Image input device and commodity recognition system

The image input device with integrated illumination and imaging capabilities, along with a recognition model creation device, addresses the challenge of clear product imaging for reliable recognition model generation and improved product identification.

JP2024169015AInactive Publication Date: 2024-12-05TOSHIBA TEC KK
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
JP2023086183
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-25
Publication Date
2024-12-05
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Conventional image input devices lack the capability to clearly image products for reliable product recognition, necessitating improved systems for generating recognition models.

Method used

An image input device comprising an imaging section, illumination section, and image output section, which captures images of products at a short distance and illuminates the imaging range, coupled with a recognition model creation device to generate and utilize recognition models for product recognition.

Benefits of technology

Enables easy and reliable generation of recognition models for accurate product recognition, enhancing the precision of product identification processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024169015000001_ABST
    Figure 2024169015000001_ABST
Patent Text Reader

Abstract

To provide an image input device and a commodity recognition system capable of easily and reliably generating a recognition model to be used when recognizing a commodity.SOLUTION: A code scanner (an image input device) includes: a camera (an imaging unit) that images a region at a short distance from itself to acquire an image of a commodity held over itself; a short-distance illumination (an illumination unit) that illuminates an imaging range of the camera; and an image transmission unit (an image output unit) that outputs an image captured by the camera to a recognition model generation device that generates a recognition model for recognizing the commodity.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] An embodiment of the present invention relates to an image input device and a product recognition system. [Background technology]

[0002] Conventionally, a processing device has been proposed that creates a recognition model for recognizing an object by detecting the object and a barcode affixed to the object from an image captured by a camera (for example, Patent Document 1).

[0003] Such a conventional processing device has a configuration for capturing an image of a commodity placed on a dedicated placement stand, but no specific details of the imaging device are mentioned. It has been desired that the image input device used for capturing an image of a commodity has a function for clearly capturing an image of only the commodity to be recognized. Summary of the Invention [Problem to be solved by the invention]

[0004] An object of the present invention is to provide an image input device and a product recognition system that can easily and reliably generate a recognition model to be used when recognizing products. [Means for solving the problem]

[0005] The image input device of the embodiment includes an imaging unit, an illumination unit, and an image output unit. The imaging unit captures images of close-ups and captures an image of a commodity held up to the imaging unit. The illumination unit illuminates the imaging range of the imaging unit. The image output unit outputs the image captured by the imaging unit to a recognition model creation device that generates a recognition model for recognizing the commodity.

[0006] In addition, the product recognition system of the embodiment includes an image input device, a recognition model creation device that uses an image of the product input by the image input device to create a recognition model for recognizing the product, and a product recognition device that recognizes the product using the image of the product input from the image input device and the recognition model generated by the recognition model creation device. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a functional block diagram showing an example of a functional configuration of a commodity recognition system using an image input device according to an embodiment. [Diagram 2] FIG. 2 is a diagram showing a product being held up to a code scanner. [Diagram 3] FIG. 3 is a timing chart showing an example of the flow of imaging processing performed by the camera control unit. [Figure 4] FIG. 4 is a diagram showing an example of a recognition model created by the recognition model creation device. [Diagram 5] FIG. 5 is a flowchart showing an example of the flow of processes performed by the product recognition system. [Figure 6] FIG. 6 is a diagram illustrating an example of the flow of a recognition model creation process performed by the recognition model creation device according to the embodiment. [Figure 7] FIG. 7 is a diagram illustrating an example of a flow of a product registration process performed by the POS terminal of the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiment described below.

[0009] A product recognition system 100 including a code scanner 110 according to an embodiment of the present invention will be described.

[0010] First, the functional configuration of a commodity recognition system 100 according to an embodiment will be described with reference to Fig. 1. Fig. 1 is a functional block diagram showing an example of the functional configuration of a commodity recognition system using an image input device according to an embodiment.

[0011] The commodity recognition system 100 includes a code scanner 110, a POS terminal 120, a recognition model creation device 130, and a commodity recognition device 140.

[0012] The code scanner 110 captures an image of the commodity 90 held up to the code scanner 110. If a code symbol such as a barcode is present in the captured image of the commodity 90, the code scanner 110 reads the coded information contained in the code symbol. The code scanner 110 is an example of an image input device in the present disclosure.

[0013] The code scanner 110 includes a camera 111, close-range illuminators 112 and 113, a camera control unit 114, a barcode recognition unit 115, and an image transmission unit .

[0014] The camera 111 captures an image of an area close to itself to obtain an image of the commodity 90 held up to itself. More specifically, the camera 111 captures an image of a substantially ellipsoidal imaging range P in the vicinity of the code scanner 110. The camera 111 is, for example, a CCD camera or a CMOS camera. The camera 111 may be a monochrome camera or a color camera. The camera 111 is an example of an imaging unit in the present disclosure.

[0015] The close-distance illuminators 112 and 113 are installed near the camera 111 and illuminate illumination ranges Q and R including the imaging range P of the camera 111. The close-distance illuminators 112 and 113 are, for example, LEDs. Note that while Fig. 1 shows an example in which two close-distance illuminators 112 and 113 are installed, the number of close-distance illuminators is not important. The close-distance illuminators 112 and 113 are an example of an illumination unit in this disclosure.

[0016] The camera control unit 114 controls the timing of imaging by the camera 111 and the timing of lighting the close-range illuminators 112 and 113. Specifically, the camera control unit 114 causes the camera 111 to capture an image. At that time, the camera control unit 114 turns on the close-range illuminators 112 and 113 at the timing when the camera 111 captures an image. The camera control unit 114 may also have a function of determining whether the product 90 is held over the code scanner 110. The functions of the camera control unit 114 will be described in detail later (see FIG. 3).

[0017] The barcode recognition unit 115 recognizes coded information (e.g., product code) contained in a barcode captured in an image captured by the camera 111. Recognition of barcodes contained in images captured by a camera is widely and generally applied, so a description of the recognition method will be omitted. Note that the recognition target of the barcode recognition unit 115 is not limited to barcodes, and for example, two-dimensional codes may also be recognized. Note that the barcode recognition unit 115 is an example of a code symbol recognition unit in this disclosure.

[0018] The image transmission unit 116 transmits the image captured by the camera 111 to the recognition model creation device 130 that generates a recognition model 132 for recognizing the commodity 90, and to the commodity recognition device 140. The image transmission unit 116 is an example of an image output unit in the present disclosure.

[0019] The POS terminal 120 performs product registration based on the product code recognized by the barcode recognition unit 115 or the product code corresponding to the product information recognized by the product recognition device 140. The POS terminal 120 also performs payment processing for the registered product.

[0020] The POS terminal 120 includes a product registration unit 121, a payment processing unit 122, an operation instruction unit 123, an operation control unit 124, and a product information storage unit 125.

[0021] The product registration unit 121 registers a product corresponding to a product code input from the code scanner 110 or a product code recognized by the product recognition device 140 .

[0022] The payment processing unit 122 performs a payment process for the price of the product registered by the product registration unit 121. The price of the product may be paid in cash, electronic money, or a credit card.

[0023] The operation instruction unit 123 issues various operation instructions to the POS terminal 120. The operation instruction unit 123 also issues operation instructions to the recognition model creation device 130 and the commodity recognition device 140.

[0024] The operation control unit 124 acquires an operation instruction given by the operation unit 127. The operation instruction is, for example, an instruction to create a recognition model 132, an instruction to execute product registration, etc. Furthermore, when recognizing the product 90, the operation control unit 124 causes the operation instructing unit 123 to output an operation instruction to the product recognition device 140 and the code scanner 110. Furthermore, when creating a recognition model of the product 90, the operation control unit 124 causes the operation instructing unit 123 to output an operation instruction to the recognition model creation device 130 and the code scanner 110.

[0025] The product information storage unit 125 stores product information corresponding to a product code. Specifically, the product information storage unit 125 stores a product master, which is a master file of product information. The product master stores additional information such as the product name, price, and discount information of the product in association with a product code that uniquely identifies the product.

[0026] In addition, a display unit 126 and an operation unit 127 are connected to the POS terminal 120 .

[0027] The display unit 126 is a display device such as a liquid crystal monitor or an organic EL monitor that displays various operational states of the POS terminal 120 .

[0028] The operation unit 127 issues various operation instructions to the POS terminal 120, issues instructions to the recognition model creation device 130 to create the recognition model 132, and presents the product name held over the code scanner 110, i.e., presents so-called teacher information, when creating the recognition model 132. The operation unit 127 is an operation device such as a touch panel or a keyboard.

[0029] The recognition model creation device 130 is a device that creates a recognition model 132 for recognizing the product 90 from an image of the product 90 captured by the camera 111.

[0030] The recognition model creation device 130 includes a recognition model creation unit 131 and a recognition model 132 .

[0031] The recognition model creation unit 131 creates and strengthens the recognition model 132 from the image of the product 90 captured by the camera 111 and training information indicating the type of product the product 90 is. More specifically, the recognition model creation unit 131 performs learning to strengthen the recognition model 132 by using images of the product 90 captured from various angles.

[0032] The recognition model 132 is a model for recognizing the product 90, created by the recognition model creation unit 131. The recognition model 132 is expressed, for example, as a neural network with a multi-layer structure, and the input / output relationships of a plurality of internal signals for recognizing the product 90 are described inside the neural network. The recognition model 132 will be described in detail later (see FIG. 4).

[0033] The commodity recognition device 140 is a device that recognizes the commodity 90 captured by the camera 111 using the recognition model 132 .

[0034] The commodity recognition device 140 includes a commodity recognition unit 141 .

[0035] The commodity recognition unit 141 inputs an image of the commodity 90 to be recognized, captured by the camera 111, to the recognition model 132. Then, the commodity recognition unit 141 recognizes which commodity registered in the commodity master the image captured by the camera 111 is, based on a numerical value indicating the commodity likeliness output by the recognition model 132. In particular, in the commodity recognition system 100 of this embodiment, the commodity recognition device 140 recognizes an unknown commodity 90 captured by the same code scanner 110, using the recognition model 132 created by the recognition model creation device 130 based on an image of the commodity 90 captured by the code scanner 110. Therefore, the recognition accuracy can be improved compared to the case where the commodity recognition device 140 recognizes the commodity 90 appearing in an image captured by another camera.

[0036] The functional configuration shown in FIG. 1 is an example, and the functional parts of each device may be appropriately interchanged between different devices.

[0037] Next, the process of capturing an image of a commodity 90 by the code scanner 110 will be described with reference to Fig. 2. Fig. 2 is a diagram showing the commodity being held up over the code scanner.

[0038] As shown in Fig. 2, the camera 111 captures an image of a commodity 90 held within the imaging range, specifically, a commodity 90 close to the camera 111 in the Z-axis direction in Fig. 2. The camera control unit 114 of the code scanner 110 is preferably provided with a function for detecting that the commodity 90 has been held over, and causes the camera 111 to capture an image when the commodity 90 is held over. Specifically, the camera control unit 114 causes the camera 111 to repeatedly capture images, and when a sudden change in brightness is observed in the captured image, the camera control unit 114 determines that the commodity 90 has been held over and causes the camera 111 to capture an image. The camera control unit 114 may also transmit ultrasonic waves toward the imaging range P of the camera 111, and cause the camera 111 to capture an image when the ultrasonic sensor detects that an object has entered the imaging range P.

[0039] In order to prevent the background other than the commodity 90 from appearing in the image captured by the camera 111, the sensitivity of the imaging element of the camera 111 is set low. More specifically, the sensitivity of the camera 111 is adjusted so that the commodity 90 held in the imaging range P is captured with appropriate exposure when the commodity 90 is illuminated by the illumination light of the close-range illumination 112 or the illumination light of the close-range illumination 113. In addition, the focal range of the optical system of the camera 111 is adjusted to match the depth range of the imaging range P. As a result, the outside of the imaging range P captured by the camera 111 is imaged in a largely blurred state, and only the commodity 90 held by the camera 111 is imaged clearly.

[0040] Next, the imaging process performed by the camera control unit 114 will be described with reference to Fig. 3. Fig. 3 is a timing chart showing an example of the flow of the imaging process performed by the camera control unit.

[0041] The camera control unit 114 outputs an imaging instruction signal that instructs the imaging timing to the camera 111, the close-distance lighting 112, and the close-distance lighting 113. The imaging instruction signal is a signal that includes a positive pulse train.

[0042] When camera 111 detects the rising edge of the pulse of the imaging instruction signal, it opens the shutter of camera 111 to capture an image. The shutter of camera 111 closes after a preset exposure time t has elapsed. That is, camera 111 exposes the imaging element to light for the exposure time t.

[0043] On the other hand, when the close-distance lighting 112 and the close-distance lighting 113 detect the rising edge of the pulse of the imaging instruction signal, they turn on the lighting. The turned-on lighting is turned off when a preset lighting time i has elapsed. Here, the lighting time i is set to be shorter than the exposure time t of the camera 111. In this way, by making the lighting time of the close-distance lighting 112 and the close-distance lighting 113 shorter than the time the shutter of the camera 111 is open, it is possible to reduce power consumption.

[0044] Next, the contents of the recognition model 132 created by the recognition model creation device 130 will be described with reference to Fig. 4. Fig. 4 is a diagram showing an example of a recognition model created by the recognition model creation device.

[0045] Fig. 4 shows the structure of a portion of the created recognition model 132. In particular, Fig. 4 is an example of a model that recognizes whether a commodity 90 is a commodity A from an image of the commodity 90 captured by the camera 111.

[0046] The recognition model creation device 130 creates a recognition model that recognizes product A by providing a plurality of images of the product 90 captured by the camera 111 and information indicating that the product 90 is product A, that is, so-called training information.

[0047] The example shown in Fig. 4 is an example of a model for recognizing product A, described by a three-layer neural network having an input layer, an intermediate layer, and an output layer. A neural network is a mathematical model that imitates a human neural circuit network. Note that the number of layers of the neural network is not limited to three, and the neural network may have multiple intermediate layers.

[0048] The multiple units constituting the input layer detect multiple feature amounts a, b, ..., n from an image captured by the camera 111. The feature amounts a, b, ..., n are preset feature amounts to be detected from an image including a product 90. Specifically, the feature amounts a, b, ..., n are feature amounts related to the brightness of an image, feature amounts related to the shape of an object, feature amounts related to the color of an image, etc.

[0049] Note that the recognition model 132 includes recognition models for a plurality of products, but the feature amounts a, b, . . . , n are set in common to the recognition models for all the products.

[0050] The output of each unit in the input layer is weighted with a different weighting coefficient and input to the intermediate layer.

[0051] In the intermediate layer, the sum of the weighted outputs of each unit is calculated.

[0052] For example, in the first unit forming the intermediate layer, the evaluation value Ea is calculated based on the formula (1).

[0053] Ea = wa*a + wb*b + ... + wn*n ... (1)

[0054] In formula (1), wa, wb, ..., wn are weighting coefficients applied to the output of the input layer. Also, in formula (1), a, b, ..., n are feature quantities calculated in the input layer. Similarly, evaluation values ​​Eb, Ec, ..., En are calculated in other units of the hidden layer. Note that different values ​​are set for the weighting coefficients wa, wb, ..., wn depending on the units forming the hidden layer.

[0055] The n evaluation values ​​Ea, Eb, . . . , En calculated in the intermediate layer are weighted with different weighting coefficients and output to the output layer.

[0056] In the output layer, an evaluation value K representing the likeliness of product A is calculated based on equation (2).

[0057] K = ua*Ea+ub*Eb+…+un*En…(2)

[0058] In equation (2), ua, ub, ..., un are weighting coefficients attached to the output of each unit in the hidden layer, and Ea, Eb, ..., En are the evaluation values ​​described above.

[0059] The recognition model creation device 130 repeats learning every time the camera 111 captures multiple images of the product 90. Then, the recognition model creation device 130 strengthens the recognition model by successively updating the weighting coefficients wa, wb, ..., wn and weighting coefficients ua, ub, ..., un each time learning is performed.

[0060] Next, the flow of processing performed by the commodity recognition system 100 will be described with reference to Fig. 5. Fig. 5 is a flowchart showing an example of the flow of processing performed by the commodity recognition system.

[0061] The operation control unit 124 of the POS terminal 120 determines whether an instruction to create the recognition model 132 has been given (step S1). If it is determined that an instruction to create the recognition model 132 has been given (step S1: Yes), the process proceeds to step S2. On the other hand, if it is determined that an instruction to create the recognition model 132 has not been given (step S1: No), the process proceeds to step S3. The operation control unit 124 of the POS terminal 120 determines whether an instruction to create the recognition model 132 has been given based on the contents of the operation instruction given to the operation unit 127.

[0062] When it is determined in step S1 that an instruction to create the recognition model 132 has been issued, the operation control unit 124 of the POS terminal 120 causes the code scanner 110 and the recognition model creation device 130 to perform a process of creating the recognition model 132 (step S2). A detailed flow of the process of creating the recognition model will be described later (see FIG. 6). Thereafter, the commodity recognition system 100 ends the process of FIG. 5.

[0063] On the other hand, if it is determined in step S1 that the creation of the recognition model 132 has not been instructed, the operation control unit 124 of the POS terminal 120 determines whether or not an instruction to execute product registration has been given (step S3). If it is determined that an instruction to execute product registration has been given (step S3: Yes), the process proceeds to step S4. On the other hand, if it is determined that an instruction to execute product registration has not been given (step S3: No), the product recognition system 100 ends the process in Fig. 5. The operation control unit 124 of the POS terminal 120 determines whether or not an instruction to execute product registration has been given based on the contents of the operation instruction given to the operation unit 127.

[0064] When it is determined in step S3 that execution of product registration has been instructed, the POS terminal 120 performs product registration processing in cooperation with the code scanner 110 and the product recognition device 140 (step S4). A detailed flow of the product registration processing will be described later (see FIG. 7).

[0065] Next, the POS terminal 120 performs a payment process for the registered product (step S5). After that, the product recognition system 100 ends the process of FIG.

[0066] Next, a flow of a recognition model creation process will be described with reference to Fig. 6. Fig. 6 is a diagram showing an example of the flow of a recognition model creation process performed by the recognition model creation device of the embodiment.

[0067] The recognition model creation unit 131 acquires product identification information, such as the product name, held over the code scanner 110 from the operation unit 127 of the POS terminal 120 (step S11).

[0068] The camera control unit 114 of the code scanner 110 determines whether the product 90 is held over the code scanner 110 (step S12). If it is determined that the product 90 is held over the code scanner 110 (step S12: Yes), the process proceeds to step S13. On the other hand, if it is not determined that the product 90 is held over the code scanner 110 (step S12: No), the determination in step S12 is repeated.

[0069] When it is determined in step S12 that the commodity 90 is held over the code scanner 110, the camera control unit 114 of the code scanner 110 causes the camera 111 to capture an image (step S13). At this time, the camera control unit 114 turns on the close-range lights 112 and 113 at the timing when the camera 111 captures the image.

[0070] The recognition model creation unit 131 of the recognition model creation device 130 updates the recognition model 132 of the product corresponding to the product 90 by applying the image of the product 90 captured in step S13 to the recognition model 132 of the product corresponding to the product 90 that is being created at that time (step S14).

[0071] The operation control unit 124 of the POS terminal 120 determines whether an instruction to end the process of creating the recognition model 132 has been given from the operation unit 127 (step S15). If it is determined that an instruction to end the process of creating the recognition model 132 has been given (step S15: Yes), the commodity recognition system 100 ends the process of Fig. 6. On the other hand, if it is not determined that an instruction to end the process of creating the recognition model 132 has been given (step S15: No), the process returns to step S12 and the creation of the recognition model 132 is repeated.

[0072] Next, the flow of the product registration process will be described with reference to Fig. 7. Fig. 7 is a diagram showing an example of the flow of the product registration process performed by the POS terminal of the embodiment.

[0073] The camera control unit 114 of the code scanner 110 determines whether the product 90 is held over the code scanner 110 (step S21). If it is determined that the product 90 is held over the code scanner 110 (step S21: Yes), the process proceeds to step S22. On the other hand, if it is not determined that the product 90 is held over the code scanner 110 (step S21: No), the determination in step S21 is repeated.

[0074] When it is determined in step S21 that the commodity 90 is held over the code scanner 110, the camera control unit 114 of the code scanner 110 causes the camera 111 to capture an image (step S22). At this time, the camera control unit 114 turns on the close-range lights 112 and 113 at the timing when the camera 111 captures the image.

[0075] The barcode recognition unit 115 of the code scanner 110 recognizes the code symbol from within the image captured by the camera 111 (step S23).

[0076] Next, the barcode recognition unit 115 judges whether the code symbol has been recognized (step S24). If it is judged that the code symbol has been recognized (step S24: Yes), the process proceeds to step S25. On the other hand, if it is not judged that the code symbol has been recognized (step S24: No), the process proceeds to step S26.

[0077] When it is determined in step S24 that the code symbol has been recognized, the product registration unit 121 of the POS terminal 120 identifies the product 90 held over the code scanner 110 based on the recognized code symbol (step S25). Then, the process proceeds to step S27.

[0078] On the other hand, if it is determined in step S24 that the code symbol has not been recognized, the commodity recognition unit 141 of the commodity recognition device 140 identifies the commodity 90 held over the code scanner 110 based on the recognition model 132 (step S26). Then, the process proceeds to step S27.

[0079] Following step S25 or step S26, the product registration unit 121 of the POS terminal 120 registers the recognized product (step S27).

[0080] The operation control unit 124 of the POS terminal 120 determines whether an instruction to end registration has been given from the operation unit 127 (step S28). If it is determined that an instruction to end registration has been given (step S28: Yes), the product recognition system 100 ends the process in Fig. 7. On the other hand, if it is not determined that an instruction to end registration has been given (step S28: No), the process returns to step S21 and repeats the registration of another product.

[0081] As described above, the code scanner 110 (image input device) of the embodiment includes a camera 111 (imaging unit) that captures an image of a close-range area from the camera 111 to obtain an image of a commodity 90 held over the camera 111, close-range illuminators 112, 113 (illumination unit) that illuminate the imaging range of the camera 111, and an image transmission unit 116 (image output unit) that outputs the image captured by the camera 111 to a recognition model creation device 130 that generates a recognition model 132 for recognizing the commodity 90. Therefore, the recognition model 132 used when recognizing the commodity 90 can be generated easily and reliably.

[0082] Furthermore, in the code scanner 110 (image input device) of the embodiment, the camera 111 (imaging unit) has sensitivity to capture an image of the commodity 90 with appropriate exposure when the commodity 90 held over the camera 111 is illuminated by close-range illuminators 112, 113 (illumination units), and the close-range illuminators 112, 113 are turned on at least within the exposure period of the camera 111. Therefore, only the commodity 90 held over the imaging range P close to the code scanner 110 can be clearly imaged, and an image suitable for creating the recognition model 132 can be obtained.

[0083] The code scanner 110 (image input device) of the embodiment further includes a barcode recognition unit 115 (code symbol recognition unit) that recognizes a code symbol appearing in an image captured by the camera 111 (imaging unit). Therefore, the code scanner 110 can be used to create a recognition model 132 in the recognition model creation device 130.

[0084] Moreover, the commodity recognition system 100 of the embodiment includes a code scanner 110 (image input device), a recognition model creation device 130 that uses an image of the commodity 90 input from the code scanner 110 to create a recognition model 132 for recognizing the commodity 90, and a commodity recognition device 140 that recognizes the commodity 90 using the image of the commodity 90 input from the code scanner 110 and the recognition model 132 created by the recognition model creation device 130. Therefore, the creation of the recognition model 132 and the recognition of the commodity 90 using the recognition model 132 can be executed by the same device. This allows the commodity 90 to be recognized with higher accuracy.

[0085] Furthermore, the commodity recognition system 100 of the embodiment operates the code scanner 110 (image input device) and the commodity recognition device 140 when recognizing the commodity 90, and operates the code scanner 110 and the recognition model creation device 130 when creating a recognition model 132 of the commodity 90. Therefore, it is possible to easily switch between creating the recognition model 132 by the recognition model creation device 130 and recognizing the commodity 90 by the commodity recognition device 140.

[0086] Although the embodiment of the present invention has been described above, this embodiment is merely an example and is not intended to limit the scope of the invention. This novel embodiment can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0087] 100 Product Recognition System 110 Code scanner (image input device) 111 Camera (imaging unit) 112,113 Near field lighting (lighting section) 114 Camera control unit 115 Barcode recognition unit (code symbol recognition unit) 116 Image transmission unit (image output unit) 120 POS terminals 121 Product Registration Department 122 Payment processing unit 123 Operation instruction section 124 Motion control section 125 Product information storage section 130 Recognition model creation device 131 Recognition Model Creation Department 132 Recognition Model 140 Product recognition device 141 Product recognition department P Image capture range Q,R Lighting range i Lighting time t Exposure time [Prior art documents] [Patent documents]

[0088] [Patent Document 1] International Publication No. 2020 / 170963

Claims

1. an imaging unit that captures an image of an area close to the imaging unit and acquires an image of a product held up to the imaging unit; an illumination unit that illuminates an imaging range of the imaging unit; and an image output unit that outputs the image captured by the imaging unit to a recognition model creation device that creates a recognition model for recognizing the product. Image input device.

2. The imaging unit has a sensitivity for capturing an image of a commodity with an appropriate exposure when the commodity is illuminated by the illumination unit. The illumination unit is turned on at least during an exposure period of the imaging unit.

2. The image input device according to claim 1.

3. The image capturing device further includes a code symbol recognition unit that recognizes a code symbol appearing in the image captured by the image capturing unit.

3. The image input device according to claim 1.

4. The image input device according to claim 3 ; a recognition model creation device that creates a recognition model for recognizing a commodity by using an image of the commodity input from the image input device; a commodity recognition device that recognizes the commodity using an image of the commodity inputted from the image input device and the recognition model created by the recognition model creation device; A product recognition system comprising:

5. The product recognition system includes: When recognizing a product, the image input device and the product recognition device are operated, and when creating a product recognition model, the image input device and the recognition model creation device are operated. The product recognition system according to claim 4.

Citation Information

Patent Citations

  • Information processing apparatus and program

    JP2014052811A

  • Recognition system, information processor and program

    JP2017215791A

  • Visual inspection device and blister packaging machine

    JP2019039762A

  • Lighting adjustment device, lighting adjustment method, and item recognition system

    JP2022178971A

  • Processing system, processing method, and program

    WO2021234842A1