Overlap determination device, overlap determination method, and program

Abstract

To provide a technology for accurately determining an overlap of objects, while improving a processing speed.SOLUTION: An overlap determination apparatus 10 for determining an overlap of a first target object and a second target object includes: a target object region generation unit 1 which generates, on the basis of image data obtained by an imaging apparatus 101 imaging the first target object and the second target object, a circular or elliptical first target object region corresponding to the first target object and a circular or elliptical second target object region corresponding to the second target object; a synthesized region generation unit 2 which synthesizes the first target object region with the second target object region to generate a synthesized region whose center is aligned with that of the first target object region; and an overlap determination unit 3 which determines that the first target object overlaps the second target object when the center of the second target object region is included in the synthesized region.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to an overlap determination device, an overlap determination method, and a program for determining whether target objects overlap each other. [Background technology]

[0002] In product production sites, there is a demand for a system that can detect overlapping of target objects when inspecting products (target objects) moving on a conveyor belt. For example, on a line that produces shrimp fritters by feeding raw shrimp carried on a conveyor into a fryer, if the raw shrimp are not transported individually, they will stick together after passing through the fryer and produce a clump of shrimp fritters. For this reason, it is necessary to detect overlapping of target objects.

[0003] A technology for tracking a moving target object requires a detection process for detecting the position of a target object region corresponding to the target object from frame images obtained by capturing an image of the target object with an imaging device, and a tracking process for tracking the target object region. For example, Patent Documents 1 and 2 disclose the detection process and tracking process.

[0004] Patent Document 1 discloses a tracking system that tracks multiple target objects moving in a predetermined direction. According to the tracking system of Patent Document 1, the positions of multiple target objects that change over time are classified using a predetermined method. Next, the tracking system performs a tracking process to track the multiple target objects based on the classification results, and adjusts the position information based on the movement speeds of the multiple target objects.

[0005] Furthermore, Patent Document 2 discloses a face tracking method for quickly tracking faces and facial parts from moving images in devices such as mobile information terminals with limited resources. According to the face tracking method of Patent Document 2, the position of each facial part is detected from the face area of ​​a frame image. Next, in this face tracking method, the facial parts are tracked in subsequent frame images within a search area of ​​a predetermined size that is set based on the positions of the facial parts. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Patent No. 6561194 [Patent Document 2] Japanese Patent Application Laid-Open No. 2006-228061 Summary of the Invention [Problem to be solved by the invention]

[0007] The technology disclosed in Patent Documents 1 and 2 can be used to track target objects. Furthermore, target objects are detected as rectangular regions surrounded by rectangular frames. Whether target objects overlap can also be determined based on whether these rectangular regions overlap. However, even when rectangular regions overlap, target objects may not actually overlap. For example, because raw shrimp have a rounded shape, a rectangular region corresponding to the raw shrimp may include areas where no raw shrimp exists. If rectangular regions overlap in these non-existent areas, the raw shrimp may be determined to overlap even when they do not. Furthermore, attempting to perform overlap determination with greater accuracy would require more complex analysis and calculation processes, slowing down the determination processing speed.

[0008] One aspect of the present disclosure is to provide an overlap determination device, an overlap determination method, and a program that can accurately determine whether or not target objects overlap while improving processing speed. [Means for solving the problem]

[0009] In order to achieve the above object, an overlap determination device according to one aspect includes: An apparatus for determining overlap between a first object and a second object, comprising: a target object region generation unit that generates a first target object region having a circular or elliptical shape corresponding to the first target object and a second target object region having a circular or elliptical shape corresponding to the second target object based on image data obtained by capturing images of the first target object and the second target object with an imaging device; a combined region generating unit that combines the first object region and the second object region to generate a combined region whose center coincides with that of the first object region; an overlap determination unit that determines that the first object and the second object overlap when the center of the second object region is included in the combined region; The present invention is characterized by comprising:

[0010] In order to achieve the above object, an overlap determination method in one aspect includes: A method for determining overlap between a first object and a second object, comprising: generating a circular or elliptical first object region corresponding to the first object and a circular or elliptical second object region corresponding to the second object based on image data obtained by capturing images of the first object and the second object with an imaging device; combining the first object region and the second object region to generate a combined region whose center is coincident with that of the first object region; determining that the first object and the second object overlap when the center of the second object region is included in the combined region; The present invention is characterized by comprising:

[0011] Furthermore, in order to achieve the above object, the program in one aspect comprises: A program for causing a computer to determine an overlap between a first target object and a second target object, The computer, generating a circular or elliptical first object region corresponding to the first object and a circular or elliptical second object region corresponding to the second object based on image data obtained by capturing images of the first object and the second object with an imaging device; combining the first object region and the second object region to generate a combined region whose center is coincident with that of the first object region; determining that the first object and the second object overlap when the center of the second object region is included in the combined region; The method includes instructions for executing the following: [Effects of the Invention]

[0012] According to one aspect of the present disclosure, it is possible to accurately determine whether or not target objects overlap with each other while improving processing speed. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram for explaining an overview of the overlap determination device. [Figure 2] FIG. 2 is a block diagram illustrating the configuration of the overlap determination device according to the embodiment. [Figure 3] FIG. 3 is a diagram illustrating a specific configuration of the overlap determination device. [Figure 4] FIG. 4 is a diagram for explaining a state in which a rectangular area is generated on image data. [Figure 5] FIG. 5 is a diagram for explaining a state in which an elliptical target object region is generated on image data. [Figure 6] FIG. 6 is a diagram for explaining a state in which a circular target object region is generated on image data. [Figure 7] FIG. 7 is a diagram for explaining a case where a synthesis region is generated from an elliptical target object region. [Figure 8] FIG. 8 is a diagram for explaining a case where a synthesis region is generated from a circular target object region. [Figure 9] FIG. 9 is a diagram for explaining a case where overlap determination is performed using rectangular areas. [Figure 10] FIG. 10 is a flowchart showing the process of the overlap determination device. [Figure 11] FIG. 11 is a block diagram illustrating an example of a computer that realizes an overlap determination device according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, an overlap determination device according to an embodiment will be described with reference to the drawings.

[0015] First, an overview will be provided to facilitate understanding of the embodiments described below. FIG. 1 is a diagram illustrating an overview of an overlap determination device 10. The overlap determination device 10 in this embodiment is used in a product production site, for example, a food processing site that produces shrimp fritters, as described above. In this food processing site, raw shrimp 100 are transported on a belt conveyor 102 and placed in a fryer (not shown) to produce shrimp fritters. If the raw shrimp 100 are placed in the fryer in an overlapping state, defective shrimp fritters will be produced in a clumped state.

[0016] The overlap determination device 10 is a device that determines whether raw shrimps 100 are overlapping on a belt conveyor 102. An imaging device 101 is installed on the belt conveyor 102. The imaging device 101 takes an image of the belt conveyor 102 as an imaging target area, and captures images of the raw shrimps 100 carried on the belt conveyor 102. The imaging device 101 is configured with an image sensor such as a CCD (Charge Coupled Device) type or a CMOS (Complementary Metal-Oxide-Semiconductor) type, and is, for example, a camera, an infrared camera, an ultrasonic camera, or an X-ray camera.

[0017] The overlap determination device 10 acquires image data from an imaging device 101. The overlap determination device 10 performs a detection process to detect the raw shrimp 100 from the acquired image data. The detection process is a process of detecting the position of the moving raw shrimp 100 using, for example, a learning model that is trained by deep learning using captured frame images and that detects the position of the moving raw shrimp 100 with high accuracy. However, the detection process is not limited to a detection process using a learning model trained by deep learning, and may be any process that can detect the position of a target object region with high accuracy. The overlap determination device 10 uses the position of the detected target object region to determine whether the raw shrimp 100 overlap. In this embodiment, the target object region is circular or elliptical.

[0018] In the following description, the object to be determined for overlapping is described as a raw shrimp, but is not limited to this. The object may be, for example, food products (vegetables, processed foods) that do not have a uniform shape.

[0019] [Device configuration] The overlap determination device 10 according to this embodiment will be described below with reference to the drawings. FIG. 2 is a block diagram illustrating the configuration of the overlap determination device 10 according to this embodiment. As shown in FIG. 2, the overlap determination device 10 includes a target object region generation unit 1, a composite region generation unit 2, and an overlap determination unit 3.

[0020] The object region generating unit 1 generates a circular or elliptical first object region corresponding to the first object and a circular or elliptical second object region corresponding to the second object from image data obtained by capturing images of the first object and the second object with the imaging device 101. In this embodiment, the first object and the second object are raw shrimp 100.

[0021] The combined region generating unit 2 combines the first object region and the second object region to generate a combined region whose center coincides with that of the first object region.

[0022] If the center of the second object region is included in the combined region, the overlap determination unit 3 determines that the first object and the second object overlap.

[0023] As described above, in this embodiment, the overlap determination between the first object and the second object is simplified by using a composite region, thereby improving the processing speed and enabling accurate overlap determination.

[0024] Next, the overlap determination device 10 of this embodiment will be described in detail. Fig. 3 is a diagram for explaining a specific configuration of the overlap determination device 10. As described above, the overlap determination device 10 includes the target object region generation unit 1, the composite region generation unit 2, the overlap determination unit 3, and the acquisition unit 4. The overlap determination device 10 is, for example, an information processing device such as a server computer or a personal computer equipped with a CPU (Central Processing Unit), an FPGA (Field-Programmable Gate Array), or both.

[0025] The acquisition unit 4 acquires image data from the imaging device 101. The imaging device 101 captures an image of a predetermined area on the belt conveyor .

[0026] The target object region generation unit 1 executes a detection process to calculate the position of a target object region corresponding to each target object (raw shrimp 100) from the frame images of the acquired image data. The detection process is a process to detect the position of a moving product using a learning model that is trained by deep learning using the captured frame images as described above and that detects with high accuracy the position of a target object region corresponding to a moving target object in the frame images. The learning model also outputs a class that represents the product classification and a reliability (score) that represents the product-likeliness.

[0027] Methods other than deep learning include a method that combines HOG (Histograms of Oriented Gradients) features with a support vector machine, or a method that combines Haar-like features with a cascade classifier.

[0028] The target object region generating unit 1 generates a frame-shaped rectangular region 21 on the image data that encompasses the raw shrimp 100, which is the target object detected by executing the detection process. Fig. 4 is a diagram for explaining the state in which the rectangular region 21 has been generated on the image data. As shown in Fig. 4, the rectangular region 21 is generated so as to encompass all of the raw shrimp 100 on the image data.

[0029] The target object region generating unit 1 generates a circular or elliptical target object region based on the generated rectangular region 21.

[0030] 5 is a diagram illustrating a state in which an elliptical target object region 22 is generated on image data. The elliptical target object region 22 is generated based on the rectangular region 21. Specifically, the target object region generation unit 1 generates a rectangular inscribed circle centered at the center of the rectangular region 21, and sets the generated inscribed circle as the elliptical target object region 22. Note that, while the rectangular region 21 encompasses the raw shrimp 100 as shown in FIG. 4, the generated elliptical target object region 22 may not cover the entire raw shrimp 100 as shown in FIG. 5.

[0031] 6 is a diagram illustrating a state in which a circular target object region 23 has been generated on image data. The circular target object region 23 is generated based on the elliptical target object region shown in FIG. 5, which is generated based on the rectangular region 21, and the target object region on the ellipse. Specifically, the target object region generation unit 1 generates a circular target object region 23 having the same center as the elliptical target object region generated based on the rectangular region 21, and with the same area as the ellipse. Note that while the rectangular region 21 encompasses the raw shrimp 100 as shown in FIG. 4, the generated circular target object region 23 may not cover the entire raw shrimp 100, as shown in FIG. 6.

[0032] The target object region generating unit 1 may generate an elliptical target object region 22 or a circular target object region 23 on the image data.

[0033] The combined region generating unit 2 generates a combined region by combining two adjacent circular or elliptical target object regions generated by the target object region generating unit 1.

[0034] First, a case where the target object region generating section 1 generates an elliptical target object region 22 will be described.

[0035] Fig. 7 is a diagram illustrating a case where a synthesis region 223 is generated from an elliptical target object region 22. In Fig. 7, an elliptical target object region 221 and an elliptical target object region 222 are generated on image data. The target object region 221 and the target object region 222 are close to each other. "Close to each other" means, for example, that the distance between the center points of the respective regions 221 and 222 is within a predetermined distance.

[0036] Hereinafter, two-dimensional coordinates on the image data are represented by the x-axis and y-axis. The center point of the target object region 221 is represented by (x1, y1), and the center point of the target object region 222 is represented by (x2, y2). The major and minor axes of the elliptical target object regions 221 and 222 are formed along the x-axis and y-axis in the two-dimensional coordinate system. The diameter of the target object region 221 along the x-axis is r x1 and the diameter along the y-axis is r y1 The diameter of the target object region 222 along the x-axis is expressed as r x2 and the diameter along the y-axis is r y2 It is expressed as:

[0037] In this embodiment, synthesis region generation unit 2 synthesizes target object region 221 and target object region 222, and generates elliptical synthesis region 223 whose center coincides with target object region 221. Note that synthesis region generation unit 2 may also synthesize target object region 221 and target object region 222, and generate an elliptical synthesis region whose center coincides with target object region 222.

[0038] The combined area generation unit 2 adds the diameters of the object areas 221 and 222 along the x-axis, and also adds the diameters of the object areas 221 and 222 along the y-axis. That is, the diameter of the combined area 223 along the x-axis is r x1 +r x2 and the diameter along the y-axis is r y1 +r y2 Furthermore, the synthesis area generation unit 2 sets the center point of the synthesis area 223 to coordinates (x1, y1).

[0039] Next, a case where the target object region generating section 1 generates a circular target object region 23 will be described.

[0040] Fig. 8 is a diagram illustrating a case where a synthesis region 233 is generated from a circular target object region 23. In Fig. 8, it is assumed that a circular target object region 231 and a circular target object region 232 are generated on image data. The target object region 231 and the target object region 232 are adjacent to each other.

[0041] The center point of target object region 231 is represented by (x1, y1), and the center point of target object region 232 is represented by (x2, y2). The radius of circular target object region 231 is represented by r1. The radius of target object region 232 is represented by r2.

[0042] In this embodiment, combined region generation unit 2 combines target object region 231 and target object region 232 to generate circular combined region 233 whose center coincides with that of target object region 231. Note that combined region generation unit 2 may combine target object region 231 and target object region 232 to generate a circular combined region whose center coincides with that of target object region 232.

[0043] Synthesis area generation unit 2 adds the radii of target object areas 231 and 232. That is, the radius of synthesis area 233 is r1+r2. Furthermore, synthesis area generation unit 2 sets the center point of synthesis area 233 to coordinates (x1, y1).

[0044] The overlap determination unit 3 determines whether or not the center of one target object region is included in a combined region that has the same center as the other target object region. For example, in the case of Fig. 7, the overlap determination unit 3 determines whether or not the center (x2, y2) of the target object region 222 is included in the combined region 223. In the case of Fig. 8, the overlap determination unit 3 determines whether or not the center (x2, y2) of the target object region 232 is included in the combined region 233.

[0045] When the combined area is elliptical (as in FIG. 7), the overlap determination unit 3 determines whether the following formula is satisfied. (x2-x1) 2 / (rx1+rx2) 2 +(y2-y1) 2 / (r y 1+r y 2) 2 ≦1

[0046] If the above formula is satisfied, the overlap determination unit 3 determines that the raw shrimp 100 included in the target object region 221 and the raw shrimp 100 included in the target object region 222 overlap.

[0047] On the other hand, when the combined area is circular (as in FIG. 8), the overlap determination unit 3 determines whether the following formula is satisfied. (x2-x1) 2 +(y2-y1) 2 ≦(r1+r2) 2

[0048] If the above formula is satisfied, the overlap determination unit 3 determines that the raw shrimp 100 included in the target object region 231 and the raw shrimp 100 included in the target object region 232 overlap.

[0049] As described above, because the shape of the raw shrimp 100 is rounded, when determining whether the raw shrimp 100 overlap based on the overlap of the rectangular regions 21 (see FIG. 4 ), there may be cases where the raw shrimp 100 do not overlap even when the rectangular regions 21 overlap, as shown in FIG. 9 . FIG. 9 is a diagram for explaining a case where overlap determination is performed using the rectangular regions 21. The same result occurs when determining whether the raw shrimp 100 overlap based on the overlap of circular or elliptical target object regions without generating a composite region. Furthermore, the circular or elliptical target object region may not encompass the entire raw shrimp 100, and part of the raw shrimp 100 may protrude from the target object region. Therefore, determining whether the raw shrimp 100 overlap based solely on the overlap of the circular or elliptical target object regions may result in an error in the determination result, resulting in a decrease in determination accuracy. In particular, attempting to improve determination accuracy using elliptical target object regions requires complex computational processing.

[0050] Therefore, by generating a synthesis area and using that synthesis area to perform overlap determination, processing can be performed using simple calculations, as shown in the above formula, which improves processing speed and enables overlap determination to be performed accurately.

[0051] Next, the operation of the overlap determination device 10 in this embodiment will be described with reference to FIG. 10. FIG. 10 is a flow diagram showing the processing of the overlap determination device 10. In the following description, in this embodiment, an overlap determination method is implemented by operating the overlap determination device 10. Therefore, the description of the overlap determination method in this embodiment will be replaced by the following description of the operation of the overlap determination device 10.

[0052] The acquisition unit 4 acquires image data from the imaging device 101 (S1). The target object region generation unit 1 executes a detection process using frame images of the acquired image data and generates a rectangular region 21 (see FIG. 4) that encompasses the detected target object, a raw shrimp 100 (S2). The target object region generation unit 1 generates a circular or elliptical target object region, as shown in FIG. 5 or 6, from the generated rectangular region 21 (S3).

[0053] The synthesis area generation unit 2 generates a synthesis area shown in Fig. 7 or 8 from the generated circular or elliptical target object area (S4). The synthesis area generation method is as described with reference to Fig. 7 or 8.

[0054] The overlap determination unit 3 determines whether or not the center of one target object region is included in a combined region that has the same center as the other target object region (S5). For example, in the case of Fig. 7, the overlap determination unit 3 determines whether or not the center (x2, y2) of the target object region 222 is included in the combined region 223. In the case of Fig. 8, the overlap determination unit 3 determines whether or not the center (x2, y2) of the target object region 232 is included in the combined region 233.

[0055] The overlap determination result may be output to an external device such as a display device including a liquid crystal display, an organic electroluminescence (EL) display, a cathode ray tube (CRT), etc. Also, a light-projecting marker may be projected onto the raw shrimp 100 determined to be overlapping by a projection device such as a projector.

[0056] As described above, according to this embodiment, overlap determination is performed using simple arithmetic operations such as the above-mentioned formula, so the overlap determination processing speed can be increased and overlap between raw shrimp 100 can be determined with high accuracy.

[0057] The program in this embodiment may be any program that causes a computer to execute steps S1 to S5 shown in Fig. 10. By installing and executing this program in a computer, the overlap determination device and overlap determination method in this embodiment can be realized. In this case, the processor of the computer functions as a target object region generation unit 1, a combined region generation unit 2, an overlap determination unit 3, and an acquisition unit 4 and performs processing.

[0058] The program in this embodiment may be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer may function as one of the target object region generating unit 1, the composite region generating unit 2, the overlap determining unit 3, and the acquiring unit 4.

[0059] A computer that implements the overlap determination device by executing the program of this embodiment will now be described with reference to Fig. 11. Fig. 11 is a block diagram showing an example of a computer that implements the overlap determination device of this embodiment.

[0060] 11, the computer 110 includes a CPU 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. These components are connected to each other via a bus 121 so as to be able to communicate data with each other. Note that the computer 110 may include a GPU (Graphics Processing Unit) or an FPGA (Field-Programmable Gate Array) in addition to or instead of the CPU 111.

[0061] The CPU 111 loads the programs (codes) of this embodiment stored in the storage device 113 into the main memory 112 and executes them in a predetermined order to perform various calculations. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). The programs of this embodiment are provided in a state stored in a computer-readable recording medium 120. The programs of this embodiment may also be distributed over the Internet connected via the communication interface 117.

[0062] Specific examples of the storage device 113 include a hard disk drive and a semiconductor storage device such as a flash memory. The input interface 114 mediates data transmission between the CPU 111 and input devices 118 such as a keyboard and a mouse. The display controller 115 is connected to a display device 119 and controls the display on the display device 119.

[0063] The data reader / writer 116 mediates data transmission between the CPU 111 and the recording medium 120, reads programs from the recording medium 120, and writes processing results from the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and other computers.

[0064] Specific examples of the recording medium 120 include general-purpose semiconductor storage devices such as CF (Compact Flash (registered trademark)) and SD (Secure Digital), magnetic recording media such as flexible disks, or optical recording media such as CD-ROMs (Compact Disk Read Only Memory).

[0065] The overlap determination device 10 in this embodiment can be realized by using hardware corresponding to each part, rather than a computer on which a program is installed. Furthermore, the overlap determination device may be partially realized by a program and the remaining part by hardware.

[0066] Some or all of the above-described embodiments can be expressed by (Supplementary Note 1) to (Supplementary Note 18) described below, but are not limited to the following descriptions.

[0067] (Appendix 1) An apparatus for determining overlap between a first object and a second object, comprising: a target object region generation unit that generates a first target object region having a circular or elliptical shape corresponding to the first target object and a second target object region having a circular or elliptical shape corresponding to the second target object based on image data obtained by capturing images of the first target object and the second target object with an imaging device; a combined region generating unit that combines the first object region and the second object region to generate a combined region whose center coincides with that of the first object region; an overlap determination unit that determines that the first object and the second object overlap when the center of the second object region is included in the combined region; An overlap determination device comprising:

[0068] (Appendix 2) 10. The overlap determination device according to claim 1, The target object region generation unit generating a first rectangular region corresponding to the first target object and a second rectangular region corresponding to the second target object based on the image data, generating the first target object region based on the first rectangular region, and generating the second target object region based on the second rectangular region; Overlap determination device.

[0069] (Appendix 3) The overlap determination device according to claim 1 or 2, When the first object region and the second object region are circular, The composite area generation unit generating the circular combined region by adding the radius of the first object region and the radius of the second object region; Overlap determination device.

[0070] (Appendix 4) 4. The overlap determination device according to claim 3, In two-dimensional coordinates, when the center coordinates of the first object region are represented by (x1, y1) and the radius is represented by r1, and the center coordinates of the second object region are represented by (x2, y2) and the radius is represented by r2, The overlap determination unit (x2-x1) 2 +(y2-y1) 2 ≦(r1+r2) 2 When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination device.

[0071] (Appendix 5) The overlap determination device according to claim 1 or 2, When the first object region and the second object region have an elliptical shape whose major axis and minor axis are aligned along the horizontal and vertical directions in a two-dimensional coordinate system, The composite area generation unit generating the combined region having an elliptical shape by adding together the major axis or the minor axis of the first object region along the horizontal direction and the major axis or the minor axis of the second object region along the vertical direction; Overlap determination device.

[0072] (Appendix 6) 6. The overlap determination device according to claim 5, In two-dimensional coordinates, when the center coordinates of the first object region are represented by (x1, y1), the diameter along the horizontal direction is represented by rx1, and the diameter along the vertical direction is represented by ry1, and the center coordinates of the second object region are represented by (x2, y2), the diameter along the horizontal direction is represented by rx2, and the diameter along the vertical direction is represented by ry2, then: The overlap determination unit (x2-x1) 2 / (rx1+rx2) 2 +(y2-y1) 2 / (r y 1+r y 2) 2 ≦1 When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination device.

[0073] (Appendix 7) A method for determining overlap between a first object and a second object, comprising: generating a circular or elliptical first object region corresponding to the first object and a circular or elliptical second object region corresponding to the second object based on image data obtained by capturing images of the first object and the second object with an imaging device; combining the first object region and the second object region to generate a combined region whose center is coincident with that of the first object region; determining that the first object and the second object overlap when the center of the second object region is included in the combined region; An overlap determination method comprising:

[0074] (Appendix 8) 8. The overlap determination method according to claim 7, further comprising: generating a first rectangular region corresponding to the first target object and a second rectangular region corresponding to the second target object based on the image data; Equipped with In the step of generating the first object region and the second object region, generating the first target object region based on the first rectangular region, and generating the second target object region based on the second rectangular region; Overlap determination method.

[0075] (Appendix 9) The overlap determination method according to Supplementary Note 7 or Supplementary Note 8, When the first object region and the second object region are circular, In the step of generating a composite area, generating the circular combined region by adding the radius of the first object region and the radius of the second object region; Overlap determination method.

[0076] (Appendix 10) 10. The overlap determination method according to claim 9, further comprising: In two-dimensional coordinates, when the center coordinates of the first object region are represented by (x1, y1) and the radius is represented by r1, and the center coordinates of the second object region are represented by (x2, y2) and the radius is represented by r2, In the step of determining that the first target object and the second target object overlap, (x2-x1) 2 +(y2-y1) 2 ≦(r1+r2) 2 When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination method.

[0077] (Appendix 11) The overlap determination method according to Supplementary Note 7 or Supplementary Note 8, When the first object region and the second object region have an elliptical shape whose major axis and minor axis are aligned along the horizontal and vertical directions in a two-dimensional coordinate system, In the step of generating a composite area, generating the combined region having an elliptical shape by adding together the major axis or the minor axis of the first object region along the horizontal direction and the major axis or the minor axis of the second object region along the vertical direction; Overlap determination method.

[0078] (Appendix 12) 12. The overlap determination method according to claim 11, In two-dimensional coordinates, when the center coordinates of the first object region are represented by (x1, y1), the diameter along the horizontal direction is represented by rx1, and the diameter along the vertical direction is represented by ry1, and the center coordinates of the second object region are represented by (x2, y2), the diameter along the horizontal direction is represented by rx2, and the diameter along the vertical direction is represented by ry2, then: In the step of determining that the first target object and the second target object overlap, (x2-x1) 2 / (rx1+rx2) 2 +(y2-y1) 2 / (r y 1+r y 2) 2 ≦1 When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination method.

[0079] (Appendix 13) A program for causing a computer to determine an overlap between a first target object and a second target object, The computer, generating a circular or elliptical first object region corresponding to the first object and a circular or elliptical second object region corresponding to the second object based on image data obtained by capturing images of the first object and the second object with an imaging device; combining the first object region and the second object region to generate a combined region whose center is coincident with that of the first object region; determining that the first object and the second object overlap when the center of the second object region is included in the combined region; A program containing instructions to execute a program.

[0080] (Appendix 14) 14. The program of claim 13, The computer, generating a first rectangular region corresponding to the first target object and a second rectangular region corresponding to the second target object based on the image data; and instructions to execute In the step of generating the first object region and the second object region, generating the first target object region based on the first rectangular region, and generating the second target object region based on the second rectangular region; program.

[0081] (Appendix 15) 15. The program according to claim 13 or 14, When the first object region and the second object region are circular, In the step of generating a composite area, generating the circular combined region by adding the radius of the first object region and the radius of the second object region; program.

[0082] (Appendix 16) 16. The program of claim 15, In two-dimensional coordinates, when the center coordinates of the first object region are represented by (x1, y1) and the radius is represented by r1, and the center coordinates of the second object region are represented by (x2, y2) and the radius is represented by r2, In the step of determining that the first target object and the second target object overlap, (x2-x1) 2 +(y2-y1) 2 ≦(r1+r2) 2 When the following equation is satisfied, it is determined that the first object and the second object overlap. program.

[0083] (Appendix 17) 15. The program according to claim 13 or 14, When the first object region and the second object region have an elliptical shape whose major axis and minor axis are aligned along the horizontal and vertical directions in a two-dimensional coordinate system, In the step of generating a composite area, generating the combined region having an elliptical shape by adding together the major axis or the minor axis of the first object region along the horizontal direction and the major axis or the minor axis of the second object region along the vertical direction; program.

[0084] (Appendix 18) 18. The program of claim 17, In the two-dimensional coordinate system, the center coordinates of the first object region are (x1, y1), and the horizontal coordinate is (x1, y1). The diameter along the direction is represented by rx1, the diameter along the vertical direction is represented by ry1, and the second object region The center coordinates of (x2, y2), the diameter along the horizontal direction is rx2, and the If the diameter is expressed as ry2, In the step of determining that the first target object and the second target object overlap, (x2-x1) 2 / (rx1+rx2) 2 +(y2-y1) 2 / (r y 1+r y 2) 2 ≦1 When the following equation is satisfied, it is determined that the first object and the second object overlap. Determine, program. [Industrial Applicability]

[0085] As described above, according to the present disclosure, it is possible to accurately determine whether or not objects overlap with each other while improving processing speed. The present disclosure is useful in fields where it is necessary to determine whether or not objects, such as transported products, overlap. [Explanation of symbols]

[0086] 1: Target object region generation unit 2: Synthesis area generation section 3: Overlap detection section 4: Acquisition part 10: Overlap detection device 21: Rectangular area 22: Target object area 23: Target object area 100: Raw shrimp 101: Imaging device 102: Belt conveyor 110: Computer 111:CPU 112: Main memory 113: Storage device 114: Input interface 115: Display Controller 116: Data reader / writer 117: Communication interface 118: Input device 119: Display device 120: Recording medium 121: Bus 221: Target object area 222: Target object area 223 :Synthesis area 231: Target object area 232: Target object area 233 :Synthesis area

Claims

1. an object region generation unit that inputs image data obtained by capturing images of a first object and a second object, each having a non-uniform shape, using an imaging device into a learning model that detects the position of the object region on the image, and generates a first object region having a circular or elliptical shape corresponding to the first object, and a second object region having a circular or elliptical shape corresponding to the second object; a synthesis region generation unit that, when a distance between a center point of the first object region and a center point of the second object region is within a predetermined distance, generates a synthesis region in a circular or elliptical shape whose center point coincides with that of the first object region, by setting the sum of the diameter in the X-axis direction of the first object region and the diameter in the X-axis direction of the second object region in two-dimensional coordinates as the diameter in the X-axis direction of the synthesis region and the sum of the diameter in the Y-axis direction of the first object region and the diameter in the Y-axis direction of the second object region as the diameter in the Y-axis direction of the synthesis region; an overlap determination unit that determines that the first object and the second object overlap when the center point of the second object region is included in the combined region; An overlap determination device comprising:

2. The overlap determination device according to claim 1 , In the two-dimensional coordinate system, when the coordinates of the center point of the first object region are represented by (x1, y1) and the radius is represented by r1, and the coordinates of the center point of the second object region are represented by (x2, y2) and the radius is represented by r2, The overlap determination unit <h2 style=";text-align:left;direction:ltr">(x2-x1)<h2 style=";text-align:left;direction:ltr"> ２ <h2 style=";text-align:left;direction:ltr"> +(y2-y1)<h2 style=";text-align:left;direction:ltr"> ２ <h2 style=";text-align:left;direction:ltr"> ≦(r1+r2)<h2 style=";text-align:left;direction:ltr"> ２ When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination device.

3. The overlap determination device according to claim 1 , In the two-dimensional coordinate system, when the coordinates of the center point of the first object region are represented by (x1, y1), the diameter along the X-axis direction is represented by rx1, and the diameter along the Y-axis direction is represented by ry1, and the coordinates of the center point of the second object region are represented by (x2, y2), the diameter along the X-axis direction is represented by rx2, and the diameter along the Y-axis direction is represented by ry2, The overlap determination unit (x2-x1) ２ / (rx1+rx2) ２ +(y2-y1) ２ / (ry1+ry2) ２ ≦1 When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination device.

4. A computer comprising: inputting image data obtained by capturing images of a first target object and a second target object, each having a non-uniform shape, using an imaging device into a learning model that detects the position of the target object region on the image, and generating a circular or elliptical first target object region corresponding to the first target object, and a circular or elliptical second target object region corresponding to the second target object; when the distance between the center point of the first object region and the center point of the second object region is within a predetermined distance, generating a circular or elliptical region whose center point coincides with that of the first object region, by setting the value obtained by adding the diameter of the first object region in the X-axis direction to the diameter of the second object region in the X-axis direction in two-dimensional coordinates as the diameter of the combined region in the X-axis direction, and setting the value obtained by adding the diameter of the first object region in the Y-axis direction to the diameter of the second object region in the Y-axis direction as the diameter of the combined region in the Y-axis direction; determining that the first object and the second object overlap when the center point of the second object region is included in the combined region; An overlap determination method that performs the following.

5. The overlap determination method according to claim 4, In the two-dimensional coordinate system, when the coordinates of the center point of the first object region are represented by (x1, y1) and the radius is represented by r1, and the coordinates of the center of the second object region are represented by (x2, y2) and the radius is represented by r2, In the step of determining that the first target object and the second target object overlap, <h2 style=";text-align:left;direction:ltr">(x2-x1)<h2 style=";text-align:left;direction:ltr"> ２ <h2 style=";text-align:left;direction:ltr"> +(y2-y1)<h2 style=";text-align:left;direction:ltr"> ２ <h2 style=";text-align:left;direction:ltr"> ≦(r1+r2)<h2 style=";text-align:left;direction:ltr"> ２ When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination method.

6. The overlap determination method according to claim 4, In the two-dimensional coordinate system, when the coordinates of the center point of the first object region are represented by (x1, y1), the diameter along the X-axis direction is represented by rx1, and the diameter along the Y-axis direction is represented by ry1, and the coordinates of the center point of the second object region are represented by (x2, y2), the diameter along the X-axis direction is represented by rx2, and the diameter along the Y-axis direction is represented by ry2, In the step of determining that the first target object and the second target object overlap, (x2-x1) ２ / (rx1+rx2) ２ +(y2-y1) ２ / (ry1+ry2) ２ ≦1 When the following equation is satisfied, it is determined that the first object and the second object overlap. Overlap determination method.

7. A computer, inputting image data obtained by capturing images of a first target object and a second target object, each having a non-uniform shape, using an imaging device into a learning model that detects the position of the target object region on the image, and generating a circular or elliptical first target object region corresponding to the first target object, and a circular or elliptical second target object region corresponding to the second target object; when the distance between the center point of the first object region and the center point of the second object region is within a predetermined distance, generating a circular or elliptical region whose center point coincides with that of the first object region, by setting the value obtained by adding the diameter of the first object region in the X-axis direction to the diameter of the second object region in the X-axis direction in two-dimensional coordinates as the diameter of the combined region in the X-axis direction, and setting the value obtained by adding the diameter of the first object region in the Y-axis direction to the diameter of the second object region in the Y-axis direction as the diameter of the combined region in the Y-axis direction; determining that the first object and the second object overlap when the center point of the second object region is included in the combined region; A program containing instructions to execute a program.

8. 8. The program according to claim 7, In the two-dimensional coordinate system, when the coordinates of the center point of the first object region are represented by (x1, y1) and the radius is represented by r1, and the coordinates of the center of the second object region are represented by (x2, y2) and the radius is represented by r2, In the step of determining that the first target object and the second target object overlap, <h2 style=";text-align:left;direction:ltr">(x2-x1)<h2 style=";text-align:left;direction:ltr"> ２ <h2 style=";text-align:left;direction:ltr"> +(y2-y1)<h2 style=";text-align:left;direction:ltr"> ２ <h2 style=";text-align:left;direction:ltr"> ≦(r1+r2)<h2 style=";text-align:left;direction:ltr"> ２ When the following equation is satisfied, it is determined that the first object and the second object overlap. program.

9. 8. The program according to claim 7, In the two-dimensional coordinate system, when the coordinates of the center point of the first object region are represented by (x1, y1), the diameter along the X-axis direction is represented by rx1, and the diameter along the Y-axis direction is represented by ry1, and the coordinates of the center point of the second object region are represented by (x2, y2), the diameter along the X-axis direction is represented by rx2, and the diameter along the Y-axis direction is represented by ry2, In the step of determining that the first target object and the second target object overlap, (x2-x1) ２ / (rx1+rx2) ２ +(y2-y1) ２ / (ry1+ry2) ２ ≦1 When the following equation is satisfied, it is determined that the first object and the second object overlap. program.

Images