JP6122655B2 - Agricultural crop selection apparatus and crop selection method - Google Patents

Description

  The present invention relates to a crop selection apparatus and a crop selection method for crops.

  Conventionally, as an example of a crop selection method, for example, an inspector visually confirms the appearance of a crop and determines whether the crop is good or bad (selection of the crop). 2. Description of the Related Art A crop selection device that recognizes the characteristics of a crop based on the crop data and makes a pass / fail judgment has been proposed. In such a crop selection device, the crops carried by the belt conveyor or the like are imaged by a camera or the like, and the computer recognizes the size, scratches, colors, scratches, etc. of the crops based on the captured information, and is constant There are those that distribute crops based on these criteria.

  An apparatus capable of recognizing and learning the appearance (size) of agricultural products is known (Japanese Patent Laid-Open No. 8-249475). In this apparatus, a crop is imaged by a camera, and a computer learns an image of a past crop and recognizes the appearance of the crop.

  However, the above device can recognize the appearance of the entire crop, but since the image of the crop is not processed such as division, components such as shimeji umbrellas and grape grains are not included. It is not suitable for selecting crops that exist in a large number of groups.

JP-A-8-249475

  The present invention has been made in view of such circumstances, and it is an object of the present invention to provide a crop selection device and a crop selection method for accurately selecting the crop by accurately recognizing the appearance of the crop having a large number of components in the assembled state. Is.

  The crop selection apparatus of the present invention made to solve the above problems includes an imaging unit that captures a crop having a large number of components in an aggregated state, and a storage unit that stores selection criterion data serving as a criterion for selection of the crop. Appearance extraction means for extracting the appearance data of the crops from the image captured by the imaging unit, and selection means for selecting the crops by comparing the appearance data extracted by the appearance extraction means with the selection reference data The appearance extraction unit recognizes the component from the image, and the selection unit selects the crop by comparing the data of the recognized component with the selection reference data.

  In the crop selection apparatus, the imaging unit captures a crop having a large number of constituent parts, and the processing unit recognizes the constituent parts of the crop from the image captured by the imaging unit. Can be ascertained for each component. Thus, the agricultural product can be selected accurately by comparing the data of this component part with the selection criterion data stored in the storage unit and serving as a criterion for selecting the crop.

  Further, it is preferable that the appearance extraction unit expresses the crop by a wire frame technique, and the selection unit selects the crop by comparing the expressed wire frame of the crop with the selection reference data. By expressing the crops using the wire frame technique, the connected state of the constituent parts of the crops can be grasped, so the density and the degree of variation of the constituent parts of the crops can be recognized. Can be used.

  Furthermore, it is preferable that the selection means selects a crop by comparing the number of the wire frames with the selection reference data. Since the number of constituent parts in a continuous state of the crop can be recognized from the number of wire frames, the density of the constituent parts of the crop can be recognized.

  Moreover, it is preferable that the selection means selects a crop by comparing the number of wire frames divided for each region with the selection reference data. By dividing the number of wireframes for each region of agricultural products and calculating the number, it is possible to recognize the number of components that are connected to each region of the crop, so the density of the components can be determined for each region. Can be recognized.

  It is preferable that the selection means selects a crop by comparing the number of end portions of the wire frame with the selection reference data. Since the number of locations where the continuous state of the constituent parts of the crop is interrupted can be grasped from the number of ends of the wire frame, the degree of variation of the constituent parts of the crop can be recognized.

  Moreover, it is preferable that the selection means selects a crop by comparing the frame length of the wire frame with the selection reference data. Since the distance between the constituent parts of the crops in a continuous state can be grasped from the frame length of the wire frame, the density of the constituent parts of the crops can be recognized.

  Moreover, in the said crop selection apparatus, it is preferable that the said process part further has a learning means which learns the external appearance data of the crop selected by the said selection means, and updates the said selection reference data. With such a configuration, it is possible to select a crop with reference to past selection results of selecting a crop, and it is possible to improve the accuracy of crop selection.

  Moreover, it is preferable that the selection means selects a crop by comparing the color information of the component and the selection reference data. By comparing the color information of the constituent parts of the crop with the selection reference data, it is possible to accurately select the crop based on the color information for each constituent part.

  Furthermore, it is preferable that the selection means selects a crop by comparing the area information of the image of the component part with the selection reference data. By comparing the area information of the constituent parts of the crop with the selection reference data, it is possible to recognize the state of variation in the area of each constituent part and to select the crop.

  Moreover, in the said crop selection apparatus, it is preferable that the said selection means uses the area information of the image of the said whole crop as reference information for crop selection. By using the area information of the image of the entire crop as reference information for selecting the crop, it is possible to select the crop from the area of the entire crop.

The crop selection method for selecting a crop according to the present invention includes a step of imaging a crop having a large number of components in an aggregated state, and a processing unit of the crop selection apparatus extracts appearance data of the crop from the captured image. And the step of selecting the crop by the processing unit by comparing the selection reference data stored in the storage unit of the crop selection device and the appearance data of the extracted crop,
In the appearance data extraction step, the component is recognized from the captured image,
In the selection step, a crop is selected by comparing the recognized component data and the selection reference data.

  In the crop selection method, an image of a crop having a large number of components in an aggregated state is captured, and the processing unit of the crop selection device recognizes the component of the crop from the captured crop image, whereby the appearance of the crop is displayed. Can be grasped for each component. Thereby, a crop can be selected accurately by comparing the data of this component and the selection reference data stored in the storage unit of the crop selection device.

  The “end of the wire frame” means a portion where a straight line expressed by the wire frame technique is interrupted without intersecting with another straight line. The “frame length of the wire frame” means the length of each straight line expressed by the wire frame technique.

  As described above, the crop selection apparatus and the crop selection method for selecting a crop according to the present invention accurately recognize the appearance of a crop having a large number of components in an aggregate state and select a crop.

It is a schematic diagram which shows the crop selection apparatus of this invention. It is a process figure at the time of selecting the crop in the crop selection apparatus of FIG. It is a process figure at the time of grasping | ascertaining the characteristic of the crop in the crop selection apparatus of FIG. It is an image figure of the image of the crop at the time of extracting the component part of the crop in the crop selection apparatus of FIG. It is an image image figure of the crop which expressed the crop by the wire frame in the crop selection apparatus of FIG. It is a figure which shows the structure of the selection reference | standard data memorize | stored in the memory | storage part in the crop selection apparatus of FIG.

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

  As shown in FIG. 1, the crop selection device 1 includes an imaging unit 2, an information processing device 3, and a mounting table 4. On this mounting table 4, a crop 5 as a selection target is mounted.

<Imaging unit 2>
The imaging unit 2 images a crop 5 as a selection target. The imaging unit 2 can employ a known device such as a CCD camera, and an image captured by the imaging unit 2 is sent to the information processing device 3. In addition, the imaging unit 2 can include a light source that irradiates the crop 5 in order to accurately capture the crop 5 as the selection target.

<Information processing apparatus 3>
The information processing apparatus 3 includes a processing unit 31 and a storage unit 32. The information processing apparatus 3 is specifically a computer, extracts the features of the crop 5 from the image captured by the imaging unit 2, and selects the crop 5. The information processing device 3 learns the characteristics of the extracted crop 5 and the selected result, and uses the learned information for selecting the crop 5.

  The information processing apparatus 3 includes an appearance extraction unit 31a, a selection unit 31b, and a learning unit 31c. The processing unit 31 functions as various units of the appearance extraction unit 31a, the selection unit 31b, and the learning unit 31c based on a program stored in the storage unit 32 described later.

  The appearance extraction unit 31 a extracts appearance data having information about the appearance characteristics of the crop 5 from the image captured by the imaging unit 2.

  The selection means 31b selects the crop 5 from the appearance data of the crop 5 extracted by the appearance extraction means 31a. When selecting the crop 5, the crop 5 is selected by comparing the appearance data extracted by the appearance extraction unit 31 a and having information on the appearance characteristics of the crop 5 and the selection reference data stored in the storage unit 32. .

  Further, the result of selecting the crop 5 is output from the information processing apparatus 5 and the crop 5 is sorted. The method for sorting the crops is not particularly limited, but the output information may be transmitted to the mounting table 4 so that the mounting table 4 can automatically sort the crops 5, and the inspector can output the output information from the information processing device 3. You may confirm and distribute the farm products 5. The selection result can be two types (for example, non-defective product / defective product), or three or more types (for example, grades of agricultural products, etc.), and the design can be changed as appropriate.

  The learning unit 31c learns the appearance data extracted by the appearance extraction unit 31a and the selection result selected by the selection unit 31b, and updates the selection reference data.

  The storage unit 32 stores appearance data and selection criterion data. As the appearance data, data extracted by the appearance extraction means 31a as described above is stored. The selection criterion data is created and updated by the person in charge of inspection, and is updated by the learning means 31c.

  The storage unit 32 stores various programs used in the appearance extracting unit 31a, the selecting unit 31b, and the learning unit 31c. The various programs are programs for realizing the processing described in the inspection method described later, and programs suitable for the information processing apparatus 5 are used.

<Mounting table 4>
The mounting table 4 is a table on which a crop 5 as an inspection object is mounted. The crop 5 placed on the mounting table 4 is imaged by the imaging unit 2. Although the mounting table 4 is not particularly limited, the mounting table 4 may be a stationary table, for example, a conveyance table for moving the crop 5 such as a belt conveyor.

  Moreover, when the mounting table 4 is a conveyance table, the structure which has the lane which the crop 5 from the process part 31 distributes the crop 5 based on selection information is employable.

<Crop 5>
In the crop selection apparatus 1, the crop 5 having a large number of components in the assembled state is a suitable selection target. Examples of suitable crops for this crop selection include, for example, a multitude of umbrellas and patterns that are a plurality of constituent parts such as shimeji, and a grain that is a plurality of constituent parts such as grapes. Many of them are densely populated.

<Selection method>
Next, the selection method of the crop 5 in the said crop selection apparatus 1 is demonstrated.

  As shown in FIG. 2, the crop selection apparatus 1 has a selection process of (I) imaging, (II) appearance extraction, (III) selection, and (IV) learning.

(I) Imaging The crop 5 placed on the mounting table 4 is imaged by the imaging unit 2. The captured image is transmitted to the information processing device 3. In addition, when the mounting table 4 is configured to move the crop such as a belt conveyor, the crop 5 is imaged by the imaging unit 2 when the crop 5 comes to a predetermined position.

  The image captured by the imaging unit 2 may be a still image or a moving image. However, when the image captured by the imaging unit 2 is a moving image, after the moving image is sent to the information processing device 3, an image in which the crop 5 is captured is extracted from the moving image and is made a still image. The

  The image captured by the imaging unit 2 may be a digital image or an analog image. However, when an analog image is captured, the image is transmitted to the information processing apparatus 3 and then converted into a digital image. A known method can be adopted for the conversion to a digital image, for example, a method of sampling and quantizing an analog image.

(II) Appearance extraction An image captured by the imaging unit 2 is sent to the information processing device 3 and stored in the storage unit 32. As described above, if the image captured by the imaging unit 2 is a moving image, a still image is extracted from the moving image, and if it is an analog image, it is converted to a digital image.

  The image stored in the storage unit 32 is read by the processing unit 31. And as shown in FIG. 3, the process part 31 is (1) extraction of the whole crop, (2) extraction of a component part, (3) acquisition of color information, (4) acquisition of area information, (5) wire A feature of the appearance of the crop is extracted from the image read by each step of expression by frame, (6) acquisition of the number of wire frames, and (7) acquisition of the number of ends of the wire frame.

  In the step of extracting the entire crop, the processing unit 31 first performs an enhancement process on the image read by the processing unit 31 to generate an enhanced image in order to emphasize the entire crop 5. A known method can be employed as the enhancement processing method, for example, a histogram equalization method, an unsharp processing method, or the like can be employed.

  Next, the processing unit 31 performs binarization processing using the enhanced image on which the enhancement processing has been performed, and extracts regions that are candidates for the entire region of the crop 5. A known method can be adopted as a threshold value setting method when performing binarization processing, and a P-tile method, a mode method, a discriminant analysis method, or the like can be adopted.

  Then, the processing unit 31 performs feature extraction from regions that are candidates for the entire region of the crop 5, and extracts the entire region of the crop 5. A known method can be employed as the feature extraction method, and for example, edge extraction, region division extraction, texture extraction, or the like can be employed. Thereby, the area | region of the whole crop 5 can be extracted.

  Furthermore, as shown in the image diagram of FIG. 4, the processing unit 31 sets an entire circle 41 that includes the entire region of the entire extracted crop 5 and is minimized. The area inside the set whole circle 41 is defined as the whole crop 5.

  The processing unit 31 performs a smoothing process on the entire region of the crop 5 extracted by feature extraction. As this smoothing method, a known method can be adopted, and a moving average method, a median filter, or the like can be adopted. By this smoothing process, it is possible to remove the noise in the region of the entire crop 5, thereby completing the extraction step for the entire crop.

  In the component extraction step, the processing unit 31 divides the region from the entire region of the crop 5 from which noise has been removed as described above. A known method can be adopted as the region dividing method, and a histogram thresholding method, a clustering method, or the like can be adopted. Thereby, the image of the whole crop 5 can be divided into a plurality of regions having the same characteristics.

  The processing unit 31 performs feature extraction from the region-divided image, and extracts the region of the constituent part of the crop 5. As this feature extraction method, a known method can be adopted, and for example, edge extraction, texture extraction, or the like can be adopted. Thereby, the area | region of the component part of the crop 5 can be extracted.

  Furthermore, as shown in the image diagram of FIG. 4, the processing unit 31 sets a component circle 42 that is the maximum circle that falls within the region of the component of the agricultural product 5. Then, the region inside the set component circle 42 is set as a component of the crop 5.

  In the color information acquisition step, the processing unit 31 acquires RGB data of a region corresponding to the component part circle 42 set in the component part extraction step from the image of the crop 5 read from the storage unit 32. The acquired RGB data can be used as the color information of the constituent parts of the crop 5. The RGB data is not particularly limited. For example, RGB components can be expressed by values from 0 to 255.

  Further, the RGB data is stored in the storage unit 32 as data of the crop component which is a part of the appearance data. Specific data stored in the storage unit 32 is based on (III) selection described later. Data corresponding to the selection criteria may be stored. In the present embodiment, the maximum RGB data among the components of the crop 5 is stored in the storage unit 32 as a part of the appearance data. In addition, depending on the selection criteria by (III) selection, the minimum value, the average value, the intermediate value, etc. may be stored as a part of the appearance data among the constituent parts of the crop 5.

  In the area information acquisition step, the processing unit 31 acquires the number of pixels of the diameter portion of the whole circle 41 set in the entire crop extraction step from the image of the crop 5 read from the storage unit 32. The acquired number of pixels can be used as area information of the entire crop. The number of pixels is stored in the storage unit 32 as part of the appearance data. In addition, depending on the selection criteria by (III) selection, the number of pixels in the area of the entire circle 41 can be acquired and used as the area information of the entire crop 5.

  Further, the processing unit 31 acquires the number of pixels of the diameter portion of the component circle 42 set in the component extraction step from the image of the crop 5 read from the storage unit 32. The acquired number of pixels can be used as area information of the constituent parts of the crop 5. The number of pixels in the region of the component circle 42 may be acquired and used as the area information of the component.

  In addition, the number of pixels is stored in the storage unit 32 as data of a component part of the crop that is a part of the appearance data, but specific data stored in the storage unit 32 is based on (III) selection described later. What is necessary is just to memorize | store according to selection criteria. In this embodiment, the difference between the maximum area and the minimum area of the constituent parts corresponding to the central part of the entire crop 5 is stored in the storage unit as a part of the appearance data. Specifically, the minimum number of pixels and the maximum number of pixels of the component partial circle 42 that exist in the region where the diameter of the whole circle 41 set in the whole crop extraction step is half are acquired, and the difference is stored in the storage unit 32. . Thereby, it can be grasped | ascertained whether the magnitude | size of the structure part of the center part of the whole crop 5 has variation.

  The processing unit 31 acquires the component part circles 42 in which the component part circles 42 set in the component part extraction step overlap in the wire frame representation step, and the respective centers 51 of the overlapping component part circles 42. Are connected by a straight line 52. As shown in FIG. 5, by performing this process on all the component part circles 42, it is possible to represent the crop by the wire frame technique.

  In the step of acquiring the number of wire frames, the processing unit 31 acquires the number of straight lines 52 in which the centers 51 of the component circles 42 of the crop 5 are connected from the information expressed by the wire frames as described above. . The obtained number is used as the number of wire frames of the crop 5, and this number is stored in the storage unit 32 as data of a constituent part of the crop that is a part of the appearance data.

  Further, the processing unit 31 divides the image of the crop 5 read from the storage unit 32 into four parts, and acquires the number of the wire frames described above for each of the divided regions. The acquired number is stored in the storage unit 32 as data of a constituent part of the crop that is a part of the appearance data.

  In the acquisition step of the number of ends of the wire frame, the processing unit 31 intersects the straight line 52 where the centers 51 of the component part circles 42 of the crop 5 are connected to other straight lines from the information expressed by the wire frame. The number of the end parts 53 which are the part interrupted without being acquired is acquired. The number of the acquired end portions can be set as the number of end portions 53 of the crop wire frame, and the number of the end portions 53 can be stored in the storage unit 32 as data of the constituent parts of the crop that is a part of the appearance data. .

(III) Selection The processing unit 31 selects the crop 5 by comparing the appearance data with the selection reference data.

  The appearance data is extracted by (II) appearance extraction as described above. On the other hand, the selection reference data is first input to the information processing apparatus 3 by the inspector as teacher data for selecting the crop 5 and stored in the storage unit 32.

  As shown in FIG. 6, the selection reference data includes feature data of the appearance of the crop 5 and selection information of the crop 5 (“selection” item in FIG. 6). The feature data of the appearance of the crop 5 in the selection criterion data is matched with the item of appearance data extracted by (II) appearance extraction.

  As described above, the feature data of the appearance of the crop 5 in the present embodiment includes the color of the constituent portion of the crop 5 (maximum RGB data), the difference between the minimum area and the maximum area of the constituent portion of the central portion of the crop 5, The area of the entire crop 5 (the number of pixels in the diameter portion), the number of wire frames of the crop 5, the number of wire frames for each region, and the number of ends of the wire frames are employed.

  In the selection reference data, the feature data of the appearance of a plurality of these crops 5 and the selection results of the crops 5 are registered.

  The processing unit 31 compares the appearance data extracted by (II) appearance extraction with the selection reference data, and searches for selection reference data similar to the appearance data. As a search method, a publicly known method can be adopted, for example, SVM method, PA method and the like can be adopted.

  Then, the processing unit 31 identifies the selection criteria data that is most similar, and sets the selection result of the specified selection criteria data as the selection result for selecting the crop that is the inspection object. Thereby, crops can be selected.

(IV) Learning After selecting the crop 5, the processing unit 31 adds (II) the appearance data extracted by appearance extraction and the selection result of (III) selection of the appearance data to the selection reference data. Thereby, the characteristics of the appearance of the examined crop 5 and the selection result of this crop 5 are added to the selection reference data, and can be used by (III) selection of the next crop 5 to be inspected.

<Advantages>
In the crop selection apparatus 1, the processing unit 31 recognizes the constituent parts of the crop 5 from the image of the crop 5 captured by the imaging unit 2, thereby grasping the appearance of the crop 5 for each constituent part. it can. Thereby, the crop 5 can be accurately selected in consideration of the components of the crop 5. Moreover, since the process part 31 acquires the color information and area information of the component part of the crop 5, it can select a crop from the color defect for every component part of a crop, or the variation in the magnitude | size of a component part.

  Furthermore, by representing the constituent parts of the crop 5 with a wire frame, the connected state of the constituent parts of the crop 5 can be grasped. Therefore, the density of the constituent parts is determined from the number of the wire frames, and the degree of variation of the constituent parts from the end of the wire frame. Can be grasped, the appearance of the crop 5 can be accurately recognized, and the crop can be selected.

  In addition, since the crop selection device 1 selects crops based on the number of wire frames divided into regions, it can grasp the variation of the constituent parts for each region of the crops, and the appearance of the crop 5 for each region. And can select crops.

  In addition, since the appearance data includes area information of the entire crop 5, in addition to the characteristics of the constituent parts of the crop 5, the area information of the entire crop 5 can be used as a reference.

  Further, since the crop selection device 1 learns the appearance data of the selected crop, the selection accuracy of the crop 5 is improved as the appearance of the crop 5 is inspected.

<Other embodiments>
In addition, the present invention is not limited to the configuration of the above embodiment, and includes the following embodiments.

  In the selection of the crop, the crop selection device includes, as described above, the color of the constituent portion of the crop 5 (maximum RGB data), the minimum area and the maximum area of the constituent portion of the central portion of the crop 5 as selection criteria items. The difference, the area of the entire crop 5 (the number of pixels in the diameter portion), the number of wire frames of the crop 5, the number of wire frames per region, and the number of ends of the wire frames are adopted. It is not limited to the combination, and may be determined according to the type of the crop that is the selection target, and may be only one item or a combination of any plurality of items.

  As a specific example, the flatness ratio of the shape of the entire circle of the crop can be added to the selection criterion item. This makes it possible to grasp the shape of the entire crop and select the crop. Other items such as the area of the constituent parts of the crop (the number of pixels in the diameter part) and the difference between the minimum area and the maximum area of the constituent parts in the entire crop may be added.

  Furthermore, the frame length of the wire frame can be added as a selection criterion item. The distance between the constituent parts of the crops in a connected state is obtained from the information represented by the wireframe of the crops by using the length of the straight line connecting the centers of the constituent part circles of the crops as a selection criterion. Since the density of the constituent parts of the crop can be recognized, the crop can be selected.

  In the above embodiment, the processing unit 31 divides the image of the crop 5 read from the storage unit 32 into four parts, and obtains the number of wire frames described above for each of the divided regions. The area to be performed is not limited to four divisions, and may be determined according to the type of agricultural department that is the selection target, and can be appropriately changed in design.

  Moreover, in the said crop selection apparatus, the structure in which a process part does not have a learning means may be sufficient. Thereby, the data amount of selection reference | standard data can be restrict | limited and the calculation speed of a process part can be improved.

  As described above, the crop selection device according to the present invention is useful as a device for selecting a crop, and can be used for, for example, determining the quality of a crop for determining the quality of the crop, and has a large number of components particularly in an aggregate state. It is suitably used for selecting crops.

DESCRIPTION OF SYMBOLS 1 Farm crop selection apparatus 2 Imaging part 3 Information processing apparatus 4 Mounting stand 5 Farm crop 31 Processing part 31a Appearance extraction means 31b Selection means 31c Learning means 32 Storage part 41 Whole circle 42 Component part circle 51 Center 52 Straight line 53 End

Claims (10)

An imaging unit for imaging a crop having a large number of components in a collective state;
A storage unit for storing selection standard data serving as a standard for selecting crops;
Appearance extraction means for extracting crop appearance data from the image captured by the imaging unit, and selection means for selecting a crop by comparing the appearance data extracted by the appearance extraction means with the selection reference data A processing unit,
The appearance extraction means,
Recognizing the component from the top Symbol image,
Acquire the overlapping area among the recognized area of the component part,
The wire frame of the crop is expressed by connecting the centers of the above overlapping areas with a straight line,
It said selecting means,
A crop selection device for selecting a crop by comparing the wire frame of the expressed crop with the selection reference data. The crop selection apparatus according to claim 1 , wherein the selection means selects a crop by comparing the number of the wire frames with the selection reference data. The crop selection apparatus according to claim 1 or 2 , wherein the selection means selects a crop by comparing the number of wire frames divided for each region with the selection reference data. It said selecting means, according to claim 1 for selecting a crop by comparing the number and the selection reference data of the end portion of the wire frame, crop selection apparatus according to claim 2 or claim 3. It said selecting means, crop selection apparatus according to any one of claims 1 to claim 4 selected crops by comparing the frame length and the selection reference data of the wire frame. The crop selection apparatus according to any one of claims 1 to 5 , wherein the processing unit further includes learning means for learning appearance data of the crop selected by the selection means and updating the selection reference data. The crop selection apparatus according to any one of claims 1 to 6 , wherein the selection means selects a crop by comparing the color information of the component and the selection reference data. The crop selection apparatus according to any one of claims 1 to 7 , wherein the selection means selects a crop by comparing area information of the image of the component and the selection reference data. The crop selection apparatus according to any one of claims 1 to 8 , wherein the selection unit uses area information of an image of the entire crop as reference information for crop selection. Imaging a crop having multiple components in an assembled state;
A step in which the processing unit of the crop selection device extracts the appearance data of the crop from the captured image;
The processing unit selects the crop by comparing the selection reference data stored in the storage unit of the crop selection device with the appearance data of the extracted crop,
In the appearance data extraction step ,
The component recognizing shooting from the image image,
Acquire the overlapping area among the recognized area of the component part,
The wire frame of the crop is expressed by connecting the centers of the above overlapping areas with a straight line,
In the above selection step ,
A crop selection method for selecting a crop by comparing the wire frame of the expressed crop with the selection criteria data.