Article detection device, article detection method, and program

Abstract

To improve detection accuracy of articles to be labeled.SOLUTION: An article detection device is provided, comprising an image acquisition unit for acquiring multiple images of an article while changing the relative position with respect to the article, an image processing unit configured to generate a composite image using the multiple images acquired by the image acquisition unit, and an article region estimation unit for estimating a region occupied by the article in the composite image generated by the image processing unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an article detection device, an article detection method, and a program. [Background technology]

[0002] Patent Document 1 discloses a labeling device that detects an article being conveyed in a conveyance direction and affixes a label to a predetermined position on the article while conveying the article.

[0003] In the labeling system disclosed in Patent Document 1, image data of the item to be adhered is acquired in a detection unit capable of capturing an image of the item, and a label can be attached to a predetermined position on the item based on the acquired image data. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2023-049479 Summary of the Invention [Problem to be solved by the invention]

[0005] In the labeling system disclosed in Patent Document 1, if it is difficult to detect the boundary between the item and the conveying section that is the background of the item in the acquired image data, it is not possible to identify the labeling position on the item, which can result in a labeling error. For this reason, there is room for further improvement in the labeling system.

[0006] Therefore, an object of the present invention is to improve the accuracy of detecting an article to which a label is to be attached. [Means for solving the problem]

[0007] According to one aspect of the present invention, an item detection device is provided that includes an image acquisition unit that acquires multiple images of an item by changing the relative position of the item with respect to the item, an image processing unit that generates a composite image using the multiple images acquired by the image acquisition unit, and an item area estimation unit that estimates the area in which the item exists in the composite image generated by the image processing unit. [Effects of the Invention]

[0008] According to one aspect of the present invention, the area where an item exists is estimated in a composite image generated using multiple images acquired at different relative positions, thereby improving the accuracy of item detection. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a schematic diagram illustrating an example of the configuration of an object detection device according to this embodiment. [Figure 2] FIG. 2 is a block diagram illustrating the object detection device according to this embodiment. [Figure 3] FIG. 3 is a flowchart illustrating the article detection method according to this embodiment. [Figure 4] FIG. 4 is a schematic diagram for explaining the process of extracting a characteristic region in the article detection method according to this embodiment. [Figure 5] FIG. 5 is a conceptual diagram illustrating the concept of generating a composite image from four feature regions extracted from an image group. [Figure 6] FIG. 6 is a schematic diagram illustrating a composite image obtained by the synthesis process. [Figure 7] FIG. 7 is a schematic diagram illustrating an example of a reference shape that is set to determine an area where an article exists. [Figure 8] FIG. 8 is a schematic diagram illustrating the labeling system according to this embodiment. [Figure 9] FIG. 9 is a schematic diagram illustrating a label attached to an article by the label attachment system. DETAILED DESCRIPTION OF THE INVENTION

[0010] The embodiments described below are not limited to the drawings described by the brief description of the drawings.

[0011] A first aspect of the present invention is an item detection device comprising an image acquisition unit that acquires multiple images of an item by changing the relative position of the item with respect to the item, an image processing unit that generates a composite image using the multiple images acquired by the image acquisition unit, and an item area estimation unit that estimates the area in which the item exists in the composite image generated by the image processing unit.

[0012] According to a first aspect of the present invention, an image acquisition unit acquires multiple images of an item by changing the relative position of the item, an image processing unit generates a composite image using the multiple images acquired by the image acquisition unit, and an item area estimation unit estimates the area in which the item exists using the composite image generated by the image processing unit.

[0013] Therefore, according to a first aspect of the present invention, the area where an item exists is estimated using a composite image generated using multiple images acquired at different relative positions, thereby improving the accuracy of item detection.

[0014] A second aspect of the present invention is an object detection device according to the first aspect, wherein the object area estimation unit estimates the composite image as an area where the object exists when the composite image of an area having characteristic pixel information in each of the plurality of images is inscribed within a predetermined reference shape.

[0015] According to a second aspect of the present invention, when a composite image of an area having characteristic pixel information in each of a plurality of images is inscribed in a predetermined reference shape, the object area estimation unit estimates the composite image as an area where an object exists, thereby improving the accuracy of object detection.

[0016] Furthermore, a third aspect of the present invention is an object detection device according to the first or second aspect, wherein at least a portion of the object is the same color as the background of the image of the object acquired by the image acquisition unit.

[0017] According to a third aspect of the present invention, a composite image is generated using multiple images acquired at different relative positions, and the area in which the object exists is estimated in the generated composite image, thereby improving the accuracy of object detection even if at least a portion of the object is the same color as the background of the object.

[0018] A fourth aspect of the present invention is the object detection device according to any one of the first to third aspects, wherein at least a portion of the object is transparent.

[0019] According to a fourth aspect of the present invention, a composite image is generated using multiple images acquired at different relative positions, and the area in which the object is present is estimated in the generated composite image, thereby improving the accuracy of detecting the object even if at least a portion of the object is transparent.

[0020] Furthermore, a fifth aspect of the present invention is an object detection device described in any one of the first to fourth aspects, wherein the image acquisition unit acquires an image of the object being transported in a predetermined transport direction.

[0021] According to a fifth aspect of the present invention, the image acquisition unit acquires images of an item being transported in a predetermined transport direction, so that even if the image acquisition unit is fixed to the item detection device, multiple images can be easily acquired with different relative positions to the item.

[0022] In addition, a sixth aspect of the present invention is an item detection device described in any one of the first to fifth aspects, wherein the image acquisition unit includes a first image acquisition unit that acquires a planar image of the item and a second image acquisition unit that acquires height information of the item, and the item area estimation unit estimates the area in which the item exists based on the area estimated based on the composite image generated using the multiple images acquired by the first image acquisition unit and the height information acquired by the second image acquisition unit.

[0023] According to a sixth aspect of the present invention, the image acquisition unit acquires a planar image of the item using a first image acquisition unit, acquires height information of the item using a second image acquisition unit, and the item area estimation unit estimates the area in which the item exists based on the area estimated based on a composite image generated using multiple images acquired by the first image acquisition unit and the height information acquired by the second image acquisition unit.

[0024] Therefore, according to the sixth aspect of the present invention, it is possible to improve the accuracy of detecting an article.

[0025] Furthermore, a seventh aspect of the present invention is an item detection device described in any one of the first to sixth aspects, wherein the image processing unit generates a first composite image using the plurality of images acquired from the image acquisition unit, and when an area similar to the predetermined outer shape of the item is not obtained in the first composite image, the image processing unit generates a second composite image using the next plurality of images acquired following the plurality of images.

[0026] According to a seventh aspect of the present invention, when the object area estimation unit cannot estimate the area where the object exists from the first composite image generated by the image processing unit, it estimates the area where the object exists in the second composite image generated by the image processing unit.

[0027] Therefore, according to the seventh aspect of the present invention, it is possible to improve the accuracy of detecting an article.

[0028] An eighth aspect of the present invention is an item detection method that acquires multiple images of an item by changing the relative position of the item with respect to the item, generates a composite image using the acquired multiple images, and estimates an area in the generated composite image in which the item exists.

[0029] According to an eighth aspect of the present invention, multiple images of an item are acquired by changing the relative position of the item, a composite image is generated using the multiple acquired images, and an area in which the item exists is estimated in the generated composite image.

[0030] Therefore, according to an eighth aspect of the present invention, the area in which an item exists is estimated in a composite image generated using multiple images acquired at different relative positions, thereby improving the accuracy of item detection.

[0031] [Item detection device] An object detection device 1 according to an embodiment of the present invention will be described.

[0032] Fig. 1 is a schematic diagram illustrating an article detection device 1 according to this embodiment. Fig. 2 is a block diagram illustrating the article detection device 1.

[0033] In this embodiment, the object M to be detected by the object detection device 1 is a food product or a part packaged in a packaging material. Therefore, the outer shape of the object M is the outer shape of the packaging material that covers the food product or part.

[0034] The item detection device 1 according to this embodiment can be suitably used particularly when at least a portion of the packaging material is transparent or when at least a portion of the packaging material is the same color as the conveying surface.

[0035] As shown in FIG. 1, the item detection device 1 includes an image acquisition unit 10 that acquires an image of an item M, and a controller 15 that executes specific image processing on the image data.

[0036] In this embodiment, an image not only means a visible image displayed on a display device such as a monitor, but also means data used for display or storage in a digital device such as a computer or monitor.

[0037] In this embodiment, the expression "at least a portion of the packaging material is transparent" means that in the image acquired by the image acquisition unit 10, it is difficult to distinguish between the packaging material covering the item M and the background image both visually and in terms of pixel information. The same applies when at least a portion of the packaging material is the same color as the conveyance surface.

[0038] In this embodiment, a conveying device 20 is also provided, and the items M are conveyed by the conveying device 20 in a conveying direction F in Fig. 1. At the destination of the conveying device 20 (not shown in Fig. 1), for example, a sorting device that sorts the items M or a labeling system that affixes labels to the items M can be disposed.

[0039] In this embodiment, the image acquisition unit 10 is a camera having a function of continuously acquiring images of the item M at predetermined intervals. The image acquisition unit 10 is provided in the conveying device 20 perpendicular to a loading surface (conveying surface 20a) on which the item M is placed, so as to face the conveying surface 20a.

[0040] In this embodiment, the image acquisition unit 10 is set so that an imaging range S including the conveying surface 20a and the article M placed on the conveying surface 20a and conveyed falls within the angle of view AF (see FIG. 4).

[0041] As shown in FIG. 2, the image acquisition unit 10 includes a 2D camera 11 as a first image acquisition unit and a 3D camera 12 as a second image acquisition unit.

[0042] As shown in FIG. 1, the 2D camera 11 captures an image on a plane extending in the conveyance direction F (X direction) of the article M and in a direction intersecting the F direction (Y direction).

[0043] The 3D camera 12 can acquire depth information (height information) of the item M in the Z direction in addition to the XY plane.

[0044] In this embodiment, the article M is conveyed at a predetermined speed by the conveying device 20. Therefore, the image acquisition unit 10 continuously acquires images of the article M being conveyed at the predetermined speed at predetermined intervals, thereby making it possible to acquire multiple images of the article M at different relative positions between the article M and the image acquisition unit 10 without changing the position of the image acquisition unit 10.

[0045] Although not shown, the controller 15 is a computer that includes a microprocessor, a storage unit such as a ROM (Read Only Memory) and a RAM (Random Access Memory), an input / output interface, and a bus that connects these.

[0046] As shown in FIG. 2, the controller 15 includes, as functional components, an image processing unit 16 that generates a composite image, and an article region estimation unit 17 that estimates the region where the article M exists from the composite image.

[0047] The controller 15 causes the image processing unit 16 and the article area estimation unit 17 to operate by executing a control program stored in a storage unit such as a ROM in a microprocessor.

[0048] The image processing unit 16 can execute processes such as extracting a region having characteristic pixel information from an image obtained from the image acquisition unit 10, and generating a composite image using multiple images acquired by the image acquisition unit 10. In this embodiment, a region having characteristic pixel information will be referred to as a characteristic region hereinafter.

[0049] In this embodiment, background subtraction processing can be applied as an example of processing for extracting characteristic regions.

[0050] When background subtraction processing is applied, the item detection device 1 stores in a storage unit an image of the conveying surface of the conveying device 20 (hereinafter referred to as a background image) that has been acquired in advance.

[0051] The controller 15 compares the background image with the image acquired by the image acquisition unit 10, and is thereby able to extract, as a characteristic region, an image that does not exist in the background image.

[0052] The image acquired by the image acquisition unit 10 is a collection of pixels having RGB digital information. In this embodiment, as an example, the luminance value obtained from the information possessed by the pixels and the information on the luminance gradation value can be used in the background subtraction process.

[0053] The image processing unit 16 can handle the multiple images (P1 to P4) acquired from the image acquisition unit 10 as a group of images G (see FIG. 4) and apply the group to image processing.

[0054] As an example, the image processing unit 16 generates a composite image VA using multiple images (P1 to P4) included in one image group G (see FIGS. 4 to 6). When the item area estimation unit 17 cannot estimate the area where the item M exists from the composite image VA, the image processing unit 16 creates a second composite image using another image group consisting of multiple images captured consecutively at a time later than the time when the image group G was acquired.

[0055] As an example, if the 2D camera 11 can capture 10 images in 0.3 seconds, multiple images can be acquired while the item M is being conveyed across the imaging range S. Furthermore, the moving item M can be photographed from various angles. In this way, the imaging interval can be adjusted according to the conveying speed of the conveying device 20 so that a sufficient number of images necessary for the image processing performed by the image processing unit 16 can be acquired.

[0056] The article region estimation unit 17 performs processing to estimate the region in which the article M exists in the composite image generated by the image processing unit 16.

[0057] When the composite image of the characteristic regions generated by the image processing unit 16 is inscribed in a preset reference shape, the article region estimation unit 17 estimates the composite image as a region where the article M exists.

[0058] In this embodiment, the reference shape is a shape corresponding to the contour of the article M. The reference shape is a rectangular shape that is set in advance in the article detection device 1 according to the outer shape of the article M as the detection target (see FIG. 7).

[0059] If the item area estimation unit 17 cannot estimate the area where the item M exists in the composite image generated by the image processing unit 16, i.e., if the generated composite image does not match the reference shape, it can estimate the area where the item M exists using a second composite image created by the image processing unit 16.

[0060] In addition, the item area estimation unit 17 can perform processing to estimate the area where the item M is located based on an area estimated based on a composite image generated using multiple images acquired by the 2D camera 11 and height information acquired by the 3D camera 12.

[0061] [Item detection method] Next, a method for detecting the article M by the article detection device 1 will be described.

[0062] Fig. 3 is a flowchart illustrating an article detection method by the article detection device 1. In Fig. 3, image processing of an image acquired by the 2D camera 11 in the image acquisition unit 10 will be described.

[0063] In step S1, the controller 15 acquires a plurality of images of the object M being conveyed by the conveying device 20 from the 2D camera 11 in the image acquisition unit 10.

[0064] In step S2, the controller 15 causes the image processing unit 16 to execute a process of extracting an area having characteristic pixel information from each of the multiple images acquired by the 2D camera 11.

[0065] In this embodiment, as an example, background subtraction processing is performed using brightness values ​​obtained from information held by pixels in the acquired image and brightness gradation value information.

[0066] FIG. 4 is a schematic diagram for explaining the process of extracting a characteristic region in the article detection method.

[0067] 4 shows an image group G that is a collection of multiple images P1, P2, P3, and P4 that were successively captured by the 2D camera 11. Image P1 is an image captured at time t1, image P2 is an image captured at time t2, image P3 is an image captured at time t3, and image P4 is an image captured at time t2.

[0068] The controller 15 extracts, as feature regions, regions made up of pixels having a brightness value exceeding a predetermined value for each of the acquired images P1, P2, P3, and P4.

[0069] The rectangle in image P1 indicates the angle of view AF of the shooting range S shown in Fig. 1. The hatched portion in image P1 indicates the conveying surface 20a and a portion of the article M that is transmitted through the conveying surface 20a, or indicates the conveying surface 20a and a portion of the article M that is the same color as the conveying surface 20a.

[0070] The white portions in the image P1 represent pixels having a brightness value exceeding a predetermined value. In this embodiment, a collection of white portions in the image P1 is represented as a characteristic region V1.

[0071] The controller 15 similarly extracts feature regions V2, V3, and V4 from the acquired images P2, P3, and P4, respectively.

[0072] Since the article M is being conveyed by the conveying device 20, as shown in FIG. 4, the positions of the characteristic regions V1 to V4 in the angle of view AF are different in each of the images P1 to P4 taken at a predetermined interval.

[0073] Next, in step S3, the controller 15 generates a composite image of the characteristic regions V1 to V4 extracted from the image group G.

[0074] FIG. 5 is a conceptual diagram illustrating the concept of generating a composite image from four feature regions V1 to V4 extracted from an image group G, and FIG. 6 is a schematic diagram illustrating the composite image obtained by the synthesis process.

[0075] The controller 15 selects a luminance value with a strong characteristic in the characteristic region V1 and sets the reference coordinates as shown in Fig. 5. In Fig. 5, (X11, Y11), (X12, Y12), and (X13, Y13) are shown as the reference coordinates.

[0076] The controller 15 sets reference coordinates (X21, Y21), (X22, Y22), and (X23, Y23) for feature region V2 that correspond to the reference coordinates of V1, sets reference coordinates (X32, Y32) and (X33, Y33) for feature region V3 that correspond to the reference coordinates of V1, and sets reference coordinates (X41, Y41), (X42, Y42), and (X43, Y43) for feature region V4 that correspond to the reference coordinates of V1.

[0077] The controller 15 superimposes the plurality of feature regions V1 to V4 based on the above-mentioned reference coordinates to generate a composite image VA (see FIG. 6) of the feature regions V1, V2, V3, and V4.

[0078] Next, in step S4, the controller 15 causes the article area estimation unit 17 to execute a process of estimating an area where the article M exists. That is, the controller 15 determines whether the composite image VA fits into a preset reference shape B.

[0079] FIG. 7 is a schematic diagram illustrating an example of a reference shape B that is set to determine an area where an article exists.

[0080] If the controller 15 determines that the composite image VA fits the reference shape B (step S4: YES), then in step S5, the controller 15 determines the area surrounded by the reference shape B in which the composite image VA is inscribed as the area where the article M exists.

[0081] If the controller 15 determines that the composite image VA does not fit into the reference shape B (step S4: NO), the controller 15 repeats the process from step S1.

[0082] The controller 15 executes the image processing from step S2 onwards using another image group consisting of a plurality of images captured consecutively at a time later than image group G.

[0083] The controller 15 can estimate the area where the article M exists from the image acquired by the 2D camera 11 by executing the series of image processing described in steps S1 to S5 above.

[0084] 4 also for the image acquired by the 3D camera 12 in the image acquisition unit 10. The controller 15 performs processing to estimate the area where the item M exists based on the area estimated based on a composite image generated using multiple images acquired by the 2D camera 11 and the height information acquired by the 3D camera 12.

[0085] [Effects of the item detection device and item detection method] According to the item detection device 1 and the item detection method executed by the item detection device 1 of the embodiment of the present invention, the image acquisition unit 10 acquires multiple images of the item by changing the relative position with respect to the item M, and the image processing unit 16 generates a composite image VA using the multiple images acquired by the image acquisition unit 10, and the area where the item M exists is estimated using the composite image VA.

[0086] Therefore, the article detection device 1 can improve the accuracy of detecting an article.

[0087] When the item area estimation unit 17 determines that the composite image VA of the feature areas V1, V2, V3, and V4 in each of the multiple images is inscribed in the reference shape B, the item detection device 1 can determine that the area surrounded by the reference shape B inscribed in the composite image VA is the area where the item M exists.

[0088] Therefore, the article detection device 1 can improve the accuracy of detecting an article.

[0089] Furthermore, according to the item detection device 1, the area where the item M exists is estimated in a composite image VA generated using multiple images P1 to P4 acquired at different relative positions, thereby improving the accuracy of item detection even when at least a portion of the item M is the same color as the background of the item M, or when at least a portion of the item M is transparent and the background color is detected inside the outer shape of the item M.

[0090] Furthermore, according to the article detection device 1, the image acquisition unit 10 acquires images of the article M being conveyed in the conveying direction F by the conveying device 20, so even if the image acquisition unit 10 is fixed to the article detection device 1, it can easily acquire multiple images with different relative positions to the article M. This makes it possible to simplify the device configuration of the article detection device 1.

[0091] Furthermore, according to the item detection device 1, the image acquisition unit 10 acquires a planar image of the item M using the 2D camera 11, and acquires height information of the item M using the 3D camera 12. The item area estimation unit 17 estimates the area where the item M exists based on the area estimated from a composite image based on the image acquired by the 2D camera 11 and the height information acquired by the 3D camera 12, thereby improving the detection accuracy of the item M.

[0092] Furthermore, according to the item detection device 1, the multiple images (P1 to P4) acquired from the image acquisition unit 10 are treated as a single image group G (see FIG. 4), and when the area where the item M exists cannot be estimated using a composite image VA obtained by image processing from this image group G, the area where the item M exists can be estimated using a composite image obtained by image processing from the next image group consisting of multiple images. Therefore, the item detection device 1 can improve the detection accuracy of the item M.

[0093] [program] The program of this embodiment is a program that can be executed by the computer of the item detection device 1, and causes the computer to execute the following steps: acquiring multiple images of item M by changing the relative position of item M; generating a composite image using the acquired multiple images; and estimating the area in which item M exists in the generated composite image.

[0094] The program is stored in a storage unit that can be used by the controller 15 of the item detection device 1. Alternatively, the program is stored in a storage medium that is detachable from the item detection device 1 and provided.

[0095] [Labeling system] As an example, the item detection device according to this embodiment can be combined with a labeling system that attaches labels to items to form a labeling system.

[0096] FIG. 8 is a schematic diagram illustrating a labeling system 100 according to this embodiment.

[0097] The label sticking system 100 according to this embodiment is used in combination with the article detection device 1 shown in FIGS. 1 and 2, and is a system that sticks a label LB to an article M detected by the article detection device 1.

[0098] The label application system 100 comprises an item detection device 1, a conveying device 20, an application arm 30, a printer 40, a label conveying unit 50, and a downstream detection unit 60, and each of the above components is controlled by a controller 70.

[0099] The article detection device 1 and the transport device 20 shown in FIG. 8 are the same as the article detection device 1 and the transport device 20 shown in FIG.

[0100] The application arm 30 includes a label holding portion 31. The tip of the label holding portion 31 is provided with a structure for sucking and holding the label LB.

[0101] The application arm 30 is provided with an actuator that allows movement in the vertical, horizontal and height directions and angle adjustment around an axis.

[0102] This allows the application arm 30 to pick up and hold the label LB conveyed by the label conveying unit 50 with the adhesive surface facing the article M. The application arm 30 can also move the label LB to the article M being conveyed and apply the label LB to a predetermined position on the article M.

[0103] The printer 40 comprises a printer body 41 for printing necessary information on a label LB, a label roll 42 in which the pre-printed label LB is temporarily attached to a continuous body of separator, and a backing roll 43 in which the continuous body of separator is collected after the label LB has been peeled off.

[0104] In this embodiment, the printer 40 is a printer that can independently set and print the print content related to the item M without being connected to an information processing terminal such as a personal computer.

[0105] The labeling system 100 is equipped with a printer 40, which allows information about the item M to be printed on the label LB before it is attached. For example, the information about the item M may include, in the case of a food product, the date of manufacture, expiration date, names of ingredients, etc.

[0106] The label conveying unit 50 includes a conveying belt 51 that conveys the labels LB peeled off from the separator continuum, and conveys the plurality of labels LB individually and continuously.

[0107] The downstream detection unit 60 is disposed downstream of the application arm 30 in the conveying device 20. In this embodiment, the downstream detection unit 60 is a camera that captures an image of the item M after the label LB has been applied. The image data acquired by the downstream detection unit 60 is used in the controller 70 in a process of determining the application state of the label LB.

[0108] The controller 70 is a computer that includes a microprocessor, storage units such as ROM and RAM, an input / output interface, and a bus that connects these.

[0109] In this embodiment, the controller 70 can determine the position where the label LB is to be attached to the item M based on the detection result of the item M obtained from the item detection device 1.

[0110] FIG. 9 is a schematic diagram illustrating a label LB that is attached to an item M by the label attachment system 100. As shown in FIG.

[0111] 9 shows states (a), (b), and (c). State (a) is a state in which the label LB is being conveyed by the conveyor belt 51. State (b) is a state in which the label LB is being picked up by the application arm 30 and is being conveyed to the article M while being rotated so that it is oriented in a predetermined direction. State (c) is a state in which the label LB is being applied to the article M.

[0112] As shown in FIG. 9, the item detection device 1 can accurately detect the item M even if the item M is packaged in a transparent packaging material or a packaging material of the same color as the conveyance surface 20a (background image).

[0113] Therefore, by acquiring data related to the item M from the item detection device 1, the labeling system 100 can attach the label LB to the item M at a position appropriate to the outer shape of the item M and in an appropriate orientation.

[0114] For example, as shown in Figure 9, the label LB can be rotated so that one side of the outer shape of the item M is parallel to one side of the label LB, and then the label LB can be moved toward the item M, and the label LB can be attached to the item M.

[0115] Therefore, the aesthetic appearance of the article M and the label LB attached to the article M can be improved.

[0116] [Other embodiments] Although the embodiments of the present invention have been described above, the above embodiments merely illustrate some of the application examples of the present invention, and are not intended to limit the technical scope of the present invention to the specific configurations of the above embodiments.

[0117] In this embodiment, the article M is being transported and the image acquisition unit 10 is fixed. In the present invention, the image acquisition unit 10 may be moving as long as it can capture multiple images of the article M by changing the relative position.

[0118] In this embodiment, instead of the 3D camera 12, a sensor may be applied that irradiates the object M with detection light such as infrared light or laser light and acquires height information based on the reflection time.

[0119] The synthesis process for generating a composite image from multiple feature regions can also apply image processing other than background subtraction. Applicable image processing in this embodiment includes an algorithm for detecting objects of known shapes from acquired images. For example, there is a method for obtaining bounding boxes (external shapes) for multiple images using R-CNN and YOLO.

[0120] Another method is to use image processing such as optical flow or known conveyance speed information to superimpose multiple images that are presumed to be the item M, and find the area in the image where the item M exists.

[0121] As a method of using the conveying speed information, the controller 15 may acquire the conveying speed of the conveying device 20 that conveys the item M, correct the amount of deviation of the feature region V2 in the image P2 from the display position of the feature region V1 in the image P1 based on the conveying speed, and perform processing to combine the feature regions V1 and V2. Similar processing is also performed for the feature regions V3 and V4. [Explanation of symbols]

[0122] 1. Item detection device 10 Image acquisition unit 11 2D Camera 12. 3D Camera 15 Controller 16 Image processing section 17 Goods area estimation section 20. Conveyor 20a Conveying surface 30 Adhesive Arm 31 Label holder 40 Printers 41 Printer body 42 label rolls 43 Mounting Paper Roll 50 Label transport unit 51 Conveyor belt 60 Lower detection section 70 Controller 100 Labeling System AF angle of view B Standard shape G Image Group LB Label M Goods P1, P2, P3, P4 images S Shooting range V1,V2,V3,V4 feature region

Claims

1. an image acquisition unit that acquires a plurality of images of an article while changing a relative position with respect to the article; an image processing unit that generates a composite image using the plurality of images acquired by the image acquisition unit; an article region estimation unit that estimates a region where the article exists using the composite image generated by the image processing unit; Equipped with Item detection device.

2. The object detection device according to claim 1 , the object region estimation unit estimates the composite image as a region where the object exists when the composite image of a region having characteristic pixel information in each of the plurality of images is inscribed in a predetermined reference shape; Item detection device.

3. The object detection device according to claim 1 , At least a portion of the article has the same color as the background of the image of the article acquired by the image acquisition unit. Item detection device.

4. The object detection device according to claim 1 , At least a portion of the article is transparent; Item detection device.

5. The object detection device according to claim 1 , the image acquisition unit acquires an image of the item being conveyed in a predetermined conveyance direction. Item detection device.

6. The object detection device according to claim 1 , The image acquisition unit a first image acquisition unit that acquires a planar image of the article; a second image acquisition unit that acquires height information of the article, The object region estimation unit an area where the item exists is estimated based on the area estimated based on the composite image generated using the plurality of images acquired by the first image acquisition unit and the height information acquired by the second image acquisition unit; Item detection device.

7. The object detection device according to claim 1 , the image processing unit generates a first composite image using the plurality of images acquired from the image acquisition unit; When the object region estimation unit cannot estimate the region where the object exists from the first composite image, the image processing unit generates a second composite image using a plurality of subsequent images acquired subsequent to the plurality of images; Item detection device.

8. acquiring a plurality of images of an article at different positions relative to the article; generating a composite image using the acquired images; Estimating an area in which the item is present in the generated composite image; Item detection methods.

9. A program executable by a computer of an item detection device, acquiring a plurality of images of an item at different positions relative to the item; generating a composite image using the plurality of acquired images; and a procedure for estimating an area in which the item exists in the generated composite image.

Images