JP2019061567A - Shape recognition program and shape recognition device for leaf-like crop - Google Patents

Abstract

To provide a shape recognition program and a shape recognition device for leaf-like crop that can accurately recognize the shape of a leaf-like crop.SOLUTION: The shape recognition program for a leaf-like crop causes a computer to function as a shape recognition device including: leaf area extraction means 61 for setting a scan area including a curved area curving outward in a radial direction with a center of gravity of a leaf-like crop being a center in order to extract a stem area from an inspection image of the leaf-like crop; stem area extraction means 62 for performing, after repeating a first extraction process of extracting an overlapping area overlapping a leaf-like crop in the scan area set by the leaf area extraction means and a second extraction process of extracting a small area from the overlapping area extracted in the first extraction process multiple times while expanding the scan area, an area combining process of combining a plurality of small areas extracted in the second extraction process to extract a stem area; and leaf blade area extraction means 63 for extracting, from an image of the leaf-like crop, a leaf blade area of the leaf-like crop excluding the stem area extracted by the stem area extraction means.SELECTED DRAWING: Figure 5

Description

  The present invention relates to a shape recognition program for recognizing the shape of leafy crops, and a shape recognition device.

  Conventionally, a system for performing operations such as sorting and packaging of leafy crops has been provided. For example, in Patent Document 1, the size of a leafy crop is determined from image data obtained by imaging the leafy crop, and the size is determined based on the determination result. A program has been introduced to operate the system to perform the task.

Unexamined-Japanese-Patent No. 2007-331782

  However, since the direction of foliage crops that appear in the inspection image is not always constant, if there is a variation (error) in the determination results of the shape or size of foliage crops, work such as sorting or packaging in the system is correct. Sometimes it does not happen. Therefore, when handling foliage crops, it is important to accurately recognize the shape of foliage crops.

  Then, this invention makes it a subject to provide the shape recognition program and shape recognition apparatus of foliage crops which can recognize the shape of foliage crops correctly in view of such a situation.

The foliage crop shape recognition program of the present invention is
Computer,
An area setting process for setting a scanning area including at least a curved area curving so as to expand radially outward around the center of gravity of the leaf crop in order to extract the stem area from the image of the leaf crop on the inspection image, the area A first extraction process for extracting an overlapping area overlapping with the leafy crop among the scanning areas set in the setting process, and further extracting a short area having a predetermined length or less from the overlapping area extracted in the first extraction process The second extraction process is performed after each process of the second extraction process is repeatedly performed a plurality of times while expanding the scanning area so that the curved area is displaced in the radial outer direction centering on the center of gravity A stalk area extraction unit that executes an area combination process that combines a plurality of short and small areas extracted in the above to extract the stalk area
It is made to function as a blade region extraction means for extracting a leaf blade region of the foliage crop by removing the stalk region extracted by the foliage crop region extraction means from the image of the foliage crop.

  According to the shape recognition program of the above configuration, the area of the stem and the area of the leaf blade, that is, the shape of the stem and the shape of the leaf blade are separately extracted from the image of the leaf crop on the inspection image. It becomes possible to recognize the shape accurately.

Further, the shape recognition program for foliage crops of the present invention is
The computer,
Function as a direction deriving means for deriving the direction of the leafy crop using information on the area of the stem extracted by the stem area extracting means and information on the area of the leaf extracted by the leaf area extracting means ,
The direction deriving means derives the center of gravity of the area of the stem extracted by the stem area extracting means and the center of gravity of the area of the leaf extracted by the leaf area extracting means, and the center of gravity of the area of the stem and the The direction of the straight line connecting the center of gravity of the leaf area may be determined as the direction of the leafy crop.

  The direction of foliage crops is deeply related to the direction and position of the stalks. According to the shape recognition program of the above configuration, after the stalk region is extracted by the stalk region extraction means, the direction deriving means is the direction of foliage crops Since it is comprised so that it may derive | lead-out, the direction of a leafy crop can be detected, considering the information regarding the area | region of a stalk.

  In addition, since the direction of the straight line connecting the center of gravity of the area of the stem and the center of gravity of the area of the leaf blade is determined as the direction of the leafy crop, the direction of the leafy crop can be easily determined.

The foliated crop shape recognition apparatus of the present invention is
A processing device for performing processing for recognizing the shape of the leafy crop from the inspection image on which the leafy crop is reflected;
The processing unit
An area setting process for setting a scan area including at least a curved area curving so as to expand radially outward around the center of gravity of the leafy crop in order to extract the area of the stem from the image of the leafy crop reflected in the inspection image; A first extraction process for extracting an overlapping area overlapping with the leafy crop among the scanning areas set in the area setting process, and further extracting a short area having a predetermined length or less from the overlapping area extracted in the first extraction process And the second extraction process is repeatedly performed a plurality of times while expanding the scanning area such that the curved area is displaced in the radially outward direction centering on the center of gravity. A stalk area extraction unit that executes an area combination process that combines a plurality of short and small areas extracted in the process to extract the stalk area.
And a leaf region extracting means for extracting a leaf region of the foliage crop by removing the pelvis region extracted by the foliage crop extraction means from the image of the foliage crop.

  According to the shape recognition device of the above configuration, the area of the stalk and the area of the leaf blade, that is, the shape of the stalk and the shape of the leaf blade are separately extracted from the image of the foliage crop shown in the inspection image. It becomes possible to recognize the shape accurately.

Moreover, the shape recognition apparatus of the foliated crop of the present invention is
It has direction deriving means for deriving the direction of the leafy crop using the information on the area of the stem extracted by the stem area extracting means and the information on the area of the leaf extracted by the leaf area extracting means,
The direction deriving means derives the center of gravity of the area of the stem extracted by the stem area extracting means and the center of gravity of the area of the leaf extracted by the leaf area extracting means, and the center of gravity of the area of the stem and the The direction of a straight line connecting the center of gravity of the leaf area may be determined as the direction of the leaf crop.

  The direction of the foliage crop is deeply related to the direction and position of the stalk. According to the shape recognition device with the above configuration, after the stalk region is extracted by the stalk region extraction means, the direction deriving means is the direction of the foliage crop It is possible to derive the direction of leafy crops after adding information on the area of the stalk.

  In addition, since the direction of the straight line connecting the center of gravity of the area of the stem and the center of gravity of the area of the leaf blade is determined as the direction of the leaf crop, the direction of the leaf crop can be easily determined.

  As described above, according to the shape recognition program and the shape recognition device for foliage crops of the present invention, it is possible to achieve an excellent effect that the shapes of foliage crops can be accurately recognized.

FIG. 1 is a schematic plan view of a shipping system provided with a shape recognition device according to an embodiment of the present invention. FIG. 2 is a schematic side view of the shipping system provided with the shape recognition device according to the embodiment. FIG. 3 is a block diagram of the hardware configuration of the shape recognition device according to the embodiment. FIG. 4 is a plan view of a stage of the shape recognition device according to the same embodiment. FIG. 5 is a functional block diagram of the shape recognition device according to the embodiment. In FIG. 6, (a) is an explanatory view of an inspection image, and (b) is an explanatory view of an inspection image in a state in which a background area and a leaf area are extracted. In FIG. 7, (a) is an explanatory view of a state in which a scan area is set in an inspection image, and (b) is an explanatory view of a state in which an overlapping area is extracted from the inspection image by the scan area. In FIG. 8, (a) is an explanatory view of a state in which the expanded scan area is set as an inspection image, and (b) is an explanatory view of a state in which the overlap area is extracted from the inspection image by the scan area. In FIG. 9, (a) is an explanatory view of a state in which a scan area expanded further is set as an inspection image, and (b) is an explanatory view of a state in which an overlap area is extracted from the inspection image by the scan area. In FIG. 10, (a) is an explanatory view of a state in which short and small regions are combined, and (b) is an explanatory view of a state in which a stem region is extracted. FIG. 11 is an explanatory view of a process of deriving the direction of the leafy crop by the shape recognition device according to the embodiment. FIG. 12 is a main flowchart for explaining the operation of the shape recognition device according to the embodiment. FIG. 13 is a sub-flowchart of leaf area extraction processing by the leaf area extraction means of the shape recognition device according to the embodiment. FIG. 14 is a sub-flowchart of the extraction process of the stalk region area by the stalk region area extraction unit of the shape recognition device according to the embodiment. FIG. 15 is a sub-flowchart of the process of deriving the direction of the large leaf by the direction deriving means of the shape recognition device according to the same embodiment.

  Hereinafter, a shape recognition program and a shape recognition apparatus for leafy crops according to an embodiment of the present invention will be described with reference to the attached drawings.

  The shape recognition program is configured to cause the computer to function as a leaf shape crop shape recognition device for specifying the shape of a leaf shape crop based on an inspection image obtained by imaging the leaf shape crop.

  In addition, the shape recognition device is incorporated into, for example, a foliage crop shipping system for inspecting and shipping preparations of foliage crops.

  In the present embodiment, the shape recognition device and the shape recognition program will be described by taking as an example that the shape recognition device is incorporated into the shipping system. In addition, although the leafy crop which concerns on this embodiment is a large leaf, as long as it is a leaf whose thickness dimension is small compared with planar dimension, other types of leafy crop may be sufficient.

  As shown in FIGS. 1 and 2, the shipping system 1 includes a container 2 for storing a plurality of large leaves, a shape recognition device 3 for recognizing the shape of the large leaves taken out from the container 2, and The transport device 4 transports large leaves whose shape is recognized by the shape recognition device 3, and the processing device 5 performs processing for acquiring and shipping the large leaves transported by the transport device 4.

  The container 2 is configured so as to be able to be accommodated therein in a state in which a plurality of large leaves are stacked.

  The shape recognition device 3 includes, as shown in FIGS. 2 and 3, a stage 30 on which a large leaf to be imaged is placed, an imaging device 31 for imaging a large leaf placed on the stage 30, and the imaging device A processing device 32 for recognizing the shape of the large leaf based on an inspection image which is an image of the large leaf captured in 31; and a storage device 33 connected to the processing device 32 by wire or wirelessly There is.

  On the stage 30, large leaves taken out of the container 2 are placed one by one. In the present embodiment, the container is held by the suction conveyance unit (hereinafter, referred to as a first suction conveyance unit) P1 that suctions and conveys the large leaf located at the top of the plurality of large leaves stacked in the container 2 It is carried from 2 to stage 30.

  In addition, the suction conveyance unit holds the large leaves by sucking one surface of the large leaves front or back facing upward, and when the suction state for the large leaves is released, the large leaves are separated (fall) from the suction conveyance unit. Is configured.

  Moreover, the stage 30 which concerns on this embodiment has the transparent mounting board part 300, and the large leaf taken out from the container 2 is mounted on this mounting board part 300 (refer FIG. 4).

  The imaging device 31 is disposed on the lower side of the stage 30, specifically, on the lower side of the mounting plate 300. Then, the imaging device 31 is configured to pick up an image of the large leaf placed on the placement plate portion 300 from one side (the front side in the present embodiment) of the front side or the back side facing downward. There is.

  The imaging device 31 is disposed on the upper side of the stage 30 (the mounting plate 300), and the large leaf placed on the mounting plate 300 is viewed from one side facing the upper side of the front surface or the back surface. It may be configured to capture an image.

  In the shape recognition device 3, the processing device 32 is configured to execute various means for recognizing the shape of the large leaf by executing the shape recognition program.

  Specifically, as shown in FIG. 5, the shape recognition device 3 executes the shape recognition program with the processing device 32 to obtain an inspection image, an image acquisition unit 60, and the entire area where the large leaf appears from the inspection image. A leaf area extraction means 61 for extracting a leaf area (hereinafter referred to as a leaf area), a stalk area extraction means 62 for extracting an area of a large leaf stalk (hereinafter referred to as a stalk area) based on the inspection image Function as leaf blade region extraction means 63 for extracting a leaf blade region (hereinafter referred to as leaf blade region) and direction deriving means 64 for deriving the direction of large leaves based on the stem region and the stem region It is done.

  The image acquisition means 60 is configured to acquire an image in which a large leaf for which a shape is to be recognized appears. The image acquisition unit 60 according to the present embodiment controls the imaging device 31 to capture a large leaf placed on the placement plate unit 300, and further, uses an image of the large leaf captured by the imaging device 31 as an inspection image. It is configured to get. That is, the image acquisition unit 60 is configured to acquire an inspection image directly from the imaging device 31.

  As shown in FIGS. 6 (a) and 6 (b), the leaf area extraction unit 61 determines from the inspection image T a leaf area RE1 and an area outside the leaf area RE1 (that is, an area in which the background appears). It is configured to identify the BE. In the present embodiment, the area BE outside the leaf area RE1 is referred to as a background area BE, and the following description will be made.

  More specifically, the leaf area extraction unit 61 executes background area identification processing for identifying the background area BE from the inspection image T and leaf area identification processing for identifying the leaf area RE1 from the inspection image T. It is configured.

  The leaf area extraction means 61 selects a background area from among the entire inspection image T based on background area identification information serving as a reference for determining whether each area of the inspection image T is the background in the background area identification processing. It is configured to identify BE.

  For example, information (color information) indicating a color included in the background area is set in the background area identification information, and the leaf area extraction unit 61 uses color information included in the background area identification information of the inspection image T. It may be configured to identify the area having the corresponding color as the background area BE.

  In the case where color information is set in the background area identification information, it is possible to determine the color information based on the color included in the image obtained by imaging the stage 30 in a state in which the large leaf is not placed. Further, the color information may be information indicating a single color, or information indicating a range of color gamut.

  Furthermore, the leaf area extraction unit 61 determines the leaf area based on the leaf area identification information as a reference for determining whether or not the large leaf appears in each area of the inspection image T in the leaf area identification process. It is configured to identify RE1.

  In the leaf area identification information, for example, information (color information) indicating a color is set, and the leaf area extraction unit 61 has a color corresponding to the color information included in the leaf area identification information in the inspection image T. It may be configured to identify the area as the background area BE.

  The color information to be set in the leaf area identification information can be set, for example, based on the image obtained by imaging the large leaf, and in the present embodiment, the leaf area identification information is based on the inspection image T obtained by imaging the large leaf by the imaging device 31 It is set.

  Further, the leaf area extraction unit 61 according to the present embodiment separates the area identified as the background area BE in the background area identification processing and the area identified as the leaf area RE1 in the leaf area identification processing in different colors. It is configured as shown. That is, the leaf area extraction unit 61 is configured to color-code the leaf area RE1 of the inspection image T and the inside of the background area BE with different colors.

  The stalk area extraction means 62 is an area setting process for setting a scan area including at least a curved area curving so as to expand radially outward around the center of gravity of the leafy crop, a scan area and a leaf area RE1 set in advance. A first extraction process for extracting an overlap area overlapping with the second extraction process, a second extraction process for further extracting a short or small area having a predetermined length or less from the overlap area extracted in the first extraction process, and set in advance The expansion process is performed by expanding the scan area and then performing the first extraction process, combining the plurality of short areas extracted by the second extraction process to form a combined area, combining by area combining process It is comprised so that the selection process which chooses the area | region with the longest dimension among these area | regions may be performed.

  In the area setting process according to the present embodiment, as shown in FIG. 7A, the area setting process has an annular shape centered on the center of gravity C1 of the leaf area RE1. An annular scan area PE1 is set.

  As described above, the first extraction unit is configured to extract the overlapping area of the scanning area PE1 and the leaf area RE1 as the overlapping area PE2. FIG. 7B illustrates a state in which a plurality of overlapping areas PE2 are extracted on one inspection image T. Further, the scan area set in advance refers to the scan area set in advance when the first extraction unit is executed (in the present embodiment, the scan area set in the area setting process, or expansion Scan area) expanded by the process.

  The second extraction process is performed subsequent to the first extraction process when the first extraction process extracts the overlapping area PE2.

  The plurality of overlapping areas PE2 includes a long overlapping area PE2 and a short overlapping area PE2, but the stalk area extracting unit 62 performs the short and small overlapping areas PE2 in the second extraction process. It is comprised so that it may extract as PE3. In addition, although the length used as a reference | standard which determines whether it is short area PE3 can be changed suitably, it is preferable to set in the range of 2 mm-5 mm.

  The expansion process is configured to be performed following the second extraction process. Further, as shown in FIG. 7A, FIG. 8A, and FIG. 9A, the expansion processing is performed in advance for the size of the scanning area PE1 set in advance, and the curved area is the center of gravity of the leaf area RE1. It is configured to be expanded and set so as to be displaced in the radial outward direction centering on C1. In the expansion process according to the present embodiment, only the diameter of the scan area PE1 is expanded without changing the width of the scan area PE1.

  The area combining process is configured to be performed subsequently to the second extraction process when the overlapping area PE2 is not extracted by the second extraction process. The stem area extraction means 62 repeatedly executes the first extraction process, the second extraction process, and the expansion process until, for example, the entire scan area PE1 expanded by the expansion process is positioned outside the leaf area RE1. It should just be comprised.

  As shown in FIG. 10A, the combining process is configured to superimpose and display the short region PE3 extracted a plurality of times by the second extraction process on one image. In the present embodiment, the short area PE3 displayed superimposed on one image in the joining process is referred to as a joining area PE4.

  In the inspection image T shown in FIG. 10A, a plurality of coupling regions PE4 are illustrated, but the plurality of coupling regions PE4 includes a coupling region PE4 configured of a single short region PE3, and A bonding region PE4 configured by a plurality of short regions PE3 being continuous is included.

  The selection process is configured to be performed subsequent to the area combination process. As shown in FIG. 10 (b), the stalk region extraction means 62 is configured to select the longest dimension region, that is, the stalk region RE2, of the combined region PE4 in the selection process.

  The leaf region extraction means 63 is configured to extract a leaf region RE3 based on the leaf region RE1 and the stem region RE2 extracted by the stem region extraction means 62. In the present embodiment, the leaf region RE3 is extracted from the leaf region RE1. A region corresponding to the stalk region RE2 is cut out, and the remaining region is set as a blade region RE3.

  As a result, a stem region RE2 that constitutes one leaf region RE1 and a leaf blade region RE3 are separately extracted (see FIG. 11). That is, the shape of the stem and the shape of the leaf blade are recognized separately.

  As shown in FIG. 11, the direction deriving means 64 derives the center of gravity C2 of the stem region RE2, derives the center of gravity C3 of the leaf region RE3, and directs the direction of the straight line L connecting the respective centers of gravity C2, C3 to the direction of the large leaf It is configured to be.

  The transport device 4 according to the present embodiment is configured to transport a large leaf whose shape is recognized by the shape recognition device 3 to a place where it can be delivered to the processing device 5. The delivery of the large leaves from the shape recognition device 3 to the transfer device 4 is also performed by the suction transfer unit (hereinafter referred to as a second suction transfer unit) P2 (see FIG. 1).

  The processing device 5 is configured to perform processing for obtaining and shipping the large leaf transported by the transport device 4. The delivery of the large leaves from the transport device 4 to the processing device 5 is also performed by a suction transport unit (hereinafter, referred to as a third suction transport unit) P3 (see FIG. 1).

  In the present embodiment, the processing for shipping the large leaves refers to a process of stacking a plurality of large leaves into one bundle, more specifically, a process of bundling the petiole part in a state in which the plurality of large leaves are stacked. However, for example, packaging of large leaves, processing such as tagging, etc. may be mentioned as an example.

  The configuration of the shipping system 1 according to the present embodiment is as described above. Subsequently, the large leaf shape recognition process (that is, the operation of the shape recognition device 3 when the shape recognition program is executed) by the shape recognition device 3 in the shipping system 1 will be described.

  In the shipping system 1, after the shape of the large leaf taken out from the container 2 is recognized by the shape recognition device 3, the large leaf is transported to the processing device 5 by the transport device 4.

  The shape recognition device 3 acquires the inspection image T by the image acquisition means 60 as shown in FIG. 12 (S1). In the present embodiment, an image obtained by imaging the large leaf placed on the placement plate 300 from the container 2 by the imaging device 31 is acquired as the inspection image T. Further, in the present embodiment, after obtaining the inspection image T, leaf area identification information is set based on the inspection image T.

  Then, the leaf area extraction unit 61 extracts the background area BE and the leaf area RE1 (S2). In the leaf area extraction means 61, as shown in FIG. 13, background area identification information is selected (S20), and an area of a color corresponding to color information of the selected background area identification information is identified as the background area BE S21). Further, the leaf area extraction means 61 identifies the area corresponding to the color information of the leaf area identification information as the leaf area RE1 (S22).

  As a result, the inspection image T is divided (colored) into the leaf area RE1 and the background area BE.

  Subsequently, as shown in FIG. 12, the stem region extraction means 62 extracts a stem region RE2 (S3). In the processing in which the stalk area extraction means 62 extracts the stalk area RE2, as shown in FIG. 14, the scan area PE1 is set on the inspection image T in the area setting process (S30). In the present embodiment, an annular scan area PE1 centered on the center of gravity C1 of the leaf area RE1 is set in the area setting process (see FIG. 7A).

  Next, as a result of performing the first extraction process, when the overlapping area PE2 is extracted (Yes in S31), the second extraction process further extracts the short / small area PE3 from the overlapping area PE2 (S32).

  Then, after the second extraction process is executed, the extension process extends the scan area PE1 (S33). More specifically, the expansion process expands only the diameter of the scan area PE1 without changing the width of the scan area PE1.

  Furthermore, after the extension process is performed, the first extraction process and the second extraction process are performed again using the new scan area PE1.

  The first extraction process, the second extraction process, and the extension process are repeatedly performed as long as the overlapping area PE2 is extracted in the first extraction process. When the overlapping area PE2 is not extracted in the first extraction process (No in S31), a plurality of short areas PE3 extracted in the second extraction process are overlapped on one image in the area combining process. By displaying it, it is set as the bonding area PE4 (S34). Subsequently, in the selection process, the longest dimension area is extracted as the stalk area RE2 from the plurality of bonding areas PE4 (S35).

  Thus, the stalk region RE2 is extracted from the leaf region RE1. That is, the stalk region RE2 (the shape of the stalk) and the blade region RE3 (the shape of the blade) are recognized separately.

  After the stem area extraction means 62 extracts the stem area RE2, as shown in FIG. 12, the leaf area extraction means 63 extracts a leaf area RE3 (S4).

  The blade region extraction means 63 cuts out the region corresponding to the stem region RE2 from the leaf region RE1, and sets the remaining region as a blade region RE3. As a result, the blade region RE3 is also extracted from the leaf region RE1, and the shapes of the large leaf petiole and the blade are recognized (specified).

  Subsequently, the direction deriving means 64 derives the direction of the large leaf (S5). The direction deriving means 64 derives the center of gravity C2 of the stem region RE2 (S50), and derives the center of gravity C3 of the leaf blade region RE3 (S51), as shown in FIG. 15, and connects straight centers of gravity C2 and C3. The direction is the direction of the large leaf (S52).

  As described above, according to the shape recognition device 3 according to the present embodiment, from the image of the large leaf which is a leaf crop on the inspection image T, the stem area RE2 and the leaf area RE3, that is, the shape of the stem and the shape of the leaf By separately extracting and, it becomes possible to accurately recognize the shape of the large leaf.

  Further, in the present embodiment, since the scan area PE1 has an annular shape, the joint area PE4 in which the short areas PE3 included in the overlap area PE2 are joined always includes the area to be the stalk area RE2. ing.

  Furthermore, since the outer peripheral edge of the large leaf blade has a shape in which a plurality of sharpened tips are aligned, the short and small area PE3 includes a portion corresponding to a part of the stem area RE2. In this embodiment, a process of excluding a long area from the overlapping area PE2 and an area having the longest dimension in the coupling area PE4 are selected in this embodiment. By performing the process, it is possible to extract only the stem region RE2 with high accuracy.

  Further, in the shape recognition device 3 according to the present embodiment, since the direction deriving unit 64 is configured to derive the direction of the large leaf, if the information on the direction of the large leaf derived by the direction deriving unit 64 is utilized, The accuracy of operations such as conveyance and processing performed on large leaves whose shapes are recognized by the recognition device 3 can be enhanced.

  The shipping system 1 according to the present embodiment includes the processing device 5 that bonds the stem part in a state in which a plurality of large leaves are stacked, so that the processing accuracy for binding the large leaves whose shape is recognized by the shape recognition device 3 It can be enhanced.

  The shape recognition program and the shape recognition device of the present invention are not limited to the above embodiment, and it goes without saying that various changes can be made without departing from the scope of the present invention.

  In the shipping system 1 of the above embodiment, the large leaf whose shape is recognized by the shape recognition device 3 is configured to be transported to the transport device 4, but the present invention is not limited to this configuration. The shipping system 1 may be configured, for example, to directly transport the large leaf whose shape is recognized by the shape recognition device 3 to the processing device 5 or another device.

  In the above embodiment, the shape recognition program is configured to capture an image of the large leaf by the imaging device 31. However, the shape recognition program may not be configured to control the imaging device 31 as long as an inspection image can be acquired. .

  In the above embodiment, the image acquisition means 60 is configured to acquire the inspection image directly from the imaging device 31, but the present invention is not limited to this configuration. For example, the image acquiring unit 60 may be configured to store the captured inspection image in the storage device 33 and read out the inspection image from the storage device 33 as needed. The inspection image stored in the external storage device connected in a wired manner or the inspection image stored in the external storage device capable of wirelessly communicating with the shape recognition device 3 may be read out.

  In the above embodiment, although not particularly mentioned, the shape recognition program may be stored in the storage device 33 of the shape recognition device 3 or if the processing device 32 can access, the shape recognition device 3 is wired It may be stored in a connected external storage device, or may be stored in an external storage device capable of wirelessly communicating with the shape recognition device 3.

  In the above embodiment, the background area identification information and the leaf area identification information are also stored in the storage device 33 of the shape recognition device 3. However, if the processing device 32 can be accessed, the shape recognition device 3 is connected by wire. It may be stored in an external storage device or an external storage device capable of wirelessly communicating with the shape recognition device 3.

  Although the area setting process is configured to set the annular scan area PE1 in the above embodiment, the present invention is not limited to this configuration. For example, the area setting process may be configured to set the scanning area PE1 having an elliptical ring shape, or may be configured to set the scanning area PE1 having a curved shape including only a curved area. .

  DESCRIPTION OF SYMBOLS 1 ... shipment system, 2 ... accommodation tool, 3 ... shape recognition device, 4 ... conveyance device, 5 ... processing device, 30 ... stage, 31 ... imaging device, 32 ... processing device, 33 ... storage device, 60 ... image acquisition means , 61 leaf area extraction means 62 leaf area extraction means 63 leaf area extraction means 64 direction derivation means 300 placement plate portion BE background area C1 center of gravity C2 center of gravity C3 ... Center of gravity, L ... straight line, PE 1 ... scanning area, PE 2 ... overlapping area, PE 3 ... short and small area, PE 4 ... combined area, RE 1 ... leaf area, RE 2 ... leaf area, RE 3 ... leaf blade area, T ... inspection image

Claims (4)

Computer,
An area setting process for setting a scanning area including at least a curved area curving so as to expand radially outward around the center of gravity of the leaf crop in order to extract the stem area from the image of the leaf crop on the inspection image, the area A first extraction process for extracting an overlapping area overlapping with the leafy crop among the scanning areas set in the setting process, and further extracting a short area having a predetermined length or less from the overlapping area extracted in the first extraction process The second extraction process is performed after each process of the second extraction process is repeatedly performed a plurality of times while expanding the scanning area so that the curved area is displaced in the radial outer direction centering on the center of gravity A stalk area extraction unit that executes an area combination process that combines a plurality of short and small areas extracted in the above to extract the stalk area
A program for recognizing the shape of a leafy crop which functions as a blade region extracting means for extracting a leaf region of the leafy crop by removing the stalk region extracted by the stalk region extracting means from the image of the leafy crop. The computer,
Function as a direction deriving means for deriving the direction of the leafy crop using information on the area of the stem extracted by the stem area extracting means and information on the area of the leaf extracted by the leaf area extracting means ,
The direction deriving means derives the center of gravity of the area of the stem extracted by the stem area extracting means and the center of gravity of the area of the leaf extracted by the leaf area extracting means, and the center of gravity of the area of the stem and the The foliage crop shape recognition program according to claim 1, wherein the direction of the straight line connecting the center of gravity of the leaf blade region is determined as the direction of the foliage crop. A processing device for performing processing for recognizing the shape of the leafy crop from the inspection image on which the leafy crop is reflected;
The processing unit
An area setting process for setting a scan area including at least a curved area curving so as to expand radially outward around the center of gravity of the leafy crop in order to extract the area of the stem from the image of the leafy crop reflected in the inspection image; A first extraction process for extracting an overlapping area overlapping with the leafy crop among the scanning areas set in the area setting process, and further extracting a short area having a predetermined length or less from the overlapping area extracted in the first extraction process And the second extraction process is repeatedly performed a plurality of times while expanding the scanning area such that the curved area is displaced in the radially outward direction centering on the center of gravity. A stalk area extraction unit that executes an area combination process that combines a plurality of short and small areas extracted in the process to extract the stalk area.
A leaf shape crop area recognition device comprising leaf blade area extraction means for extracting a leaf blade area of the foliage crop by removing the stalk area extracted by the stalk crop area extraction means from the image of the foliage crop. It has direction deriving means for deriving the direction of the leafy crop using the information on the area of the stem extracted by the stem area extracting means and the information on the area of the leaf extracted by the leaf area extracting means,
The direction deriving means derives the center of gravity of the area of the stem extracted by the stem area extracting means and the center of gravity of the area of the leaf extracted by the leaf area extracting means, and the center of gravity of the area of the stem and the The foliage crop shape recognition device according to claim 3, wherein the direction of a straight line connecting the center of gravity of the leaf blade region is determined as the direction of the foliage crop.