A method and apparatus for identifying a product
By narrowing down the product feature library based on shelf product information or standard display information, the problem of high resource and time consumption in product identification is solved, achieving more efficient and accurate product identification.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANGHAI HANSHI INFORMATION TECH CO LTD
- Filing Date
- 2022-12-26
- Publication Date
- 2026-06-26
AI Technical Summary
Existing technologies consume significant resources and time for product identification, making it difficult to achieve efficient and accurate product identification and impacting intelligent shelf management.
The scope of the product feature library is narrowed down based on shelf product information or standard display information. The narrowed feature library is then used for product retrieval and identification, reducing the number of features to be compared and the time consumed.
It achieves more accurate product recognition, saves resources and time costs, and improves recognition efficiency.
Smart Images
Figure CN115861781B_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of artificial intelligence technology, specifically relating to a product recognition method and apparatus. Background Technology
[0002] With the rapid development of artificial intelligence, smart retail has also developed rapidly. In order to achieve efficient and intelligent merchandise management, smart retail comprehensively utilizes technologies such as the Internet, big data, and artificial intelligence to empower or upgrade existing supermarkets and convenience stores, making them more digital and intelligent.
[0003] Intelligent digital shelves are a crucial component of smart retail operations. Functions such as intelligent detection and identification of goods, stockout monitoring, and display supervision necessitate intelligent management of these shelves. Goods identification, a key aspect of intelligent shelf management and a prerequisite for applications like shelf layout analysis and stockout analysis, is a core technology for achieving intelligent shelf management. Therefore, accurate goods identification plays a vital role in intelligent shelf management. However, due to the diverse range of products, frequent packaging updates, and changing environmental conditions, achieving high-precision goods identification presents significant challenges. This requires not only massive amounts of data but also highly accurate goods identification technology.
[0004] To achieve high-precision product recognition, the most effective method currently is image-based product retrieval. This involves a feature vector database; during retrieval, the target feature is compared to the database using distance or similarity calculations, and the category of the feature with the highest similarity is used as the target feature's category. However, due to the vast variety of products, product recognition consumes significant resources and time, hindering accuracy and efficiency, and consequently impacting subsequent intelligent shelf management.
[0005] Application content
[0006] The purpose of this application is to provide a product identification method and apparatus to address the shortcomings of existing product identification technologies, which involve significant resource and time consumption.
[0007] To solve the above-mentioned technical problems, this application is implemented as follows:
[0008] Firstly, a product identification method is provided, including the following steps:
[0009] Based on the display location of the product to be identified, determine the product category corresponding to the product to be identified;
[0010] Based on the product characteristics of the product to be identified, a comparative search is performed in the product feature library corresponding to the product category, and the feature category with the highest similarity to the product characteristics is selected from the product feature library as the category of the product to be identified.
[0011] Secondly, a product identification device is provided, comprising:
[0012] The determination module is used to determine the product category corresponding to the product to be identified based on the display location of the product to be identified;
[0013] The retrieval module is used to perform a comparative retrieval in the product feature library corresponding to the product category based on the product features of the product to be identified, and select the feature category with the highest similarity to the product features from the product feature library as the category of the product to be identified.
[0014] This application embodiment determines the corresponding product category based on the display location of the product to be identified, thereby narrowing down the product categories in the product feature library, and uses the narrowed product feature library for product retrieval and identification, thereby saving resource and time costs and achieving more accurate product identification. Attached Figure Description
[0015] Figure 1 This is a flowchart of a product identification method provided in an embodiment of this application;
[0016] Figure 2 This is a schematic diagram of a shelf layer for displaying goods provided in an embodiment of this application;
[0017] Figure 3 This is a schematic diagram of the structure of a product identification device provided in an embodiment of this application. Detailed Implementation
[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0019] In this embodiment, images of all products requiring identification within the supermarket can be collected to construct a product identification feature library. During product identification, all product features within the entire feature library are compared, and the category of the feature with the highest similarity is ultimately used as the category of the product to be identified. The aforementioned product identification technology uses a feature library constructed from all products for product feature comparison and retrieval. However, the overall feature library contains millions or even tens of millions of product features, making accurate product identification extremely difficult. Furthermore, comparing and retrieving features from such a massive amount consumes significant resources and time, making real-time performance impossible.
[0020] Therefore, this application proposes a product identification method based on shelf information, utilizing shelf product information (or more detailed standard shelf product display information) provided by supermarkets. This method can reduce the feature categories in the feature library to be compared and retrieved to the categories of products placed on the shelf. If the supermarket can provide more detailed standard shelf product display information, the feature types in the feature library to be compared and retrieved can be reduced to a smaller range of product types, thereby reducing the number of features in the feature library. The method uses shelf product information or standard shelf product display information to obtain the category, type, and even the order in which the products are placed on the shelf. Then, based on the product category or type information, the product feature library is narrowed down to minimize product features irrelevant to the product to be identified.
[0021] In real-world scenarios, supermarkets typically group similar products onto the same shelf (or the same shelf level) for easier management, meaning the product categories on a shelf are fixed (or known). Furthermore, some supermarkets, based on brand requirements, arrange products according to standard shelf display diagrams specified by the brand. In this case, the standard display diagrams provide more detailed product placement information. The product identification method proposed in this application, based on shelf product information or standard shelf display information, first narrows down the product feature database for comparison before product retrieval. Then, it performs more precise product retrieval and identification within this narrowed feature database. This not only reduces the number of product features compared and improves the accuracy of product identification but also reduces resource and time consumption during product retrieval, making product identification truly fast and accurate.
[0022] The following description, in conjunction with the accompanying drawings, details a product identification method provided by this application through specific embodiments and application scenarios.
[0023] like Figure 1 The diagram shown is a flowchart of a product identification method provided in an embodiment of this application. The method includes the following steps:
[0024] Step 101: Determine the product category corresponding to the product to be identified based on the display location of the product to be identified.
[0025] Specifically, the product category corresponding to the shelf where the product to be identified is located can be determined as the product category corresponding to the product to be identified; wherein, all products on the shelf belong to the product category. Alternatively, based on the standard shelf product display information, a first product category corresponding to the display position of the product to be identified on the shelf, and a second product category corresponding to the position adjacent to the display position can be determined; based on the product categories corresponding to the first and second product categories, the product category corresponding to the product to be identified can be determined; wherein, the standard shelf product display information is used to determine the category of the product placed at each position on the shelf.
[0026] In this embodiment, the product categories corresponding to the first product category and the second product category can be determined as the product categories corresponding to the product to be identified.
[0027] Step 102: Based on the product features of the product to be identified, perform a comparative search in the product feature library corresponding to the product category, and select the feature category with the highest similarity to the product features from the product feature library as the category of the product to be identified.
[0028] In this embodiment, before determining the product category corresponding to the product to be identified based on its display location, the position of the shelf to which the product to be identified belongs can be located based on the center point coordinates of the detection frame of the product to be identified. Accordingly, the first product category corresponding to the position of the shelf to which the product to be identified belongs and the second product category corresponding to the positions of the multiple shelves adjacent to the shelf can be determined based on the standard display information of the shelf products.
[0029] This application embodiment determines the corresponding product category based on the display location of the product to be identified, thereby narrowing down the product categories in the product feature library, and uses the narrowed product feature library for product retrieval and identification, thereby saving resource and time costs and achieving more accurate product identification.
[0030] In this embodiment, the types and number of features in the product feature library to be compared and retrieved are reduced by using shelf product information or standard shelf product display information provided by supermarkets. This allows the product to be identified to be compared with fewer features in the feature library, resulting in more accurate product identification results and lower resource and time costs. Here, "product category" represents a broad category of products, such as beverages, biscuits, personal care products, etc., while "product type" represents the specific category of a particular product. "Shelf" represents a fixed area on the shelf, such as... Figure 2The dashed box shown is based on the shelf layer.
[0031] In one specific implementation of this application embodiment, when using supermarket shelf product information, the product identification method based on the shelf product information to narrow down the product feature library is implemented in the following detailed steps:
[0032] (1) Assume that there are 10 categories of products in the overall product feature library, such as A, B, C, D, E, F, G, H, I, J;
[0033] (2) For a product to be identified, all products on the shelf it is located in belong to category A or B;
[0034] (3) Use the shelf product information (category A and B) provided in (2) to narrow down the product feature library to include only the features of products in categories A and B;
[0035] (4) Compare and search the features of the product to be identified with the features in the feature library of the product category narrowed down in (3);
[0036] (5) Select the feature category with the highest similarity as the category of the product to be identified.
[0037] In another specific implementation of this application embodiment, the product identification method based on a product feature library narrowed down from standard shelf product display information is implemented in the following detailed steps:
[0038] (1) Based on the standard display information of the goods on the shelves, the categories of goods placed in each position on each shelf can be obtained. Assume... Figure 2 The products labeled on the first shelf of the middle shelf are the products to be identified. The coordinates of the center point of the detection frame can be calculated based on the detection frame of the product to be identified.
[0039] (2) Based on the coordinates of the center point of the commodity inspection frame calculated in (1), its position on the shelf can be located, such as... Figure 2 The sheds belonging to category A4;
[0040] (3) Select n racks to the left and right of the rack to which the product belongs, and get the product categories within the range of 2n+1 racks; assuming n=1, select racks A3, A4, and A5, and get the three product categories A3, A4, and A5;
[0041] (4) Based on the 2n+1 product categories selected in (3), the categories of features in the product feature library are reduced to 2n+1 categories;
[0042] (5) Compare and search the features of the product to be identified with the features in the feature library of the product category narrowed down in (4);
[0043] (6) Select the feature category with the highest similarity as the category of the product to be identified.
[0044] Before identifying goods on the shelf, this embodiment selects features from a full-category product feature library based on the product information of the shelf or the standard product display information of the shelf. Features related to the product category on the shelf or the product category of the selected shelf are selected and used for subsequent product identification. This can significantly reduce the categories and types of goods in the product feature library, reduce the number of features that need to be compared, and then use the reduced feature library to perform product retrieval and identification. This not only saves a lot of resources and time costs, but also achieves more accurate product identification.
[0045] like Figure 3 The diagram shown is a structural schematic of a product identification device provided in an embodiment of this application, comprising:
[0046] The determination module 310 is used to determine the product category corresponding to the product to be identified based on the display position of the product to be identified.
[0047] Specifically, the determining module 310 can be used to determine the product category corresponding to the shelf where the product to be identified is located as the product category corresponding to the product to be identified; wherein all products on the shelf belong to the product category. It can also be used to determine, based on the standard shelf product display information, a first product category corresponding to the display position of the product to be identified on the shelf, and a second product category corresponding to the position adjacent to the display position; and to determine the product category corresponding to the product to be identified based on the product categories corresponding to the first and second product categories; wherein the standard shelf product display information is used to determine the category of the product placed at each position on the shelf.
[0048] In this embodiment, the determining module 310 is specifically used to determine the product category corresponding to the first product category and the second product category as the product category corresponding to the product to be identified.
[0049] The retrieval module 320 is used to perform a comparative retrieval in the product feature library corresponding to the product category based on the product features of the product to be identified, and select the feature category with the highest similarity to the product features from the product feature library as the category of the product to be identified.
[0050] In this embodiment, the above-mentioned device further includes:
[0051] The positioning module is used to locate the position of the shelf compartment of the product to be identified based on the coordinates of the center point of the detection frame of the product to be identified;
[0052] Accordingly, the determining module 310 is specifically used to determine, based on the standard display information of the goods on the shelf, the first product category corresponding to the position of the shelf to which the product to be identified belongs, and the second product category corresponding to the positions of the multiple shelf adjacent to the shelf.
[0053] This application embodiment determines the corresponding product category based on the display location of the product to be identified, thereby narrowing down the product categories in the product feature library, and uses the narrowed product feature library for product retrieval and identification, thereby saving resource and time costs and achieving more accurate product identification.
[0054] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0055] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.
[0056] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.
Claims
1. A product identification method, characterized in that, Includes the following steps: Based on the display location of the product to be identified, determine the product category corresponding to the product to be identified; specifically including: Based on the product information displayed on the shelf or the standard product display information on the shelf, the product category corresponding to the shelf where the product to be identified is located is determined as the product category corresponding to the product to be identified; Based on the product characteristics of the product to be identified, a comparative search is performed in the product feature library corresponding to the product category, and the feature category with the highest similarity to the product characteristics is selected from the product feature library as the category of the product to be identified. The step of determining the product category corresponding to the product to be identified based on its display location specifically includes: Based on the standard display information of the goods on the shelf, determine the first product category corresponding to the display position of the product to be identified on the shelf, and the second product category corresponding to the position adjacent to the display position; Based on the product categories corresponding to the first product category and the second product category, determine the product category corresponding to the product to be identified; Among them, the standard display information of goods on the shelves is used to determine the category of goods placed in each position on the shelves.
2. The method according to claim 1, characterized in that, The step of determining the product category corresponding to the shelf where the product to be identified is located, based on the product information displayed on the shelf, specifically includes: For the product to be identified, all products on the shelf where it is located belong to the product category.
3. The method according to claim 1, characterized in that, The step of determining the product category corresponding to the product to be identified based on the first product category and the second product category specifically includes: The product categories corresponding to the first product category and the second product category are determined as the product categories corresponding to the product to be identified.
4. The method according to claim 1, characterized in that, Before determining the product category corresponding to the product to be identified based on its display location, the method further includes: Based on the coordinates of the center point of the detection frame of the product to be identified, locate the position of the shelf compartment to which the product belongs on the shelf; The step of determining the first product category corresponding to the display position of the product to be identified on the shelf, and the second product category corresponding to the position adjacent to the display position, based on the standard display information of the shelf merchandise, specifically includes: Based on the standard display information of the goods on the shelf, determine the first product category corresponding to the position of the shelf to be identified, and the second product category corresponding to the positions of the multiple shelf adjacent to the shelf.
5. A product identification device, characterized in that, include: The determination module is used to determine the product category corresponding to the product to be identified based on the display location of the product to be identified; Specifically, it is used to determine the product category corresponding to the shelf where the product to be identified is located based on the product information displayed on the shelf or the standard product display information on the shelf; The retrieval module is used to perform a comparative retrieval in the product feature library corresponding to the product category based on the product features of the product to be identified, and select the feature category with the highest similarity to the product features from the product feature library as the category of the product to be identified; The determining module is specifically used to determine, based on the standard shelf display information, the first product category corresponding to the display position of the product to be identified on the shelf, and the second product category corresponding to the position adjacent to the display position; and to determine the product category corresponding to the product to be identified based on the product categories corresponding to the first product category and the second product category; wherein, the standard shelf display information is used to determine the category of the product placed at each position on the shelf.
6. The apparatus according to claim 5, characterized in that, The step of determining the product category corresponding to the shelf where the product to be identified is located, based on the product information displayed on the shelf, specifically includes: For the product to be identified, all products on the shelf where it is located belong to the product category.
7. The apparatus according to claim 5, characterized in that, The determining module is specifically used to determine the product category corresponding to the first product category and the second product category as the product category corresponding to the product to be identified.
8. The apparatus according to claim 5, characterized in that, Also includes: The positioning module is used to locate the position of the shelf compartment of the product to be identified based on the coordinates of the center point of the detection frame of the product to be identified; The determining module is specifically used to determine, based on the standard display information of the goods on the shelf, the first product category corresponding to the position of the shelf to be identified, and the second product category corresponding to the positions of the multiple shelf adjacent to the shelf.