Product detection method and device based on artificial intelligence, equipment and storage medium

Abstract

The invention relates to a product detection method and device based on artificial intelligence, equipment and a storage medium. The method comprises the steps that identification codes on a bottle body and a packaging box of a to-be-detected product are obtained through a camera; inputting the identification code into a deep learning network model, and if it is determined that the bottle body identification code and the packaging box identification code of the to-be-detected product meet preset requirements through the deep learning network model, determining that the to-be-detected product is a first product; if it is determined that the to-be-detected product is the first product, preset associated information is obtained, and the associated information represents the association relationship between the bottle body identification code and the packaging box identification code; and if the bottle body identification code and the packaging box identification code of the to-be-detected product accord with the association relationship, determining that the to-be-detected product is a qualified product. The detection efficiency and accuracy can be improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular, to a product detection method, device, equipment, and storage medium based on artificial intelligence. Background Art

[0002] In the field of liquor product detection, traditional methods mainly rely on manual detection. For the detection of products and corresponding packaging boxes, through two manual workstations, one for detecting the bottle body and the other for detecting the packaging box, and then manually holding a barcode scanner to scan the QR codes on the bottle body and the packaging box for spot checks. In the traditional production process, the accuracy and integrity of this information often need to be checked manually, which not only increases the production cost but also may lead to the risks of missed inspections and misjudgments.

[0003] Traditional existing technologies cannot detect systematically and on a large scale through manual detection, and their detection efficiency is low and the detection accuracy is limited. Therefore, there is an urgent need for a method to solve the above problems. Summary of the Invention

[0004] Based on this, this application provides a product detection method, device, equipment, and storage medium based on artificial intelligence to solve the problems existing in the prior art.

[0005] In a first aspect, a product detection method based on artificial intelligence is provided. The method includes:

[0006] Obtain the identification codes on the bottle body and the packaging box of the product to be detected through a camera respectively;

[0007] Input the identification code into a deep learning network model. If it is determined through the deep learning network model that both the bottle body identification code and the packaging box identification code of the product to be detected meet the preset requirements, then determine that the product to be detected is a first product;

[0008] If it is determined that the product to be detected is a first product, then obtain the pre-set associated information, where the associated information represents the association relationship between the bottle body identification code and the packaging box identification code;

[0009] If the bottle body identification code and the packaging box identification code of the product to be detected meet the association relationship, then determine that the product to be detected is a qualified product.

[0010] According to an implementable manner in the embodiments of this application, the method further includes:

[0011] If it is determined through the deep learning network model that the bottle body identification code or the packaging box identification code of the product to be detected does not meet the preset requirements, then determine that the product to be detected is an unqualified product and reject it.

[0012] According to an implementable manner in an embodiment of the present application, the method further includes:

[0013] If the bottle identification code and the packaging box identification code of the product to be detected do not conform to the association relationship, it is determined that the product to be detected is a non-conforming product and is rejected.

[0014] According to an implementable manner in an embodiment of the present application, the identification code includes a first two-dimensional code and a first inkjet code located on the bottle body, and a second two-dimensional code and a second inkjet code located on the packaging box;

[0015] When inputting the identification code into the deep learning network model, if it is determined through the deep learning network model that both the bottle identification code and the packaging box identification code of the product to be detected meet the preset requirements, it is determined that the product to be detected is a first product, including:

[0016] Input the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code into the deep learning network model respectively. If it is determined through the deep learning network model that the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code all meet the preset requirements, it is determined that the product to be detected is a first product.

[0017] According to an implementable manner in an embodiment of the present application, when inputting the identification code into the deep learning network model, if it is determined through the deep learning network model that both the bottle identification code and the packaging box identification code of the product to be detected meet the preset requirements, it is determined that the product to be detected is a first product, including:

[0018] Obtain multiple identification code samples and corresponding annotation files to construct a data set, where the annotation file represents the annotation information corresponding to the identification code, and the annotation information includes that the identification code is normal or the identification code is abnormal;

[0019] Input the data set into a convolutional neural network to generate a deep learning network model;

[0020] Input the identification code of the product to be detected into the deep learning network model, and determine the annotation information of the bottle identification code and the packaging box identification code through the deep learning network model;

[0021] If the annotation information of the bottle identification code and the annotation information of the packaging box identification code are both normal, it is determined that the product to be detected is a first product.

[0022] According to an implementable manner in an embodiment of the present application, the association information includes:

[0023] The association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code;

[0024] If the bottle body identification code and the packaging box identification code of the product to be detected conform to the association relationship, it is determined that the product to be detected is a qualified product, including:

[0025] If the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code all conform to the pre-set association information, it is determined that the product to be detected is a qualified product.

[0026] According to an implementable manner in the embodiments of the present application, the identification code is abnormal, including:

[0027] One or more of the identification code being skewed, the identification code being incomplete, the position of the identification code exceeding the preset range, the identification code being repeated, and the identification code not existing.

[0028] In a second aspect, a product detection device based on artificial intelligence is provided. The device includes:

[0029] An acquisition module: used to obtain the identification codes on the bottle body and the packaging box of the product to be detected through a camera respectively;

[0030] An identification module: used to input the identification code into a deep learning network model. If it is determined through the deep learning network model that the bottle body identification code and the packaging box identification code of the product to be detected both meet the preset requirements, it is determined that the product to be detected is a first product;

[0031] An acquisition module: used to obtain the pre-set association information if it is determined that the product to be detected is a first product. The association information represents the association relationship between the bottle body identification code and the packaging box identification code;

[0032] A determination module: used to determine that the product to be detected is a qualified product if the bottle body identification code and the packaging box identification code of the product to be detected conform to the association relationship.

[0033] In a third aspect, a computer device is provided, including:

[0034] At least one processor; and

[0035] A memory communicatively connected to the at least one processor; wherein,

[0036] The memory stores computer instructions executable by the at least one processor. The computer instructions are executed by the at least one processor so that the at least one processor can execute the method involved in the first aspect above.

[0037] In a fourth aspect, a computer-readable storage medium is provided, on which computer instructions are stored, characterized in that the computer instructions are used to cause a computer to execute the method described in the first aspect above.

[0038] According to the technical content provided by the embodiments of the present application, the identification codes on the bottle body and the packaging box of the product to be detected are respectively obtained through a camera, and the identification codes are input into a deep learning network model. If it is determined through the deep learning network model that both the identification code on the bottle body and the identification code on the packaging box of the product to be detected meet the preset requirements, it is determined that the product to be detected is a first product. If it is determined that the product to be detected is a first product, the pre-set associated information is obtained. The associated information represents the association relationship between the identification code on the bottle body and the identification code on the packaging box. If the identification code on the bottle body and the identification code on the packaging box of the product to be detected conform to the association relationship, it is determined that the product to be detected is a qualified product. By obtaining the identification codes on the bottle body and the packaging box of the product to be detected through a camera and using a deep learning network model for identification and verification, it is effectively ensured that the identification codes meet the preset requirements. This solution can automatically determine whether both the identification code on the bottle body and the identification code on the packaging box meet the preset requirements, and further verify the association relationship between the two through the pre-set associated information, thereby accurately determining the qualification of the product. This method not only improves the efficiency and accuracy of product detection, but also greatly reduces the errors and costs of manual detection, and can achieve systematic and large-scale detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic flowchart of a product detection method based on artificial intelligence in an embodiment;

[0040] Figure 2 It is a flowchart of a product detection method based on artificial intelligence in an embodiment;

[0041] Figure 3 It is a structural block diagram of a product detection device based on artificial intelligence in an embodiment;

[0042] Figure 4 It is a schematic structural diagram of a computer device in an embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] The following further describes the present application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0044] Figure 1 It is a flowchart of a product detection method based on artificial intelligence provided by an embodiment of the present application, as Figure 1 shown, and in combination with Figure 2 , the method may include the following steps:

[0045] Step 101: Obtain the identification codes on the bottle body and the packaging box of the product to be detected through a camera respectively.

[0046] Specifically, the product can be products such as wine bottles (such as white wine, red wine), cosmetics, medicines, etc. The product includes a bottle body and a packaging box, and identification codes are respectively pasted or printed on the bottle body and the packaging box. In the embodiments of the present application, taking a wine bottle as an example, the identification codes on the bottle body and the packaging box of the product to be detected are obtained through a camera respectively.

[0047] Step 102: Input the identification codes into a deep learning network model. If it is determined through the deep learning network model that both the identification code on the bottle body and the identification code on the packaging box of the product to be detected meet the preset requirements, then determine that the product to be detected is a first product.

[0048] Specifically, input the identification codes on the bottle body and the packaging box of the product into the deep learning network model, and identify the identification codes through the deep learning network model. If it is determined through the deep learning network model that both the identification code on the bottle body and the identification code on the packaging box of the product to be detected meet the preset requirements, where the preset requirements indicate that the identification code is complete, clear and within the position range. If it is determined through the deep learning network model that both the identification code on the bottle body and the identification code on the packaging box of the product to be detected meet the preset requirements, then determine that the product to be detected is a first product, and continue to perform the detection in steps 103 - 104 on the first product.

[0049] Step 103: If it is determined that the product to be detected is a first product, then obtain the pre-set associated information, and the associated information represents the association relationship between the identification code on the bottle body and the identification code on the packaging box.

[0050] Specifically, if it is determined that the product to be detected is a first product, then obtain the pre-set associated information, and the associated information represents the association relationship between the identification code on the bottle body and the identification code on the packaging box. This association relationship is pre-stored in the host computer, and the corresponding association relationship is pre-stored for each product. If a product with an incorrect detected association relationship is considered a non-conforming product, it is rejected and an alarm is issued.

[0051] Step 104: If the identification code on the bottle body and the identification code on the packaging box of the product to be detected conform to the association relationship, then determine that the product to be detected is a qualified product.

[0052] Specifically, if the identification code on the bottle body and the identification code on the packaging box of the product to be detected conform to the association relationship, then determine that the product to be detected is a qualified product. If a product with an incorrect detected association relationship, that is, a product that does not conform to the association relationship, is considered a non-conforming product, it is rejected and an alarm is issued.

[0053] It can be seen that in the embodiments of the present application, the camera is used to obtain the identification codes on the bottle body and the packaging box of the product to be detected respectively, and the identification codes are input into the deep learning network model. If it is determined through the deep learning network model that both the identification code on the bottle body and the identification code on the packaging box of the product to be detected meet the preset requirements, it is determined that the product to be detected is a first product. If it is determined that the product to be detected is a first product, the pre-set associated information is obtained. The associated information represents the association relationship between the identification code on the bottle body and the identification code on the packaging box. If the identification code on the bottle body and the identification code on the packaging box of the product to be detected meet the association relationship, it is determined that the product to be detected is a qualified product. By using the camera to obtain the identification codes on the bottle body and the packaging box of the product to be detected and using the deep learning network model for identification and verification, it is effectively ensured that the identification codes meet the preset requirements. This solution can automatically determine whether both the identification code on the bottle body and the identification code on the packaging box meet the preset requirements, and further verify the association relationship between the two through the pre-set associated information, so as to accurately determine the qualification of the product. This method not only improves the efficiency and accuracy of product detection, but also greatly reduces the errors and costs of manual detection, and can achieve systematic and large-scale detection.

[0054] In an embodiment of the present application, the method further includes: if it is determined through the deep learning network model that the identification code on the bottle body of the product to be detected or the identification code on the packaging box of the product to be detected does not meet the preset requirements, it is determined that the product to be detected is a non-conforming product and is removed.

[0055] Specifically, if it is determined through the deep learning network model that the identification code on the bottle body of the product to be detected or the identification code on the packaging box of the product to be detected does not meet the preset requirements, it is determined that the product to be detected is a non-conforming product and is removed. For example, not meeting the preset requirements may be one or more of the identification code being skewed, the identification code being incomplete, the position of the identification code exceeding the preset range, the identification code being repeated, and the identification code not existing.

[0056] In an embodiment of the present application, the method further includes: if the association relationship between the identification code on the bottle body and the identification code on the packaging box of the product to be detected does not hold, it is determined that the product to be detected is a non-conforming product and is removed.

[0057] Specifically, if the association relationship between the identification code on the bottle body and the identification code on the packaging box of the product to be detected holds, it is determined that the product to be detected is a qualified product. If a product with an incorrect detected association relationship, that is, a product that does not meet the association relationship, can be regarded as a non-conforming product, removed and an alarm is given.

[0058] Embodiments of the present application verify the preset requirements of the bottle body identification code and the packaging box identification code through a deep learning network model. Once any non-conformities are found, such as the identification code being skewed, incomplete, having a position deviation, being repeated or missing, etc., the product is determined to be unqualified and immediately removed. At the same time, the correct association relationship between the bottle body identification code and the packaging box identification code is ensured. Any product with an inconsistent association is also regarded as unqualified and removed, and the alarm mechanism is triggered, significantly improving the strictness and accuracy of product detection.

[0059] In an embodiment of the present application, the identification code includes a first two-dimensional code and a first inkjet code on the bottle body, and a second two-dimensional code and a second inkjet code on the packaging box; the identification code is input into the deep learning network model. If it is determined through the deep learning network model that both the bottle body identification code and the packaging box identification code of the product to be detected meet the preset requirements, it is determined that the product to be detected is a first product, including: inputting the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code into the deep learning network model respectively. If it is determined through the deep learning network model that the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code all meet the preset requirements, it is determined that the product to be detected is a first product.

[0060] Specifically, both the bottle body and the packaging box are printed with a two-dimensional code and an inkjet code. The inkjet code consists of several lines of numbers, and the two-dimensional code is a black and white rectangular matrix. The identification codes on the bottle body and the packaging box include a first two-dimensional code and a first inkjet code on the bottle body, and a second two-dimensional code and a second inkjet code on the packaging box.

[0061] Input the identification code into the deep learning network model. If it is determined through the deep learning network model that both the bottle body identification code and the packaging box identification code of the product to be detected meet the preset requirements, it is determined that the product to be detected is a first product, including: inputting the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code into the deep learning network model respectively. If it is determined through the deep learning network model that the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code all meet the preset requirements, it is determined that the product to be detected is a first product. Specifically, for a product to be detected, input the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code on the bottle body and the packaging box into the deep learning network model respectively. If it is determined through the deep learning network model that the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code all meet the preset requirements, it is determined that the product to be detected is a first product.

[0062] In an embodiment of the present application, the associated information includes: the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code; determining that the product to be detected is a qualified product if the bottle body identification code and the packaging box identification code of the product to be detected conform to the association relationship includes: determining that the product to be detected is a qualified product if the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code all conform to the pre-set associated information.

[0063] Specifically, the associated information includes: the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code. It should be noted that the associated information may also include the association relationship between the first two-dimensional code and the second two-dimensional code, and only one of the association relationship between the first two-dimensional code and the second two-dimensional code or the association relationship between the first inkjet code and the second inkjet code is required. Determining that the product to be detected is a qualified product if the bottle body identification code and the packaging box identification code of the product to be detected conform to the association relationship includes: determining that the product to be detected is a qualified product if the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code all conform to the pre-set associated information. It may also be that if the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first two-dimensional code and the second two-dimensional code all conform to the pre-set associated information, then it is determined that the product to be detected is a qualified product.

[0064] The embodiment of the present application not only performs inkjet code recognition on wine boxes and wine bottles, compares them after deep learning, but also pulls data from the database of the previous station by identifying two-dimensional codes for associated data comparison. On the one hand, it is the comparison of inkjet codes and two-dimensional code recognition, and on the other hand, it is the comparison of the association of inkjet code and two-dimensional code data, with stronger logic and more accurate comparison.

[0065] In an embodiment of the present application, the identification code is input into the deep learning network model. If it is determined through the deep learning network model that both the bottle body identification code and the packaging box identification code of the product to be detected meet the preset requirements, then it is determined that the product to be detected is a first product, including: obtaining a plurality of identification code samples and corresponding annotation files to construct a data set, where the annotation file represents the annotation information corresponding to the identification code, and the annotation information includes that the identification code is normal or the identification code is abnormal; inputting the data set into a convolutional neural network to generate a deep learning network model; inputting the identification code of the product to be detected into the deep learning network model, and determining the annotation information of the bottle body identification code and the packaging box identification code through the deep learning network model; if the annotation information of the bottle body identification code and the annotation information of the packaging box identification code are both normal, then it is determined that the product to be detected is a first product.

[0066] Specifically, first, a dataset is constructed by collecting a large number of identification code samples and their corresponding annotation files. These annotation files record in detail whether each identification code is normal or abnormal, providing the necessary supervision information for subsequent model training. Then, this dataset is used to train a convolutional neural network (CNN) to generate a deep learning network model. This model can learn the features of the identification code and accordingly determine whether a new identification code sample is normal or abnormal. Then, in actual applications, the bottle identification code and the packaging box identification code of the product to be detected are respectively input into this deep learning network model. The model will quickly analyze these identification codes and output their respective annotation information, that is, determine whether these identification codes are normal or abnormal. Finally, if the annotation information of both the bottle identification code and the packaging box identification code is determined to be normal, then it can be determined that this product to be detected is a first product that meets specific standards or requirements.

[0067] Specifically, in an embodiment of the present application, an abnormal identification code includes one or more of the following: the identification code is skewed, the identification code is incomplete, the position of the identification code exceeds a preset range, the identification code is repeated, or the identification code does not exist.

[0068] In an embodiment of the present application, it further includes: setting different alarm sounds for different reasons for abnormal identification codes. In case the spraying content of multiple bottles is the same, the laser spraying or even the production line should be interrupted.

[0069] The embodiments of the present application utilize the high efficiency and accuracy of deep learning to achieve rapid detection of product identification codes, greatly improving the efficiency and accuracy of product detection.

[0070] On the basis of the above embodiments, and in combination with Figure 2 , in a specific embodiment of the present application, the method provided by the present application includes the following steps:

[0071] 1. If the spraying content of two adjacent bottles or multiple bottles is the same, it is regarded as unqualified, and the bottles are removed and an alarm is given;

[0072] 2. Using the data obtained from the host computer, compare whether the spraying data and the two-dimensional code are correctly associated. Products with incorrect association are regarded as unqualified products, removed, and an alarm is given;

[0073] 3. The system needs to identify whether the spraying is skewed. Products that exceed the allowed range are regarded as unqualified products, removed, and an alarm is given;

[0074] 4. For products removed due to different reasons, the alarm sounds should be different. In case the spraying content of multiple bottles is the same, the laser spraying or even the production line should be interrupted.

[0075] According to the specific embodiments provided by the present application, the technical solutions provided by the present application may have the following advantages:

[0076] (1) directly replaces the original method of manually holding a barcode scanner for detection, achieving system detection and large-scale detection;

[0077] (2) Utilizes the efficiency and accuracy of deep learning to achieve rapid detection of product identification codes, greatly improving the efficiency and accuracy of product detection;

[0078] (3) Not only identifies the inkjet codes on wine boxes and wine bottles, conducts comparison after deep learning, but also pulls data from the database of the previous station by identifying QR codes for associated data comparison. On the one hand, it is the comparison of inkjet codes and QR code recognition, and on the other hand, it is the comparison of associated data of inkjet codes and QR codes, with stronger logic and more accurate comparison.

[0079] It can be understood that implementing any method or product of this application does not necessarily need to achieve all the above-mentioned advantages simultaneously.

[0080] Figure 3 The structural schematic diagram of a product detection device based on artificial intelligence provided by an embodiment of this application is as Figure 3 shown, and the device may include:

[0081] Acquisition module 301: used to respectively obtain the identification codes on the bottle body and packaging box of the product to be detected through a camera;

[0082] Recognition module 302: used to input the identification code into a deep learning network model. If it is determined through the deep learning network model that both the bottle body identification code and the packaging box identification code of the product to be detected meet the preset requirements, then determine that the product to be detected is a first product;

[0083] Obtaining module 303: used to obtain pre-set associated information if it is determined that the product to be detected is a first product, and the associated information represents the association relationship between the bottle body identification code and the packaging box identification code;

[0084] Determination module 304: used to determine that the product to be detected is a qualified product if the bottle body identification code and the packaging box identification code of the product to be detected meet the association relationship.

[0085] In an embodiment of this application, it further includes: if it is determined through the deep learning network model that the bottle body identification code of the product to be detected or the packaging box identification code of the product to be detected does not meet the preset requirements, then determine that the product to be detected is an unqualified product and reject it.

[0086] In an embodiment of this application, it further includes: if the bottle body identification code and the packaging box identification code of the product to be detected do not meet the association relationship, then determine that the product to be detected is an unqualified product and reject it.

[0087] In an embodiment of the present application, the identification code includes a first two-dimensional code and a first inkjet code located on the bottle body, and a second two-dimensional code and a second inkjet code located on the packaging box;

[0088] The step of inputting the identification code into the deep learning network model, and if it is determined through the deep learning network model that both the bottle body identification code and the packaging box identification code of the product to be detected meet the preset requirements, then determining that the product to be detected is a first product, includes:

[0089] Input the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code into the deep learning network model respectively. If it is determined through the deep learning network model that the first two-dimensional code, the first inkjet code, the second two-dimensional code, and the second inkjet code all meet the preset requirements, then determine that the product to be detected is a first product.

[0090] In an embodiment of the present application, the step of inputting the identification code into the deep learning network model, and if it is determined through the deep learning network model that both the bottle body identification code and the packaging box identification code of the product to be detected meet the preset requirements, then determining that the product to be detected is a first product, includes:

[0091] Obtain multiple identification code samples and corresponding annotation files to construct a data set, where the annotation file represents the annotation information corresponding to the identification code, and the annotation information includes that the identification code is normal or the identification code is abnormal;

[0092] Input the data set into a convolutional neural network to generate a deep learning network model;

[0093] Input the identification code of the product to be detected into the deep learning network model, and determine the annotation information of the bottle body identification code and the packaging box identification code through the deep learning network model;

[0094] If the annotation information of the bottle body identification code and the annotation information of the packaging box identification code are both normal, then determine that the product to be detected is a first product.

[0095] In an embodiment of the present application, the association information includes:

[0096] The association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code;

[0097] The step of determining that the product to be detected is a qualified product if the bottle body identification code and the packaging box identification code of the product to be detected meet the association relationship includes:

[0098] If the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code all conform to the pre-set association information, it is determined that the product to be detected is a qualified product.

[0099] In an embodiment of the present application, the identification code is abnormal, including:

[0100] One or more of the identification code being skewed, the identification code being incomplete, the position of the identification code exceeding the preset range, the identification code being repeated, and the identification code not existing.

[0101] According to the specific embodiments provided by the present application, the technical solutions provided by the present application may have the following advantages:

[0102] (1) It directly replaces the original method of manually holding a barcode scanner for detection, realizing system detection and large-scale detection;

[0103] (2) Utilizing the efficiency and accuracy of deep learning, it realizes the rapid detection of product identification codes, greatly improving the efficiency and accuracy of product detection;

[0104] (3) It not only identifies the inkjet codes on wine boxes and wine bottles, conducts comparison after deep learning, but also pulls data from the database of the previous station by identifying two-dimensional codes for associated data comparison. On the one hand, it is the comparison of inkjet codes and two-dimensional codes, and on the other hand, it is the comparison of the associated data of inkjet codes and two-dimensional codes, with stronger logic and more accurate comparison.

[0105] For the same or similar parts among the above embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the description of the method embodiment.

[0106] It should be noted that the use of user data may be involved in the embodiments of the present application. In actual applications, user-specific personal data can be used in the solutions described herein within the scope permitted by applicable laws and regulations (such as when the user clearly consents, is effectively notified, and clearly authorizes, etc.) in compliance with the applicable laws and regulations of the country where it is located.

[0107] According to the embodiments of the present application, the present application also provides a computer device and a computer-readable storage medium.

[0108] Such as Figure 4As shown, it is a block diagram of a computer device according to an embodiment of the present application. The computer device is intended to represent various forms of digital computers or mobile devices. Among them, digital computers can include desktop computers, portable computers, workstations, personal digital assistants, servers, mainframe computers, and other suitable computers. Mobile devices can include tablet computers, smart phones, wearable devices, etc.

[0109] As Figure 4 shown, device 400 includes a computing unit 401, a ROM 402, a RAM 403, a bus 404, and an input / output (I / O) interface 405. The computing unit 401, the ROM 402, and the RAM 403 are connected to each other through the bus 404. The input / output (I / O) interface 405 is also connected to the bus 404.

[0110] The computing unit 401 can execute various processes in the method embodiments of the present application according to computer instructions stored in the read-only memory (ROM) 402 or computer instructions loaded from the storage unit 408 into the random access memory (RAM) 403. The computing unit 401 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. The computing unit 401 can include, but is not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. In some embodiments, the method provided by the embodiments of the present application can be implemented as a computer software program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 408.

[0111] The RAM 403 can also store various programs and data required for the operation of the device 400. Part or all of the computer program can be loaded and / or installed onto the device 400 via the ROM 802 and / or the communication unit 409.

[0112] The input unit 406, the output unit 407, the storage unit 408, and the communication unit 409 in the device 400 can be connected to the I / O interface 405. Among them, the input unit 406 can be, for example, a keyboard, a mouse, a touch screen, a microphone, etc.; the output unit 407 can be, for example, a display, a speaker, an indicator light, etc. The device 400 can exchange information, data, etc. with other devices through the communication unit 409.

[0113] It should be noted that the device may also include other components necessary for normal operation. It may also only include the components necessary to implement the solution of the present application, and does not necessarily include all the components shown in the figure.

[0114] The various embodiments of the systems and techniques described herein can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof.

[0115] The computer instructions for implementing the methods of the present application can be written in any combination of one or more programming languages. These computer instructions can be provided to the computing unit 401 such that when the computer instructions are executed by the computing unit 401, such as a processor, the steps involved in the method embodiments of the present application are executed.

[0116] The computer-readable storage medium provided by the present application can be a tangible medium that can contain or store computer instructions for executing the steps involved in the method embodiments of the present application. The computer-readable storage medium can include, but is not limited to, storage media in the forms of electronic, magnetic, optical, electromagnetic, etc.

[0117] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present application shall be included within the protection scope of the present application.

Claims

1. A product detection method based on artificial intelligence, characterized in that: The method includes: Use the camera to obtain the identification codes on the bottle and packaging box of the product to be tested; Inputting the identification code into a deep learning network model, and if the deep learning network model determines that both the bottle identification code and the packaging box identification code of the product to be detected meet preset requirements, determining that the product to be detected is a first product; If it is determined that the product to be detected is the first product, then obtaining pre-set association information, wherein the association information indicates an association relationship between the bottle identification code and the packaging box identification code; If the bottle identification code and the packaging box identification code of the product to be tested conform to the association relationship, it is determined that the product to be tested is a qualified product.

2. The product detection method based on artificial intelligence according to claim 1, characterized in that: The method further comprises: If it is determined through the deep learning network model that the bottle identification code of the product to be tested or the packaging box identification code of the product to be tested does not meet the preset requirements, the product to be tested is determined to be an unqualified product and is discarded.

3. The product detection method based on artificial intelligence according to claim 1, characterized in that: The method further comprises: If the bottle identification code and the packaging box identification code of the product to be tested do not conform to the association relationship, the product to be tested is determined to be an unqualified product and is discarded.

4. The product detection method based on artificial intelligence according to claim 1, characterized in that: The identification code includes a first two-dimensional code and a first spray code located on the bottle body, and a second two-dimensional code and a second spray code located on the packaging box; The step of inputting the identification code into the deep learning network model, and determining that the product to be detected is the first product if the bottle identification code and the packaging box identification code of the product to be detected are determined by the deep learning network model to meet preset requirements, comprises: The first two-dimensional code, the first inkjet code, the second two-dimensional code and the second inkjet code are respectively input into the deep learning network model. If the deep learning network model determines that the first two-dimensional code, the first inkjet code, the second two-dimensional code and the second inkjet code all meet the preset requirements, the product to be inspected is determined to be the first product.

5. The product detection method based on artificial intelligence according to claim 1, characterized in that: The step of inputting the identification code into the deep learning network model, and determining that the product to be detected is the first product if the bottle identification code and the packaging box identification code of the product to be detected are determined by the deep learning network model to meet preset requirements, comprises: Acquire multiple identification code samples and corresponding annotation files to construct a data set, wherein the annotation file represents annotation information corresponding to the identification code, and the annotation information includes whether the identification code is normal or the identification code is abnormal; Inputting the data set into a convolutional neural network to generate a deep learning network model; Inputting the identification code of the product to be tested into the deep learning network model, and determining the labeling information of the bottle identification code and the packaging box identification code through the deep learning network model; If the labeling information of the bottle identification code and the labeling information of the packaging box identification code are both normal, it is determined that the product to be detected is the first product.

6. The product detection method based on artificial intelligence according to claim 4, characterized in that: The associated information includes: The association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code; If the bottle identification code and the packaging box identification code of the product to be detected meet the association relationship, determining that the product to be detected is a qualified product includes: If the association relationship between the first two-dimensional code and the first inkjet code, the association relationship between the second two-dimensional code and the second inkjet code, and the association relationship between the first inkjet code and the second inkjet code all conform to the preset association information, the product to be tested is determined to be a qualified product.

7. The product detection method based on artificial intelligence according to claim 5, characterized in that: The identification code is abnormal, including: One or more of the following: the identification code is skewed, the identification code is incomplete, the identification code position exceeds a preset range, the identification code is repeated, and the identification code does not exist.

8. A product detection device based on artificial intelligence, characterized in that: The device includes: Acquisition module: used to obtain the identification codes on the bottle and packaging box of the product to be tested through the camera; Identification module: used for inputting the identification code into the deep learning network model, and if it is determined through the deep learning network model that the bottle identification code and the packaging box identification code of the product to be detected meet the preset requirements, then determining that the product to be detected is the first product; An acquisition module: used for acquiring preset association information if it is determined that the product to be detected is the first product, wherein the association information indicates the association relationship between the bottle identification code and the packaging box identification code; Determination module: used for determining that the product to be detected is a qualified product if the bottle identification code and the packaging box identification code of the product to be detected meet the association relationship.

9. A computer device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores computer instructions that can be executed by the at least one processor, and the computer instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer instructions stored thereon, characterized in that: The computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 7.

Images