JP2021082015A - Electronic apparatus and control method of the same - Google Patents

Abstract

To provide an electronic apparatus capable of improving a user's operability when performing an image determination on the basis of learning data.SOLUTION: A display control device performs control such that a partial area of a first image is selected, any display form among a plurality of display forms is set in association with information on the selected area. Then, a second image corresponding to the first image and displaying an area different from the area selected from the first image and the selected area in the display form set by setting means is displayed with the first image.SELECTED DRAWING: Figure 2

Description

The present invention relates to an electronic device and a control method thereof, and more particularly to an image display method when making a determination based on learning data.

There is a technique for determining each area in an image based on learning data. Patent Document 1 discloses a technique for determining a subject category by analyzing an image into a small area and detecting a feature amount for each area.

JP-A-2019-12426

In Patent Document 1, since the display method of the determination result cannot be changed, it may be difficult for the user to grasp the determination result.

In view of the above problems, an object of the present invention is to provide an electronic device that improves user operability when determining an image based on learning data.

In order to achieve the above object, the electronic device of the present invention associates a selection means capable of selecting a part of the area of the first image with the information of the selected area, and a plurality of display modes. A setting means for setting one of the display modes, an area different from the selected area in the first image, and the selected area are displayed in the display form set by the setting means. It is characterized by having a control means for controlling the second image corresponding to the first image to be displayed together with the first image.

According to the present invention, it is possible to improve the operability of the user when determining an image based on the learning data.

(A) External view of a personal computer as an example of a device to which the configuration of the present embodiment can be applied, (b) A block diagram showing a configuration example of a personal computer as an example of a device to which the configuration of the present embodiment can be applied. Flowchart showing image display processing in this embodiment An example of the main screen in this embodiment An example of an image of each wavelength filter in this embodiment An example of the setting screen in this embodiment An example of the display color setting screen in this embodiment An example of the wavelength of the selected region in this embodiment An example of displaying the determination result in this embodiment The figure for demonstrating the inspection system in this embodiment The figure explaining the inspection of meat in this embodiment The figure explaining the part of the meat to be inspected in this embodiment

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

<Hardware>
FIG. 1 shows an example of the configuration of the display control device 100 to which the present invention can be applied. The display control device 100 can be configured by using a personal computer (hereinafter, PC) or the like.

In FIG. 1, a CPU 101, a memory 102, a non-volatile memory 103, an image processing unit 104, a display 105, an operation unit 106, a recording medium I / F 107, an external I / F 109, and a communication I / F 110 are connected to the internal bus 150. ing. Each part connected to the internal bus 150 can exchange data with each other via the internal bus 150.

The memory 102 includes, for example, a RAM (such as a volatile memory using a semiconductor element). The CPU 101 controls each part of the display control device 100 by using the memory 102 as a work memory according to a program stored in the non-volatile memory 103, for example. The non-volatile memory 103 stores image data, audio data, other data, various programs for operating the CPU 101, and the like. The non-volatile memory 103 is composed of, for example, a hard disk (HD) or a ROM.

Based on the control of the CPU 101, the image processing unit 104 acquired the image data stored in the non-volatile memory 103 or the recording medium 108, the video signal acquired via the external I / F 109, and the communication I / F 110. Performs various image processing on image data and the like. The image processing performed by the image processing unit 104 includes A / D conversion processing, D / A conversion processing, image data coding processing, compression processing, decoding processing, enlargement / reduction processing (resizing), noise reduction processing, and color conversion. Processing etc. are included. The image processing unit 104 may be configured by a dedicated circuit block for performing specific image processing. Further, depending on the type of image processing, the CPU 101 may perform image processing according to a program without using the image processing unit 104.

The display 105 displays an image, a GUI screen that constitutes a GUI (Graphical User Interface), and the like based on the control of the CPU 101. The CPU 101 controls each part of the display control device 100 so as to generate a display control signal according to a program, generate a video signal to be displayed on the display 105, and output the video signal to the display 105. The display 105 displays an image based on the output image signal. The display control device 100 itself may include an interface for outputting a video signal to be displayed on the display 105, and the display 105 may be configured by an external monitor (television or the like).

The operation unit 106 is an input device for accepting user operations, including a character information input device such as a keyboard, a pointing device such as a mouse and a touch panel, a button, a dial, a joystick, a touch sensor, and a touch pad. The touch panel is an input device that is superposed on the display 105 and is configured in a plane so that coordinate information corresponding to the contacted position is output.

A recording medium 108 such as a memory card, a CD, or a DVD can be attached to the storage medium I / F 107, and data can be read from the attached recording medium 108 or data can be written to the recording medium 108 based on the control of the CPU 101. I do. The external I / F 109 is an interface for connecting to an external device by a wired cable or wirelessly to input / output video signals and audio signals. The communication I / F 110 is an interface for communicating with an external device, the Internet 111, or the like to send and receive various data such as files and commands.

Next, with reference to FIGS. 9 to 11, a meat inspection, which is an example of an inspection target used as learning data in the present embodiment, will be described. In the present embodiment, the meat mass is inspected by X-ray, and the wavelength of each region of the meat is acquired by the inspection system 200 shown in FIG. Based on the wavelength data acquired from the inspection system 200 and the image of the meat, it is made to learn which part of the meat is bone, lean meat, cartilage, fat, etc., and stored as learning data. In this embodiment, as an example, the input of learning data when inspecting a chunk of meat will be described. That is, when the user specifies, for example, the fat region while looking at the meat to be inspected and specifies the display color, the region having the same wavelength characteristics is set as the fat region based on the wavelength information of the specified region. Memorize the learning data that there is. As a result, the fat area can be immediately grasped with the color specified by the user without the user checking all the meat chunks. In particular, although cartilage is described below, the present embodiment is not limited to meat and cartilage, and any one that can determine the characteristics based on the wavelength may be used.

First, the configuration of the inspection system according to the present embodiment will be described with reference to FIG. FIG. 1 is a block diagram showing a configuration of an inspection system according to the present embodiment.

The inspection system 200 according to the present embodiment includes a PC 240 (personal computer), a measuring device 210 (X-ray device, hyperspectral camera), and a display unit 220. The PC 240 to the display unit 220 may be partially or wholly connected via the network 230. Note that the one shown in FIG. 9 is an example, and each component may be dispersed in more devices, or may be grouped in one device. The PC 240 described later and the PC 100 shown in FIG. 1 may be the same PC. Each component further allows the PC 240 to communicate with or control a device for controlling meat in a factory (not shown). The network 230 includes a LAN (Local Area Network) and the Internet.

The PC 240 includes a CPU 201, a classifier 202, a database 203 that stores learned data applied to the classifier 202, and an operation unit 204.

The CPU 201 is an arithmetic circuit that controls the entire PC 240 according to a program stored in a memory (not shown), controls the measuring device 210, and causes the classifier 202 to analyze the spectrum data acquired from the measuring device 210. The classifier 202 is composed of a GPU (Graphics Processing Unit) and an FPGA (field-programmable gate array). For example, in the present embodiment, the classifier 202 uses the discriminant parameters generated by machine learning stored in the database 203, and based on the spectral data obtained from the measuring device 200, the meat part to be inspected. Is discriminated. Database 203 stores identification parameters obtained by performing machine learning in advance using a cloud or a server. A plurality of types of the identification parameters are prepared according to the type of the inspection target, the inspection level, or the teacher data, and the identification parameters selected by the CPU 201 are used for the identification process of the classifier 202. The operation unit 204 is an input device for accepting user operations, including a character information input device such as a keyboard, a pointing device such as a mouse and a touch panel, a button, a dial, a joystick, a touch sensor, and a touch pad. The touch panel is an input device that is superposed on the display unit 220 and is configured in a plane so that coordinate information corresponding to the contacted position is output.

The measuring device 210 includes a controller 211, a measuring unit 212, and a lighting device 213. The controller 211 is a circuit that communicates with the CPU 201 and controls the entire measuring device 210 according to the instruction of the CPU 201.

The measurement unit 212 is controlled by the controller 211 to measure the signal level of each spectral component from the meat to be inspected (not shown) and generate spectral data.

The spectral data is not particularly limited as long as it is data in which an intensity value (referred to as “spectral intensity”) of the spectral component is stored for each of the plurality of spectral components. As the spectral data, for example, the response intensity (corresponding to the spectral intensity) of the response generated when the meat to be inspected is stimulated is stored with respect to the measurement parameter (corresponding to the spectral component). Data can be used. The "stimulus" here includes electromagnetic waves, sound, electromagnetic fields, temperature, and humidity.

When the spectrum data is spectroscopic spectrum data in the ultraviolet or visible or infrared region, or Raman spectroscopic spectrum data, the spectral component can be a wavelength or a wave number. When the spectrum data is mass spectrum data, the spectrum component can be a mass-to-charge ratio or a mass number.

The spectral data belongs to one of the groups corresponding to a plurality of constituents contained in the meat to be inspected. The spectral component and the spectral intensity differ depending on each of the constituent components of the meat to be inspected, which are included in the measurement region in which the spectral data is acquired. Therefore, by analyzing the spectrum data, the group to which the spectrum data belongs can be identified, and each spectrum data can be assigned to each component.

In the present embodiment, the spectral components are set to dozens to hundreds of wavelength components in the wavelength bands of visible light and near-infrared light, and the measuring unit 212 is for obtaining image data corresponding to each spectral component. It consists of an image sensor or a line sensor.

The lighting device 213 has a light source for irradiating the meat to be inspected. In this embodiment, a halogen lamp that covers near-infrared light and visible light is used, and by irradiating the meat, the intensity of each spectral component reflected by the meat can be increased.

The display unit 220 is a portion that displays the identification result of the PC 240. As the display unit 220, for example, a liquid crystal display, an organic EL display, or the like can be used. The display unit 220 can display image data or the like transmitted via the network 230.

The processing process of meat in the present embodiment will be described with reference to FIG. After passing through a meat wholesaler or a livestock farmer, the meat to be processed is transported to a factory where it is processed. In the meat processing process, there are mainly what is called block meat in which the same part is solidified, and what is called end meat which is a mass of meat pieces of various parts that did not enter the block meat mass. Is done. In the present embodiment, the process of processing meat for producing a hamburger will be described, but it goes without saying that the present embodiment is not limited to this and can be applied to sausages and other processed meat foods.

The block meat and the cut meat proceed to the meat processing process in a frozen state in order to maintain the quality of the meat. In particular, the end meat, which is a collection of meat pieces from various parts, is a large mass.

Hereinafter, a description will be given with reference to FIG. 10 (a). First, an inspection using X-rays is performed to detect bones contained in block meat and end meat and to confirm whether or not foreign substances are contained. The inspection is performed with the block meat and the end meat lined up on the belt conveyor, and if there are bones or foreign substances, they are removed in the subsequent process. In the X-ray inspection, the belt conveyor can be inspected at a speed of 15 meters / minute and 25 meters / minute.

Next, an inspection using a hyperspectral camera having a wide wavelength range that can be irradiated at one time, which is called a hyperspectral inspection in the present embodiment, is performed. In the hyperspectral test, the classifier 202 discriminates the spectrum data acquired by the measuring device 210 using the discriminating parameters obtained based on machine learning, and the cartilage / muscle / tendon is located at any position. It outputs the result such as whether it is in the meat or not in the meat. In addition, although it is assumed here that the output of several tens to several hundreds of wavelengths is used, the hyperspectrum is used, but even if the multispectrum is based on the assumption that the output of several wavelengths is used, this book is used. It is possible to apply the invention.

In the hyperspectral inspection, the meat is placed on the belt conveyor and inspected in the same manner as the X-ray inspection. Therefore, it is desirable to set the processing speed so that the belt conveyor can be operated at the same speed as the X-ray inspection. However, if it takes time to output the identification result by the classifier 202, more hyperspectral cameras may be installed than the X-ray device, or the number of classifiers may be increased. After the X-ray examination and the hyperspectral examination, the block meat and the end meat excluding the parts such as bones, foreign substances, cartilage, muscles, and tendons are minced, and the process proceeds to mold the hamburger steak.

Next, the state of the X-ray inspection and the hyperspectral inspection will be described with reference to FIG. 10 (b). In the present embodiment, the X-ray and hyperspectral inspection of the meat 201 is performed on the belt conveyor 1002. As shown in FIG. 10B, meat 1001 which is end meat or block meat flows on the belt conveyor 1002, and is first inspected by the X-ray inspection apparatus 1003, and then the hyper, which is the inspection system of the present embodiment. A spectral test is performed. The moving direction of the belt conveyor 1002 is the direction of the arrow shown in FIG. 10 (b). In this way, each inspection is performed by shining a light wave on the meat. In hyperspectral inspection, light reaches only near the surface to be inspected, compared to X-ray inspection. Therefore, the hyperspectral inspection result is also limited to a shallow range from the irradiation surface in the hyperspectral inspection as compared with the X-ray inspection. Therefore, in the hyperspectral test, it is desirable to slice the meat to be tested into a range of 2 cm or 5 cm. Alternatively, for example, the meat 1001 may be sliced to a thickness of 10 cm or 8 cm, and both sides of the meat 1001 may be inspected. Furthermore, as will be described later, as a result of the inspection, meat that may proceed to the normal mincing process as it is (referred to as A pattern), meat that requires further human judgment (referred to as B pattern), and meat that requires further human judgment (referred to as B pattern) are normal as they are. The classification of meat may be subdivided, as in meat (C pattern), which should not proceed to the minced process. Since the meat of which part is known at the stage of dividing into block meat, the block meat may not be subject to X-ray or hyperspectral inspection, but only the end meat may be subject to X-ray or hyperspectral inspection. ..

The lighting device 213 is provided with a pair of light sources for irradiating light having directivity so as to sandwich the measuring unit 212, and the irradiation directions of the respective light sources are inclined in opposite directions with respect to the vertical direction. By irradiating light at different angles, it is possible to irradiate the part that is shaded by the irradiation of one light source with the other light source, and by collecting the light emitted from both light sources at the same place, More light can be applied to the inspection target.

That is, in the inspection system 200, the wavelength of each region of the meat can be acquired as learning data by using X-rays and hyperspectrums. Then, in the image display process described later, the user connects the area and the color and specifies the area, so that the area determined to have the same wavelength characteristics as the specified area is displayed in the color specified by the user. Can be done. For example, if the user specifies the cartilage part as red while checking the image, the other images also have the same wavelength characteristics as the cartilage part (region specified by the user) based on the learning data, and may be cartilage. The high part can be shown in red.

Next, a meat mass called end meat in the present embodiment will be described with reference to FIGS. 11 (a) and 11 (b). FIG. 11A shows an example of block meat, and FIG. 11B shows an example of end meat. FIG. 11A shows block meat containing lean meat and fat meat. In the block meat, since the positional relationship of each part is known, for example, it may be possible to know that the lean and fat block meat shown in FIG. 11 (a) does not contain cartilage without inspection. is there. As shown in FIG. 11B, in the end meat, the finely cut portion is a lump regardless of the positional relationship of the original portion.

Next, the image display processing in the present embodiment will be described. This process is realized by expanding the program recorded in the non-volatile memory 103 into the memory 102 and executing it by the CPU 101. This process starts when the power of the PC 100 is turned on and the display on the display unit 112 and the acquisition of learning data become possible.

In S201, the CPU 101 displays a main screen for performing image display processing on the display unit 112. FIG. 3 shows an example of the main screen 301 in the present embodiment.

In S202, the CPU 101 displays RGB images with seven wavelengths. An RGB image is displayed as a result of passing an image (input image) input to the user through seven wavelengths of 470, 540, 610, 710, 770, 820, and 920. The RGB image 302 of FIG. 3 shows an example of the image displayed in S202. The input image is an image that the user has previously captured in the PC 100, but an image obtained by capturing an object inspected by the inspection system 200 may be newly selectable.

In S203, the CPU 101 displays a predicted image predicted by using AI (artificial intelligence) based on the RGB image in S202. FIG. 3 shows an example of the predicted image 303. The predicted image 303 is more likely to be displayed in a color closer to the actual determination result than the RGB image 302 composed of only seven wavelengths.

In S204, it is determined whether or not the user selects a point in the RGB image and gives an instruction to display a wavelength graph of the selected point. The user can select a point on the RGB image by moving the cursor, for example, points 302a, b, c, d. Up to four points can be selected, and four colors of Red, Green, Blue, and Orange are associated and selected as shown in the color selection frame 305. 1. 1. When any point of the RGB image 302 is selected after selecting Red, the wavelength graph of the selected point is displayed in red. If it is determined in S204 that the user has instructed to display the wavelength graph of the selected point, the process proceeds to S205, and if not, the process proceeds to S206.

In S205, the CPU 101 displays the wavelength graph of the points selected in S204 on the display unit 112. The wavelength graph 307 of FIG. 3 shows an example of a wavelength graph of four selected points. That is, it is a graph in which the results of passing four points 302a to d through seven wavelength filters are combined. The horizontal axis shows the wavelength and the vertical axis shows the intensity. From the wavelength graph 307, it can be seen that the wavelengths of the points 302a and 302d are close to each other, and the wavelengths of the points 302c and 302d are close to each other.

In S206, the CPU 101 determines whether or not an instruction to display an image of each wavelength has been given. The instruction to display the image of each wavelength can be given by selecting the item 306 (Display 7 Times) of FIG. The graph of each wavelength is an image passed through each of the above-mentioned seven wavelength filters. If it is determined that the instruction to display the image of each wavelength has been given, the process proceeds to S207, and if not, the process proceeds to S208.

In S207, the CPU 101 displays an image that has passed through each wavelength filter. FIG. 4 shows an example of an image that has passed through each wavelength filter. The RGB image 302 of FIG. 3 is an image obtained by combining the images passing through each of these wavelengths.

In S208, the CPU 101 determines whether or not an instruction to set machine learning has been given. The machine learning instruction can be given by selecting the item 308 (Setting) of FIG. 3 and selecting the item 501 of the setting screen 502 of FIG. 5 displayed by selecting the item 308. If it is determined that the machine learning instruction has been given, the process proceeds to S209, and if not, the process proceeds to S220.

In S209, the CPU 101 reads an image and acquires wavelength information on the display color setting screen 601 in which the user sets a color for displaying the determination result based on the learning data. The image to be read can be selected from the image of the object whose wavelength data has been acquired by the inspection system 200. If there is no wavelength data, it is not possible to associate the wavelength with the color in the first place, so it cannot be used as learning data. An image with wavelength data is learned by the user by associating the area / position of the image with the color so that, for example, the area having the wavelength characteristic of the cartilage part can be displayed in a specified color. Can be done. In S209, the wavelength information of the object in the image (wavelength information of each position in the image) is acquired together with the image. Since these pieces of information are acquired in the inspection system 200 described with reference to FIGS. 9 to 11, the CPU 101 acquires wavelength information from the database 203 on the inspection system 200 side. When the PC 100 and the PC 240 in the inspection system 200 are the same, the CPU 201 acquires the wavelength information and the image from the database 203.

In S210, the CPU 101 determines whether or not a color is selected from the colors in the classification 602. If it is determined that the color has been selected, the process proceeds to S211. If not, the process waits until the color is selected. Up to 10 colors can be selected. By selecting a color, the position (wavelength) selected in S211 and the color can be set in association with each other (can be set). In classification 602, there are two for each color. This is useful, for example, when the user wants to detect multiple types of parts, cartilage and bone, to remove them from a mass of meat. For example, when both the cartilage and the bone part are learned as Red, the part to be detected by the user is displayed as Red. Therefore, the user can determine whether or not there is a part to be removed from this meat mass just by looking at the image. Alternatively, it is possible to quickly grasp how much of the part to be removed occupies.

In S211 the CPU 101 determines whether or not the position is selected from the image 603 (whether or not a part of the area of the image is selected). The position 604 of the image 603 is an example. After selecting a color in the classification 602, the position can be selected by moving the frame (frame 604a indicating the position 604) displayed in the selected color. If it is determined that the position has been selected, the process proceeds to S212, and if not, the process waits until the position is selected. The position selected in S211 becomes the teacher data. Therefore, if the user wants to display the cartilage in Red, 2. Red or 12. After selecting Red, by selecting the position that seems to be cartilage in the image 603, it can be accumulated as learning data. If the color is reselected in S210 here, the position corresponding to the newly selected color can be selected. FIG. 7 is a diagram showing the wavelengths of the positions 604 and 606. FIG. 7A shows the wavelengths that have passed through the seven filters at position 604, and FIG. 7B shows the wavelengths that have passed through the seven filters at position 606. For example, at position 604, the wavelength 470 is large and the wavelengths 610 and 710 are small. At position 606, the wavelengths 710 and 770 are large and the wavelength 470 is small. In this way, the wavelength characteristics differ depending on the position in the image. By learning the characteristics of the wavelength in this way, the system described later can determine which region of the newly selected image is cartilage or lean meat.

In S212, the CPU 101 determines whether or not an instruction to change the size of the position selected in S211 has been given. The position in S211 indicates an area to be the teacher data, but the size of the teacher data (image area) can be changed by using the bar 605. The larger the size of the teacher data, the longer it takes to process because the learning is performed according to the wavelength in the selected region. On the other hand, if the size of the teacher data is too small, the detection accuracy may decrease. That is, even if there are 500 wavelength patterns of the portion likely to be cartilage, if there is only one teacher data, there is a high possibility that the cartilage region cannot be detected accurately. If it is determined that the instruction to change the size of the selected position has been given, the process proceeds to S213, and if not, the process proceeds to S214.

In S213, the CPU 101 changes the size of the teacher data.

In S214, the CPU 101 associates the wavelength of the selected position with the color being selected and stores it in the memory 102 as learning data. The selected position and the color may be associated with each other. As the learning data, data created by the user can be accumulated not only for one image but also for a plurality of images.

In S215, the CPU 101 determines whether or not to train the system based on the learning data. Here, the system is a system that performs machine learning, and can determine which part is included based on the learning data. To train the system, it can be done by selecting item 607. If it is determined that the system is to be trained, the process proceeds to S216, otherwise the process returns to S210, and other colors and positions can be set as training data.

In S216, the CPU 101 displays the image 608 as a result of the determination on the image 603 by the system learned based on the learning data stored in the memory 102 as the current learning data. That is, if the user determines the image based on the current learning data while the user is storing the learning data using the image 603, what kind of determination result is obtained is displayed. From this, the user can know whether or not the region recognized as cartilage in the image 603 is determined as cartilage by the system (whether it has the same classification) by looking at the image 608. When there are three places that the user recognizes as cartilage, the user selects one of them, and 0. Learn by associating with Red. At this time, by checking the image 608, if the other two places recognized as cartilage by the user are also displayed in Red (red), it is highly possible that the system can recognize the cartilage. I understand. On the other hand, when the other two places are displayed in other colors, it can be seen that the amount of training data may be insufficient or the accuracy of the training data may be low. Therefore, the user can confirm whether the region selected in S211 is really a cartilage region and does not include other regions such as lean meat, and can increase the size of the teacher data. Then, by letting the system learn again in S215, it is possible to know whether or not the determination result requested by the user is obtained from the current learning data. Furthermore, since there is no training data in the uncolored part, it is possible to immediately grasp which part of the training data is lacking. Further, the display of S216 may be performed each time the position is selected in S211. In that case, the result image of S216 is updated every time the size of the teacher data is changed in S212.

In S217, the CU 101 determines whether or not the learning data is stored (stored) in the memory 102. The training data can be saved by selecting item 608. If it is determined that the training data is stored in the memory 102, the process proceeds to S218, and if not, the process proceeds to S219.

In S218, the CPU 101 stores the learning data so far in the memory 102.

In S219, the CPU 101 determines whether or not to end the machine learning setting. The end of the machine learning setting can be performed by closing the display color setting screen 601 by selecting the item 609. If it is determined that the machine learning setting is finished, the process proceeds to S224, and if not, the process returns to S210.

In S220, the CPU 101 determines whether or not to analyze the image. That is, based on the learning data stored in S218, it is determined whether or not to analyze a new image. Image analysis can be performed by selecting item 309 (Analyze) on the main screen 301. If it is determined that the image is to be analyzed, the process proceeds to S221, and if not, the process proceeds to S224.

In S221, the CPU 101 reads an image on the image analysis screen displayed when S220 is Yes. FIG. 8 shows an image analysis screen 801. When the item 802 (Open Images) on the image analysis screen 801 is selected, the image can be selected and read. Image 803 shows an example of the read image.

In S222, the CPU 101 reads the learning data from the memory 102. The training data can be changed by selecting item 804 (Add). For example, the learning data changes depending on the origin and type of meat. The algorithm can be changed by selecting the pull-down 805 on the image analysis screen 801. The algorithm to be selected can be selected even if it is different from when the training data was created. Further, when the item 806 (Tech) is selected, the system learns the training data based on the selected algorithm. When the item 807 (Draw All) is selected after the system has learned the training data, the determination result of the image 803 is displayed as the image 808 (S223). In the image 808, each part is displayed in a color set by the user. As a result, the user can immediately grasp which part is included by looking at the image without checking the meat mass in detail. Alternatively, even if the user is not the user, the processing may be automatically changed according to the percentage of the total red color using the image 808 of the determination result.

In S224, the CPU 101 determines whether or not to end the image display process. If it is determined that the image display process is finished, the flowchart of FIG. 2 is finished, and if not, the process returns to S208.

According to the embodiment described above, since the user can obtain the determination result of the image based on the currently generated learning result during the generation of the learning data, it becomes easier to generate the learning data more accurately. Currently, by displaying the image in which the user associates the position and the color as learning data and the result of the system's judgment on the same image side by side, does the user's recognition and the judgment result match? Is easy to grasp. That is, since the determination result based on the learning data can be seen in real time, it is possible to confirm the lack and accuracy of the learning data. Therefore, the intended result cannot be obtained after all the training data has been input, and the possibility that the training data must be input again and again is reduced.

In addition, although the above-described embodiment has been described by taking color as an example as a display method corresponding to each part, the present embodiment is not limited to this, and the display form may be changed such as dots, fills, and diagonal lines. Further, the colors to be displayed may not be selected by the user, and the colors may be assigned in a predetermined order in the order in which the positions are selected in S211.

Further, in the above-described embodiment, the description is made for a chunk of meat, but the description is not limited to this, and it can be used for various foods such as fruits, vegetables, fish, grains, water, oil, spices, grains, and seaweed. is there. Furthermore, it goes without saying that it is effective not only in food products but also in recycling factories.

The various controls described above as those performed by the CPU 101 may be performed by one hardware, or the entire device may be controlled by sharing the processing among a plurality of hardware.

Further, although the present invention has been described in detail based on the preferred embodiments thereof, the present invention is not limited to these specific embodiments, and various embodiments within the scope of the gist of the present invention are also included in the present invention. included. Furthermore, each of the above-described embodiments is merely an embodiment of the present invention, and each embodiment can be combined as appropriate.

Further, in the above-described embodiment, the case where the present invention is applied to the PC100 has been described as an example, but this is not limited to this example, and is an electronic device capable of acquiring the wavelength information of the image and the subject in the image. If there is, it is applicable. That is, the present invention can be applied to mobile phone terminals, portable image viewers, digital photo frames, music players, game machines, electronic book readers, tablet PCs, smartphones, projection devices, home appliances devices having a display unit, and the like. ..

(Other embodiments)
The present invention is also realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiment is supplied to the system or device via a network or various recording media, and the computer (or CPU, MPU, etc.) of the system or device reads the program code. This is the process to be executed. In this case, the program and the recording medium storing the program constitute the present invention.

Claims (12)

A selection means that can select a part of the area of the first image,
A setting means for setting one of a plurality of display forms in association with the information of the selected area, and
A second image corresponding to the first image in which an area different from the selected area and the selected area of the first image are displayed in a display form set by the setting means is displayed. , An electronic device comprising a control means for controlling display together with the first image. The electronic device according to claim 1, further comprising an acquisition unit capable of acquiring information on the wavelength of the area selected by the selection means. The electron according to claim 2, wherein the control means controls so that the second image is displayed based on the wavelength information of the first image acquired by the acquisition means. machine. The control means displays the wavelength information of the area selected by the selection means and the area having the wavelength information of the same classification as the selected area in the display form set by the setting means. The electronic device according to claim 2 or 3, wherein the electronic device is controlled. The electronic device according to any one of claims 2 to 4, wherein the acquisition means also acquires information on the wavelength of an area not selected by the selection means in the first image. The electronic device according to any one of claims 1 to 5, wherein the selection means can change the size of the area to be selected in the first image. The electronic device according to any one of claims 1 to 6, wherein the setting means can be set by associating an area selected by the selection means with a color. The electronic device according to any one of claims 1 to 7, wherein the setting means can set a display form for a plurality of areas. The control means any one of claims 1 to 8, wherein the control means controls the area selected by the selection means as shown in a display form corresponding to the display form set by the setting means. Electronic equipment described in. A selection step that allows you to select a part of the first image,
A setting step for setting one of a plurality of display modes in association with the information of the selected area, and
A second image corresponding to the first image in which an area different from the selected area and the selected area of the first image are displayed in the display form set in the setting step is displayed. , A method of controlling an electronic device having a control step for controlling display together with the first image. A program for causing a computer to function as each means of the electronic device according to any one of claims 1 to 9. A computer-readable storage medium containing a program for causing the computer to function as each means of the electronic device according to any one of claims 1 to 9.

Images