditions are the same, the first image can be used as the second image as is, so S405 to S407 are skipped and the process proceeds to S408. On the other hand, if the CPU 201 determines that the shooting conditions are different, the process proceeds to S405.
In S405, the CPU 201 transmits the shooting conditions 306 specified in S403 to the image capturing apparatus 102 and the lighting apparatus 103 via the communication I/F 207. The image capturing apparatus 102 and the lighting apparatus 103 change the settings in accordance with the received shooting conditions 306.
Next, in S406, the CPU 201 acquires a second image captured under the shooting condition 306 from the imaging device 102 via the communication I/F 207 at the timing when the same fish as the fish detected in S402 arrives at the shooting position of the imaging device 102.
Next, in S407, the CPU 201 detects a fish region from the second image acquired in S406.
Next, in S408, the CPU 201 inputs the image of the fish region detected in S407 to the second model 307 specified in S403, and identifies the fish species. Note that if the shooting conditions 306 specified in S403 are the same as the shooting conditions used when the first image was captured, the image of the fish region detected in S402 is input.
Next, in S409, the CPU 201 notifies the automatic sorting machine 107 of the identified fish species via the output I/F 206. The automatic sorting machine 107 sorts the fish detected in S402 according to the notified fish species. Then, the series of identification processes shown in the flowchart ends.

According to the classification process of this embodiment as described above, images captured under shooting conditions suitable for each fish species are used, so the accuracy of fish species classification can be improved. Furthermore, because the suitable shooting conditions are used for classification within a group of the same fish species, the number of classes to be classified can be reduced. Therefore, the time required for fish species classification can be shortened.

Next, the learning process in the learning phase of the information processing device 100 according to this embodiment will be described with reference to Fig. 6. Fig. 6 is a flowchart showing the flow of the learning process. The flowchart in Fig. 6 is realized by the CPU 201 expanding a program stored in the ROM 202 or the like into the RAM 203 and executing it. Fig. 7 shows data generated by the learning process in Fig. 6.
First, as learning data for learning, a plurality of captured images are prepared by capturing images of a plurality of types of fish under different shooting conditions using the imaging devices 101 and 102 according to the present embodiment. Each captured image is labeled with the shooting conditions and fish species used when capturing the image as correct answer data.
First, in S601, the CPU 201 learns the second model using a DNN (deep neural network). The learning method will be described in detail later. This learning is performed for each shooting condition in FIG. 5. The CPU 201 then performs learning so that each fish species becomes a separate class, and obtains the identification accuracy for each fish species. When learning for all shooting conditions is completed, the CPU 201 tallies the identification accuracy for each shooting condition for each fish species. The CPU 201 then determines the shooting condition suitable for identification for each fish species based on the tallied results.

Figure 7(a) shows an example of the results of tallying up the identification accuracy for each shooting condition by fish species. Figure 7(a) shows the identification accuracy for shooting conditions A to D for horse mackerel, mackerel, squid, thresher, red scorpionfish, and rockfish. Note that shooting conditions A to D in Figures 7(a) to 7(c) are the same as shooting conditions A to D shown in Figure 5. Here, the CPU 201 determines the shooting condition suitable for identification as the shooting condition with the highest identification accuracy. Note that the highest identification accuracy is not limited to the case where the identification accuracy is higher than the others. The parts marked with a star (★) in Figure 7(a) indicate the shooting condition with the highest identification accuracy. For example, in the case of horse mackerel, the identification accuracy for shooting conditions A to D is 32%, 47%, 66%, and 95%, so shooting condition D is determined. Also, for example, in the case of mackerel, the identification accuracy for shooting conditions A to D is 18%, 33%, 52%, and 94%, so shooting condition D is determined. For example, in the case of squid, the identification accuracy for shooting conditions A to D is 99%, 75%, 83%, and 67%, so shooting condition A is selected. Similarly, for thresher, redfish, and rockfish, the shooting condition suitable for identification is selected based on the identification accuracy for each shooting condition. FIG. 7(b) shows an example of a list of fish species suitable for identification under each shooting condition A to D. As shown in FIG. 7(b), fish species can be grouped according to the shooting conditions suitable for fish species identification. As a result, in S403 of FIG. 4, fish can be classified into the same number of groups as the number of shooting conditions. In the example of FIG. 7(b), horse mackerel, mackerel, and pacific saury are classified into the same group as fish species suitable for identification under shooting condition D.

Next, in S602, the CPU 201 learns a first model using a DNN. This learning is performed using learning data of one of the multiple shooting conditions (for example, shooting condition A). This learning is performed so that fish species in groups with the same shooting condition are classified into the same class, and fish species in groups with different shooting conditions are classified into different classes. The model learned in this step is used as the first model in the identification process of FIG. 4. In addition, the shooting condition of the first image acquired in S401 is the shooting condition used in the learning in this step (here, shooting condition A). In this manner, a first model that identifies the shooting condition suitable for identifying fish species is generated from the image captured under shooting condition A.
Next, in S603, the CPU 201 re-learns the second model learned in S601. The learning in S601 was performed to identify all fish species. In this step, the learning is performed for each shooting condition to identify fish species belonging to the same shooting condition. For example, using the learning data for shooting condition A, learning is performed so that fish determined to belong to shooting condition A in S601 are classified into different classes for each fish species. When shooting condition A is identified in S403 in the identification process of FIG. 4, the model learned in this step is used as the second model linked to shooting condition A. Similarly, the second model is re-learned for shooting conditions B to D. The fish species to be identified by the second model differs depending on the associated shooting condition. In this manner, a second model for each shooting condition is generated.

In this embodiment, the shooting condition for the first image acquired in S401 is shooting condition A. Here, the CPU 201 may learn under which shooting condition the first image should be captured. For example, the learning data for shooting conditions A to D is used to perform learning in S602 for each shooting condition. Then, the identification accuracy is obtained for each shooting condition. When learning for all shooting conditions is completed, the CPU 201 determines the shooting condition suitable for identification based on the identification accuracy under each shooting condition. FIG. 7(c) shows an example of the identification accuracy under each shooting condition. Here, the CPU 201 determines the shooting condition suitable for identification as the shooting condition with the highest identification accuracy. Note that the highest identification accuracy is not limited to the case where the identification accuracy is higher than the others. The part marked with a star (★) in FIG. 7(c) indicates the shooting condition with the highest identification accuracy, and in the example of FIG. 7(c), it is shooting condition C. This shows that it is optimal to learn the first model using an image captured under shooting condition C. In this case, in the identification process of FIG. 4, in S401, a first image captured under shooting condition C is obtained, and in S403, a first model trained using the image captured under shooting condition C is used.

Figure 8 shows an example of the configuration of a learning system used for learning. Below, we explain the learning method when the learning system is configured with a DNN (deep neural network). In this embodiment, for example, LeNet described in Non-Patent Document 2 is used as the neural network.

In FIG. 8, an image 81 is input to a DNN 82, which outputs a vector 83 whose number of dimensions is the number of fish species to be identified in the cases of S601 and S603, and the number of shooting conditions in the case of S602. Each element of vector 83 is a probability between 0 and 1, and the total value is 1. A loss function 84 is then calculated using the correct vector of image 81. A correct vector is a vector in which only the correct elements are 1, and the other elements are 0. The cross-entropy error H(p, q) in the following formula (1) is used as the loss function 84.

In the above formula (1), p is the correct vector of the image 81, and q is the vector 83 of the output of the DNN 82. x is the number of dimensions, i.e., the number of fish species or the number of shooting conditions. Learning is performed by updating the weights of the DNN 82 using the calculated value of the loss function 84 through the error propagation method. The first model and the second model are learned using the above learning system.

According to the information processing system of this embodiment as described above, the shooting conditions suitable for identifying fish species are determined in the first stage of classification, and the fish species are identified in the second stage of classification using images captured under those shooting conditions. In the second stage of classification, classification is performed using images captured under shooting conditions suitable for each fish species, so the accuracy of fish species classification can be improved compared to when images captured under fixed shooting conditions are used. Furthermore, in the second stage of classification, the suitable shooting conditions are used to identify within a group of the same fish species, so the number of classes to be classified can be reduced. Therefore, the time required for fish species classification can be shortened.

In this embodiment, an example of application to identifying fish species has been described, but the invention can be applied to identifying various other objects, such as identifying the type of parts on a conveyor belt in an assembly line in a factory. Parts have different characteristics, such as the presence or absence of gloss, unevenness, shape, and size. In the first stage of identification, appropriate shooting conditions are identified according to the above-mentioned characteristics of the parts, and in the second stage of identification, the type of part is identified using images captured under the identified shooting conditions. This makes it possible to identify the type of part with high accuracy.

Other Embodiments
The present invention can also be realized by a process in which a program for implementing one or more of the functions of the above-described embodiments is supplied to a system or device via a network or a storage medium, and one or more processors in a computer of the system or device read and execute the program. The present invention can also be realized by a circuit (e.g., ASIC) that implements one or more of the functions.

The present invention has been described above in conjunction with embodiments, but the above embodiments are merely illustrative examples of how the present invention may be implemented, and the technical scope of the present invention should not be interpreted in a limiting manner based on these embodiments. In other words, the present invention can be implemented in various forms without departing from its technical concept or main features.

Below, a modification of the above embodiment will be described.
As a first modified example, instead of a method in which the objects 105 and 106 are transported by the belt conveyor 104 and images for identification are taken of the first stage and the second stage at different points, images for identification may be taken at the same point under different shooting conditions. In this case, the belt conveyor 104 does not need to be used.
As a second variant, images may be taken at multiple points on the conveyor belt 104 for a number of shooting conditions, and an image taken under shooting conditions suitable for classification may be selected from those images to perform the second stage of classification.

As a third variant, the objects 105 and 106 may be arranged in an aligned state on the conveyor belt 104 so that the number of objects appearing in the first image and the second image is limited.
As a fourth modified example, the learning process shown in Fig. 6 may be executed by another device that performs learning. In that case, the first model and the second model used in the classification process may be obtained from the other device.

100: Information processing device, 101, 102: Imaging device, 103: Lighting device, 104: Belt conveyor, 105, 106: Object, 107: Automatic sorting machine

Claims (19)

a first identification means for inputting a first image obtained by capturing an object under a first capturing condition into a first model, thereby identifying a capturing condition suitable for identifying a type of the object;
a second identification means for identifying the type by inputting a second image obtained by capturing the object under the second photographing condition identified by the first identification means into a second model;
having
An information processing device characterized in that the second model is prepared for each shooting condition, and the second identification means uses the second model associated with the second shooting condition among the multiple second models . The information processing apparatus according to claim 1 , wherein the type of the object to be identified differs for each of the second models. 3. The information processing device according to claim 1, wherein the photographing conditions suitable for identifying the type are determined for each type based on a result of tallying up the identification accuracy of the second model for each of a plurality of photographing conditions. The information processing device according to any one of claims 1 to 3, further comprising a first learning means for learning the first model so that images of the object captured under the first shooting conditions are classified into classes according to the types having the same shooting conditions suitable for identifying the type. 5. The information processing device according to claim 1, further comprising a second learning means for learning the second model so that images of the object captured under shooting conditions suitable for identifying the type are classified according to the type . The information processing device according to claim 4, characterized in that the first learning means determines the shooting conditions to be used as the first shooting conditions based on a result of tallying up the identification accuracy of the first model for each of a plurality of shooting conditions. 7. The information processing device according to claim 1, wherein the second identification means inputs the first image into the second model when the second shooting conditions are the same as the first shooting conditions. 8. The information processing apparatus according to claim 1, further comprising a detection unit for detecting an area of the object from at least one of the first image and the second image. 9. The information processing apparatus according to claim 1 , wherein the photographing conditions include at least one of a photographing position and settings related to camera parameters. 10. The information processing apparatus according to claim 1 , wherein the photographing conditions include at least one of settings related to a lighting position, a lighting amount, and a lighting color. The imaging device further includes a control unit for controlling the imaging device that captures an image of the object.
The control means controls the setting of the imaging device to be changed to the second photographing condition,
11. The information processing apparatus according to claim 1, wherein the second identification means acquires the second image from the imaging device. 12. The information processing apparatus according to claim 11 , wherein the control means further controls an illumination device of the imaging device, and controls so as to change the setting of the illumination device to the second photographing condition. the object is transported on a conveyor belt;
the control means controls a first imaging device and a second imaging device provided downstream of the first imaging device in a conveying direction of the belt conveyor;
The first identification means acquires the first image from the first imaging device,
The information processing device according to claim 11 or 12, characterized in that the second identification means acquires the second image from the second imaging device at a timing when it is predicted that the object same as the object imaged by the first imaging device has reached the shooting position of the second imaging device based on the distance between the first imaging device and the second imaging device and the conveying speed of the conveyor belt . 14. The information processing apparatus according to claim 1, wherein the second identification means selects an image captured under the second shooting condition from among images captured under a plurality of shooting conditions. 15. The information processing apparatus according to claim 1, wherein the object is a fish. 16. The information processing apparatus according to claim 1, wherein at least one of the first model and the second model is configured as a neural network. An information processing system including an imaging device that images an object and an information processing device that identifies the object,
The information processing device includes:
a first acquisition means for acquiring a first image captured under a first photographing condition from the imaging device;
a first identification means for identifying a photographing condition suitable for identifying the type of the object by inputting the first image into a first model;
A control means for controlling the image capturing device to change the setting to a second photographing condition;
a second acquisition means for acquiring a second image captured under the second photographing condition from the imaging device;
a second identification means for identifying the type by inputting the second image into a second model;
having
An information processing system characterized in that the second model is prepared for each shooting condition, and the second identification means uses the second model associated with the second shooting condition from among the multiple second models . a first identification step of inputting a first image of an object captured under a first imaging condition into a first model, thereby identifying an imaging condition suitable for identifying a type of the object;
a second classification step of classifying the type by inputting a second image of the object captured under the second photographing condition classified in the first classification step into a second model;
Including,
The information processing method is characterized in that the second model is prepared for each shooting condition, and in the second identification process, the second model associated with the second shooting condition is used from among the multiple second models . A program for causing a computer to function as each of the means of the information processing apparatus according to any one of claims 1 to 16 .