Image review method and related apparatus

Abstract

This application proposes an image review method and related apparatus, relating to the field of image processing technology. The method includes: performing a first review process on an acquired first image to be reviewed; if the first image to be reviewed passes review, acquiring target attribute information of a target product in the first image to be reviewed; determining a first review item corresponding to a second image to be reviewed based on the target attribute information; wherein the second image to be reviewed includes a product image of the target product; the first review item refers to a review item used to review the target product; and performing the first review process on the acquired second image to be reviewed based on the first review item to obtain a review result. The technical solution provided by this application can solve the problem in the prior art of how to obtain valid and usable product images and product packaging images in the scenario of applying for national subsidies.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and more specifically, to an image review method and related apparatus. Background Technology

[0002] In recent years, in order to promote consumption upgrading and green energy conservation, the government has introduced subsidy policies for items such as mobile phones, tablets, and laptops.

[0003] According to current subsidy application requirements, merchants must provide the review agency with a complete chain of transaction evidence. This chain of evidence must include at least proof that the goods have actually been delivered to the consumer. Specifically, this proof can be images of the goods and their packaging, which must include the information required for the application, such as the product's serial number (SN). However, obtaining valid and usable images of the goods and their packaging has become a pressing technical problem that needs to be solved. Summary of the Invention

[0004] Based on the aforementioned defects and shortcomings of the prior art, this application proposes an image review method and related apparatus, which can solve the problem of how to obtain effective and usable product images and product packaging images in the national subsidy application scenario in the prior art.

[0005] According to a first aspect of the embodiments of this application, an image review method is provided, the method comprising: The first review process is executed on the first image to be reviewed; wherein the first image to be reviewed is the outer packaging image of the target product collected in the logistics process, and the outer packaging image includes the coding information of the target product; If the first image to be reviewed passes the review, the target attribute information of the target product in the first image to be reviewed is obtained; wherein, the target attribute information includes at least one of the following: the brand information of the target product and the type information of the target product; The first review item corresponding to the second image to be reviewed is determined based on the target attribute information; wherein, the second image to be reviewed is an image of the target product obtained after the first image to be reviewed has passed the review, and the second image to be reviewed includes a product image of the target product; the first review item refers to the review item used to review the target product; The first review process is executed on the second image to be reviewed based on the first review item to obtain the review result.

[0006] According to a second aspect of the embodiments of this application, an image review apparatus is provided, the apparatus comprising: The first review module is used to perform a first review process on the first image to be reviewed; wherein, the first image to be reviewed is the outer packaging image of the target product collected in the logistics process, and the outer packaging image includes the coding information of the target product; The attribute acquisition module is used to acquire target attribute information of the target product in the first image to be reviewed, provided that the first image to be reviewed has passed the review; wherein, the target attribute information includes at least one of the following: brand information of the target product and type information of the target product; The determining module is used to determine the first review item corresponding to the second image to be reviewed based on the target attribute information; wherein, the second image to be reviewed is an image of the target product obtained after the first image to be reviewed has passed the review, and the second image to be reviewed includes a product image of the target product; the first review item refers to the review item used to review the target product; The second review module is used to execute the first review process on the obtained second image to be reviewed according to the first review item, and obtain the review result.

[0007] According to a third aspect of the embodiments of this application, an electronic device is provided, including: a memory and a processor; The memory is connected to the processor and is used to store programs; The processor is used to implement the image review method as described in the first aspect by running the program in the memory.

[0008] According to a fourth aspect of the embodiments of this application, a storage medium is provided, on which a computer program is stored, and when the computer program is run by a processor, it implements the image review method as described in the first aspect.

[0009] According to a fifth aspect of the embodiments of this application, a computer program product or a computer program is provided, the computer program product including the computer program, wherein when a processor executes the computer program, it implements the steps in the image review method as described in the first aspect.

[0010] The technical solution provided in this application allows for the collection of images of goods and their outer packaging during the logistics process. These images are then reviewed based on the national subsidy application review criteria, facilitating the selection of valid and usable image credentials. Regarding the review of goods images, since the national subsidy application review criteria may differ for different brands and types of goods, this application can first review the images of the goods' outer packaging. If the outer packaging image passes review, the target attribute information of the goods (such as brand information, type information, etc.) is obtained from the outer packaging image. Then, based on this target attribute information, the first review item corresponding to the goods image is determined. Subsequently, the goods image can be reviewed based on this first review item. This enables precise review of goods images and helps obtain more valid and usable image credentials. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0012] Figure 1 This is a flowchart illustrating an image review method provided in an embodiment of this application.

[0013] Figure 2 This is a schematic diagram of the review process provided for an embodiment of this application.

[0014] Figure 3 This is one of the flowcharts illustrating an evidence collection process provided in this application embodiment.

[0015] Figure 4 This is a second flowchart illustrating an evidence collection process provided in an embodiment of this application.

[0016] Figure 5 This is a block diagram of an image review device provided in an embodiment of this application.

[0017] Figure 6 This is a schematic diagram of the structure of an electronic 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, and 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] Application Overview In recent years, to promote consumption upgrading and green energy conservation, the government has introduced subsidy policies for electronic products such as mobile phones, tablets, and laptops. Consumers can also enjoy preferential prices when purchasing these products through online channels. During the implementation of this policy, merchants typically advance the subsidy amount, allowing consumers to complete the transaction directly at the subsidized price. Merchants then apply to the relevant government departments for the remaining subsidy funds.

[0020] According to current subsidy application requirements, merchants must provide a complete chain of transaction evidence to the review agency to prove the authenticity and compliance of their sales activities. This chain of evidence generally includes: online sales records, goods leaving the warehouse and logistics records, and vouchers proving that the goods have actually been delivered to consumers. Sales records and warehouse records are relatively easy for merchants to obtain and provide because they rely on the data retention of e-commerce platforms and their logistics systems. Valid evidence of delivery to consumers can be images of the goods and their packaging, which must include information required for the national subsidy application, such as the product's serial number (SN). However, obtaining valid and usable images of the goods and their packaging has become a pressing technical problem that needs to be solved.

[0021] To address this, this application provides a solution that allows for the collection of images of goods and their packaging during the logistics process. These images are then reviewed based on the national subsidy application requirements, facilitating the selection of valid and usable image credentials based on the review results. Details are as follows.

[0022] Exemplary methods This application also provides an image review method applied to an electronic device. The electronic device can be a server, such as a cloud server, or a terminal device, such as a mobile phone or computer. The electronic device is equipped with a system for image review.

[0023] The method is described in detail below through some embodiments. The following embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0024] like Figure 1 As shown, the image review method may include steps 101 to 103, as described below.

[0025] Step 101: Perform a first review process on the acquired first image to be reviewed.

[0026] The first image to be reviewed here is the outer packaging image of the target commodity collected in the logistics link. The outer packaging image includes the review content during national subsidy application, such as the coding information of the target commodity (SN code, International Mobile Equipment Identity (IMEI), etc.). This first image to be reviewed is one of the vouchers (hereinafter referred to as delivery vouchers) for proving that the target commodity sold by the merchant has been delivered to the user.

[0027] The target commodity described here can be a commodity purchased by a consumer on an e-commerce platform, such as electronic products like mobile phones, tablet computers, and laptop computers.

[0028] The logistics link described here can be the link where the courier delivers the express package of the target commodity to the user. After obtaining the user's consent, the courier can use a first terminal (such as a mobile phone) to collect relevant images of the target commodity enjoying the subsidy policy in the express package, such as the first image to be reviewed.

[0029] Images taken casually or images not taken in accordance with the subsidy application requirements may not be used as vouchers for applying for national subsidies. Therefore, in order to obtain effective and usable image vouchers, an image review system is set in the electronic device in the embodiments of the present application, and the first review process can be performed on the effectiveness of the collected images through this image review system to determine whether the images can be used as vouchers for applying for national subsidies.

[0030] For the effectiveness review of images, it can at least include reviewing whether the image content meets the subsidy application requirements. For example, the voucher images required for subsidy application need to include: the outer packaging of the commodity and the product serial number on the outer packaging. If the image content in the image does not meet this requirement, it means that the image is an invalid image; on the contrary, if the image content in the image meets this requirement, it means that the image is a valid image. It should be noted that this is only an example, and the specific subsidy application requirements shall be subject to the actual situation.

[0031] For the first image to be reviewed, it can be determined whether it is valid through the first review process. Specifically, it can be determined whether: the first image to be reviewed includes the outer packaging, whether the outer packaging is complete, whether there is the coding information of the target commodity on the outer packaging, and whether the coding information is clear, etc.

[0032] Step 102: When the first image to be reviewed passes the review, obtain the target attribute information of the target commodity in the first image to be reviewed.

[0033] The target attribute information mentioned herein may include, but is not limited to, at least one of the following: brand information of the target product, type information of the target product.

[0034] Step 103: Determine the first review item corresponding to the second image to be reviewed based on the target attribute information.

[0035] The second image to be reviewed mentioned here is an image of the target product obtained after the first image to be reviewed has passed the review. It is also one of the credentials used to prove that the target product sold by the merchant has been delivered to the user.

[0036] The second image to be reviewed includes an image of the target product. Specifically, the second image to be reviewed may include: a first image and a second image, wherein the first image is an image of the target product, and the second image is a combined image of the target product and its outer packaging.

[0037] For example, the courier can first collect the first image to be reviewed, and then upload it to the image review system for the first review process; after the first image to be reviewed passes the first review process, the courier can then collect the first image and upload it to the image review system for the first review process; after the first image passes the first review process, the courier can then collect the second image and upload it to the image review system for review.

[0038] The first review item mentioned here refers to the review item used to review the target product.

[0039] Step 104: Perform the first review process on the obtained second image to be reviewed according to the first review item, and obtain the review result.

[0040] The verification items for delivery documents may differ for different types of goods. For example, for different types of electronic products, such as mobile phones and tablets, the former requires verification of whether the IMEI code is included, compared to the latter. Similarly, for electronic products of the same type but different brands, such as brand A mobile phones and brand B mobile phones, the former requires verification of whether the phone is connected to the internet (connection can be determined based on network identifiers, such as Wi-Fi or mobile network identifiers). These differences in verification items generally arise from the second image to be verified, while some attribute information of the product (such as type and brand) can be obtained from the image of the product's outer packaging.

[0041] Therefore, in this embodiment, different review items for different products can be pre-set. For example, review items can be configured according to the merchant's requirements, and a correspondence between review items and product attributes can be established. After the first image to be reviewed (i.e., the product packaging image) passes the review, the packaging information in the first image to be reviewed can be identified, such as through character information recognition or trademark image recognition, to determine the target attribute information of the target product. Then, based on the target attribute information, the corresponding first review item can be determined. In this way, the validity of the second image to be reviewed can be accurately reviewed based on the first review item, improving review efficiency and the reliability of review results, so as to select more effective and usable image credentials.

[0042] Furthermore, after the courier delivers the package to the user with the user's consent, images of the target goods purchased online and eligible for subsidies are collected from the package. This provides merchants with visual evidence of the goods reaching the user at the delivery site, offering them intuitive and traceable delivery proof. By reviewing the collected images, it can be ensured that they meet the formal and substantive requirements for evidence (such as clearly showing the goods), effectively filtering out invalid or non-compliant images, improving the reliability of the evidence collection process, and reducing the risk of review failure due to flawed evidence.

[0043] The following is based on Figure 2 Taking an example, the above embodiments will be further explained and illustrated.

[0044] First, customized configurations for verification items can be made for merchants.

[0045] Secondly, consumers place orders with merchants for eligible government-subsidized goods. Merchants then package the goods according to the order information and ship them to their destination via courier. After receiving the package, the courier delivers it to the customer. With the customer's consent, a delivery receipt is obtained, as described below: First, photograph the outer packaging of the target product to obtain Evidence 1 (corresponding to the first image to be reviewed). Then, review Evidence 1 according to the preset review items and call the first review process in the image review system. If the review passes, remind the courier to collect Evidence 2 (corresponding to the first image in the second set of images to be reviewed); if the review fails, remind the courier to re-collect Evidence 1 and re-review it. After completing the final review, the review process for Evidence 1 ends.

[0046] After obtaining Evidence 1, a classification model can be invoked to determine the target attribute information of the target product based on the packaging information, and to determine the first review item for reviewing Evidence 2 and Evidence 3 (corresponding to the second image in the second image to be reviewed) based on the target attribute information. Optionally, the classification model can be a pre-trained qwen2.5-vl-3B model.

[0047] After the final review of Evidence Collection 1 is completed, the courier takes a photo of the target product to obtain Evidence Collection 2. Evidence Collection 2 is then reviewed according to the first review item and the first review process. If the review passes, the courier is reminded to collect Evidence Collection 3; if the review fails, the courier is reminded to re-collect Evidence Collection 2 and re-review it. The review process for Evidence Collection 2 ends after the final review is completed.

[0048] After the final review of Evidence Collection 2 is completed, the courier takes photos of the target product and its outer packaging to obtain Evidence Collection 3. Following the preset review items of Evidence Collection 1 and the first review item of Evidence Collection 2, the first review process is invoked to review Evidence Collection 3. If the review passes, the courier is notified to complete the review; if the review fails, the courier is notified to re-collect Evidence Collection 3 and re-review it. After the final review is completed, the review process for Evidence Collection 3 ends.

[0049] Optionally, in this embodiment of the application, it can be determined whether to send the image to be reviewed to the merchant based on the review result of the first review process of the image to be reviewed (such as the first image to be reviewed and the second image to be reviewed).

[0050] For example, if the review result indicates that the review has passed, the image to be reviewed can be automatically sent to the merchant's end (i.e., the second terminal), enabling the merchant to obtain a valid and approved delivery voucher in a timely manner, thus improving delivery efficiency. Simultaneously, a first notification message can also be sent to the first terminal, indicating that the image to be reviewed has passed the review, so that other images to be reviewed can be obtained.

[0051] If the review result indicates that the review has failed, the image to be reviewed can be retrieved again, and the first review process can be re-executed until the image to be reviewed passes the review or the number of reviews reaches the preset number.

[0052] When the number of times an image to be reviewed reaches a preset number (e.g., 3 times) and the last review result indicates failure, considering the trend of increasingly higher image quality taken by the courier, the last image taken is likely the best quality frame. Therefore, the last received image to be reviewed can be sent back to the merchant. Alternatively, the clearest frame among all received images to be reviewed can be sent to the second terminal. Or, the frame that best meets the review requirements can be sent back to the merchant. For example, if four images (A, B, C, and D) are received for the first review, and the review requirements include 10 criteria, and after review, image A meets 5 criteria, image B meets 8, image C meets 6, and image D meets 7, then image B can be sent back to the merchant.

[0053] Based on this, in this embodiment of the application, the target attribute information of the target product in the first image to be reviewed can be obtained after the final review of the first image to be reviewed is completed. The final review here refers to either passing the review or reaching a preset number of reviews.

[0054] Optionally, if the second image to be reviewed is a product image of the target product (i.e., the first image), the first review process can be used to determine whether the product image includes the target product and whether the target product includes its coding information. If the target product is an electronic product, it can also be determined whether the electronic product in the product image has a lit screen and whether the screen displays an SN code or IMEI code (required when the electronic product is a mobile phone). If the second image to be reviewed is a composite image (i.e., the second image), the review can be conducted by combining the review items of the first image to be reviewed and the first image.

[0055] In some optional embodiments, since the coding information is the unique identifier of the electronic product, when the target commodity is an electronic product, the image information required for applying for national subsidies can at least include the coding information of the electronic product. Therefore, the review item for the delivery certificate can at least include whether the delivery certificate includes the coding information of the electronic product. Specifically, it can be: whether the packaging in the first image to be reviewed includes the coding information of the electronic product; whether the electronic product in the first image in the second image to be reviewed includes the coding information; and whether both the outer packaging and the electronic product in the second image in the second image to be reviewed include the coding information.

[0056] like Figure 3As shown, the review process for encoded information can begin by using Optical Character Recognition (OCR) technology to recognize characters and obtain the recognition results. Then, based on the encoded information of the electronic device provided by the merchant, regular expression matching is performed on the character recognition results to determine if the encoded information is included. If included, the review item passes; otherwise, it fails. Optionally, if the required encoded information is not extracted from the character recognition results using regular expression matching, named entity recognition (NER) technology can be used for encoding information extraction. Optionally, during the extraction of encoded information based on regular expression matching and / or NER technology, a pre-trained model (such as a Visual Language Model, VLM) can also be used to extract encoded information from the target image. The results from the two branches are then fused. For example, if extracting encoded information based on regular expression matching and / or NER technology fails, but extracting encoded information from the image using a pre-trained model succeeds, the model's result is used, and the extracted result is reviewed. Considering the hallucination problem in the model, when the required encoding information is extracted from both branches, the encoding information extracted based on regular expression matching or NER technology is used first.

[0057] After extracting the required encoded information from the character recognition results, the image region containing the encoded information can also be reviewed for sharpness to ensure that the encoded information is clear and usable. For example, a sharpness classification network (such as a classification network based on Mobilenet-v3-large) can be used to accurately determine the sharpness of the encoded information region.

[0058] In some optional embodiments, the first review process and the second review may further include a second review item, which includes at least one of the following: random shooting review, re-shooting review, and clarity review.

[0059] like Figure 2 As shown, the embodiments of this application can also perform at least one of the following checks on the image to be reviewed: random shooting review, photocopying review, and clarity review.

[0060] In the context of the "random shooting" review, "random shooting" refers to taking photos haphazardly, such as randomly photographing non-target objects like floors or walls. Images obtained in this way are obviously unacceptable as proof of delivery. Therefore, in this embodiment, before reviewing whether the images to be reviewed contain the image content required for the national subsidy application, a review of whether the images to be reviewed were taken randomly can be conducted first. In this embodiment, the target object should be the target product itself, its outer packaging, etc.

[0061] In the context of the "reproduction review," reproduction refers to photographing and copying images. For example, a courier might photograph images of the target product or its packaging sent to them by a customer. Reproduced images cannot prove they were taken at the delivery site of the target product; therefore, such images cannot serve as proof of delivery. Thus, in this embodiment, before reviewing whether the images to be reviewed include the image content required for the national subsidy application, a review of whether the images to be reviewed were reproduced can be conducted first.

[0062] Since blurry images can affect the review results, the basis for reviewing whether the image to be reviewed contains the image content required for the national subsidy application is that the image to be reviewed is clear. Therefore, in this embodiment of the application, before reviewing whether the image to be reviewed contains the image content required for the national subsidy application, the clarity of the image to be reviewed can be checked first.

[0063] Optionally, in the first review process, random and clear photos can be reviewed using the mobilenetv3-large network, and photos re-shot can be reviewed using the mobilenetv3-large network. In the second review process, convnext-large can be used for random and clear photo identification, and convnext-large can be used for photo re-shot identification.

[0064] In some alternative embodiments, considering that the review results of the first review process for the image to be reviewed may contain errors—for example, although the review result indicates that the review has passed, the actual image content may not fully meet the review items—the delivery documents can be manually reviewed to reduce review errors. However, manual review is inefficient and easily affected by subjective factors. To overcome this problem, a more precise review process can be used for review, as described below.

[0065] If the image to be reviewed passes the first review process, the method may further include steps B1 to B3, as described below: Step B1: Perform the second review process on the image to be reviewed.

[0066] The image to be reviewed can be at least one of the first image to be reviewed and the second image to be reviewed.

[0067] In this embodiment, the second review process can be performed on only the first image to be reviewed or the second image to be reviewed, or it can be performed on both the first and second images to be reviewed, depending on actual needs. For example, since the target attribute information of the target product needs to be obtained through the first image to be reviewed, the review of the first image to be reviewed can be strengthened, so the second review process can be performed on the first image to be reviewed. As another example, the second image to be reviewed includes a product image of the target product, which is a more important image; therefore, the second review process can be performed on the second image to be reviewed. Yet another example, both the first and second images to be reviewed are images required by the merchant; in order to deliver higher quality images to the merchant, the second review process can be performed on both the first and second images to be reviewed.

[0068] The second review process has a higher accuracy rate but a longer review time compared to the first review process.

[0069] Since delivery personnel need to determine whether to re-collect images based on the review results of the first review model, to reduce their waiting time, the first review process can be implemented using a first review model based on a small network. Such models are characterized by high processing speed. Generally speaking, the first review model can control the review time for one frame of image to within 10 seconds, thus quickly completing the review process.

[0070] For a more precise second-stage review process, a second-stage review model based on a large-scale network can be implemented. Such models are characterized by high processing accuracy. Of course, while the processing accuracy is high, the computational complexity is also higher, so the review time is relatively long (e.g., within 3 minutes). Therefore, the review process can be set as a seamless process, meaning that the courier does not need to wait for the review results and can leave after all the images to be reviewed have completed the first-stage review process. If a prompt for re-collection of evidence is received later due to a failed review, then further negotiation with the user can be held to arrange for on-site evidence collection.

[0071] Step B2: If the image to be reviewed fails the second review process, send a first prompt message to the first terminal.

[0072] The first prompt message is used to remind users to re-capture the image to be reviewed. The first terminal refers to the terminal device used to acquire the image to be reviewed.

[0073] Step B3: If the image to be reviewed passes the second review process, send the image to be reviewed to the second terminal.

[0074] The second terminal refers to the terminal device associated with the user who requests the image to be reviewed.

[0075] In this embodiment of the application, it is also possible to determine whether to send the image to be reviewed to the merchant based on the review result of the second review process of the image to be reviewed.

[0076] In this embodiment, the review of the image to be reviewed can be completed quickly through the first review process. To reduce review errors, a second, more precise review process can be used to further verify the image, such as... Figure 4 As shown. If the verification fails, the courier will be reminded to re-collect the target image to further improve the reliability of the delivery receipt.

[0077] Optionally, step B1: Performing a second review process on the image to be reviewed may include: Step B11: Identify the encoded information in the first image to be reviewed using optical character recognition algorithm and regular expression matching algorithm.

[0078] In the second review process, the OCR results can be output by combining the OCR-server version and the OCR-mobile version.

[0079] Step B12: If the coded information in the first image to be reviewed cannot be identified by the optical character recognition algorithm and the regular expression matching algorithm, the coded information in the first image to be reviewed is identified by the named entity recognition algorithm.

[0080] Step B13: If the named entity recognition algorithm fails to identify the encoded information in the first image to be reviewed, the encoded information in the first image to be reviewed is identified by multiple pre-trained large models.

[0081] In the second review process, BERT-large can be used to extract encoded information.

[0082] Step B14: Use the same recognition results from multiple large models as the recognition results of the encoded information.

[0083] In the second review process, the review of coded information can begin with OCR algorithm-based character recognition to obtain the results. Then, based on the coded information of the electronic device provided by the merchant, regular expression matching is performed on the coded information to determine whether it is included in the result. If it is, the review item passes; otherwise, it fails.

[0084] If the required encoded information cannot be extracted from the character recognition results using OCR and regular expression matching algorithms, the named entity recognition (NER) technology can also be used to extract the encoded information.

[0085] When NER technology fails to identify and extract the required encoded information, multiple pre-trained large models can be used to identify the encoded information in the image to be reviewed. For example, a VLM large model based on retrained qwen2.5-vl-72B can be used to identify the encoded information, and then commercial large models gemini and chat-gpt-4o can be used simultaneously for identification. The results of the three large models are then fused using a voting strategy: if more than one model identifies the same encoded information, that identification result is used as the final identification result.

[0086] After obtaining the encoded information, the image region containing the encoded information can be reviewed for sharpness to ensure that the encoded information is clear and usable. In the second review process, a transformer-based convnext-large network can be used to identify the sharpness of the encoded information region.

[0087] It should be noted that the use of the three main models is not limited to the recognition of coded information, but also applicable to the extraction of information required for other audit items. The result fusion strategy is similar to that described above.

[0088] In some optional embodiments, since the target attribute information of the target product needs to be obtained through the first image to be reviewed, the review of the first image to be reviewed can be strengthened. Therefore, the method may further include: performing a second review process on the first image to be reviewed; if the first image to be reviewed passes the second review process, then performing the step of obtaining the target attribute information of the target product in the first image to be reviewed. If the first image to be reviewed fails the second review process, a first prompt message is sent to the first terminal. The first prompt message is used to remind the user to re-acquire the first image to be reviewed. The first terminal refers to a terminal device used to acquire the first image to be reviewed.

[0089] In some alternative embodiments, during the execution of the first or second review process on the second image, the method may further include steps C1 and C2, as described below: Step C1: Using a preset segmentation algorithm, identify the target product and its outer packaging in the second image, and segment the second image according to the position of the target product and its outer packaging in the second image to obtain a first sub-image and a second sub-image.

[0090] The first sub-image is an image that includes the target product, and the second sub-image is an image that includes the outer packaging.

[0091] Step C2: Review the first sub-image and the second sub-image respectively.

[0092] Since the second image includes the target product image and its outer packaging image, in order to review the target product image and its outer packaging image separately, it is necessary to use a preset segmentation algorithm to identify the target product and its outer packaging in the second image, and then segment the second image to obtain the target product image and the outer packaging image, thereby achieving more accurate review.

[0093] Optionally, in the first review process, the preset segmentation algorithm can be implemented using a YOLOv11-S network. In the second review process, the preset segmentation algorithm can be implemented using a more accurate YOLOv11-M network.

[0094] In summary, in this embodiment, after the courier delivers the package to the user and obtains the user's consent, images of the target goods purchased online by the user and eligible for subsidies can be collected from the package to obtain image evidence for the merchant to apply for national subsidies. To improve the reliability of the image evidence, this embodiment can also review the image evidence, enhancing the reliability of the evidence collection process and reducing the risk of review due to flawed evidence. Furthermore, to achieve precise review for different goods, this embodiment can determine the target attribute information of the target goods based on the outer packaging information, then determine the corresponding review items based on the target attribute information, and then achieve precise review of the target goods' images based on these review items, improving the reliability of the review results. Further, to improve the accuracy of the review results, this embodiment also uses a more precise review process (i.e., a second review process) to verify the image evidence, further enhancing its reliability.

[0095] Exemplary device Accordingly, this application also provides an image review device applied to an electronic device. The electronic device can be a server, such as a cloud server, or a terminal device, such as a mobile phone or computer. The electronic device is equipped with a system for image review.

[0096] like Figure 5 As shown, the device may include: The first review module 501 is used to perform the first review process on the first image to be reviewed.

[0097] The first image to be reviewed is an image of the outer packaging of the target product collected during the logistics process, and the outer packaging image includes the coding information of the target product.

[0098] The attribute acquisition module 502 is used to acquire the target attribute information of the target product in the first image to be reviewed when the first image to be reviewed has passed the review.

[0099] The target attribute information includes at least one of the following: the brand information of the target product and the type information of the target product.

[0100] The determination module 503 is used to determine the first review item corresponding to the second image to be reviewed based on the target attribute information.

[0101] Wherein, the second image to be reviewed is an image of the target product obtained after the first image to be reviewed has passed the review, and the second image to be reviewed includes the product image of the target product; the first review item refers to the review item used to review the target product.

[0102] The second review module 504 is used to execute the first review process on the obtained second image to be reviewed according to the first review item, and obtain the review result.

[0103] Optionally, if the first image to be reviewed passes the first review process, the device further includes: The third review module is used to perform the second review process on the first image to be reviewed.

[0104] The second review process has a higher accuracy rate but a longer review time compared to the first review process.

[0105] The first sending module is used to send a first prompt message to the first terminal when the first image to be reviewed fails to pass the second review process.

[0106] The first prompt message is used to remind users to re-capture the first image to be reviewed; the first terminal refers to the terminal device used to acquire the first image to be reviewed.

[0107] The second sending module is used to send the first image to be reviewed to the second terminal when the first image to be reviewed passes the second review process.

[0108] The second terminal refers to the terminal device associated with the user who requests the first image to be reviewed.

[0109] Optionally, if the target image to be reviewed passes the first review process, the device further includes: The fourth review module is used to perform a second review process on the target image to be reviewed.

[0110] The target image to be reviewed includes the first image to be reviewed and the second image to be reviewed; the second review process has a higher accuracy rate but a longer review time compared to the first review process.

[0111] The third sending module is used to send a first prompt message to the first terminal when the target image to be reviewed fails to pass the second review process.

[0112] The first prompt message is used to remind the user to re-acquire the target image to be reviewed; the first terminal refers to a terminal device used to acquire the target image to be reviewed.

[0113] The fourth sending module is used to send the target image to be reviewed to the second terminal when the target image to be reviewed has passed the second review process.

[0114] The second terminal refers to a terminal device associated with the user who requests the target image to be reviewed.

[0115] Optionally, the third review module includes: The first recognition unit is used to recognize the encoded information in the first image to be reviewed by using an optical character recognition algorithm and a regular expression matching algorithm.

[0116] The second recognition unit is used to recognize the encoded information in the first image to be reviewed by a named entity recognition algorithm when the encoded information in the first image to be reviewed cannot be recognized by optical character recognition algorithm and regular expression matching algorithm.

[0117] The third recognition unit is used to recognize the encoded information in the first image to be reviewed by using multiple pre-trained large models when the named entity recognition algorithm cannot recognize the encoded information in the first image to be reviewed.

[0118] The recognition result determination unit is used to take the same recognition result in the multiple large models as the recognition result of the encoded information.

[0119] Optionally, the second image to be reviewed includes: a first image and a second image, wherein the first image is a product image of the target product, and the second image is a composite image of the target product and its outer packaging.

[0120] Optionally, during the execution of the first review process or the second review process on the second image, the apparatus may further include: The segmentation module is used to identify the target product and its outer packaging in the second image using a preset segmentation algorithm, and to segment the second image according to the positions of the target product and its outer packaging in the second image to obtain a first sub-image and a second sub-image.

[0121] The first sub-image is an image including the target product, and the second sub-image is an image including the outer packaging.

[0122] The fifth review module is used to review the first sub-image and the second sub-image respectively.

[0123] Optionally, the first review process may further include a second review item, which may include at least one of the following: review of random shooting, review of re-shooting, and review of image clarity.

[0124] Optionally, during the first review process for the first image to be reviewed, the first review module is specifically used to: if the first image to be reviewed fails the review, re-acquire the first image to be reviewed and re-execute the first review process until the first image to be reviewed passes the review or the number of reviews reaches a preset number.

[0125] The image review device provided in this embodiment belongs to the same concept as the image review method provided in the above embodiments of this application. It can execute the image review method provided in any of the above embodiments of this application and has the corresponding functional modules and beneficial effects of the method. Technical details not described in detail in this embodiment can be found in the specific processing content of the image review method provided in the above embodiments of this application, and will not be repeated here.

[0126] It should be understood that the modules in the indicator light display control device described above can be implemented by a processor calling software. For example, the device includes a processor connected to a memory containing instructions. The processor calls the instructions stored in the memory to implement any of the above methods or to realize the functions of each unit in the device. The processor can be a general-purpose processor, such as a CPU or microprocessor, and the memory can be internal or external to the device. Alternatively, the units in the device can be implemented as hardware circuits. By designing the hardware circuits, some or all of the unit functions can be realized. The hardware circuit can be understood as one or more processors. For example, in one implementation, the hardware circuit is an ASIC, and the functions of some or all of the above units are realized by designing the logical relationships between the components within the circuit. In another implementation, the hardware circuit can be implemented using a PLD, such as an FPGA, which can include a large number of logic gates. The connection relationships between the logic gates are configured through a configuration file, thereby realizing the functions of some or all of the above units. All units of the above device can be implemented entirely by a processor calling software, entirely by hardware circuits, or partially by a processor calling software with the remaining parts implemented by hardware circuits.

[0127] In this application embodiment, a processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction reading and execution capabilities, such as a CPU, microprocessor, GPU, or DSP. In another implementation, the processor can implement certain functions through the logical relationships of hardware circuits. These logical relationships are fixed or reconfigurable. For example, the processor may be a hardware circuit implemented as an ASIC or PLD, such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document and configuring the hardware circuit can be understood as the processor loading instructions to implement the functions of some or all of the above units. Furthermore, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as an NPU, TPU, or DPU.

[0128] As can be seen, each unit in the above device can be one or more processors (or processing circuits) configured to implement the above methods, such as: CPU, GPU, NPU, TPU, DPU, microprocessor, DSP, ASIC, FPGA, or a combination of at least two of these processor forms.

[0129] Furthermore, the units in the above devices can be integrated in whole or in part, or they can be implemented independently. In one implementation, these units are integrated together and implemented in the form of a System-on-Chip (SoC). The SoC may include at least one processor for implementing any of the above methods or implementing the functions of the units in the device. The at least one processor may be of different types, such as CPU and FPGA, CPU and artificial intelligence processor, CPU and GPU, etc.

[0130] Exemplary electronic devices This application also provides an electronic device, such as... Figure 6 As shown, the electronic device includes a memory 600 and a processor 610.

[0131] The memory 600 is connected to the processor 610 and is used to store programs.

[0132] The processor 610 is used to implement the image review method in the above embodiments by running the program stored in the memory 600.

[0133] Specifically, the aforementioned electronic device may also include: a communication interface 620, an input device 630, an output device 640, and a bus 650.

[0134] The processor 610, memory 600, communication interface 620, input device 630, and output device 640 are interconnected via a bus. Among them: Bus 650 may include a pathway for transmitting information between various components of a computer system.

[0135] The processor 610 can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0136] The processor 610 may include a main processor, as well as a baseband chip, modem, etc.

[0137] The memory 600 stores a program that executes the technical solution of this invention, and may also store an operating system and other key business functions. Specifically, the program may include program code, which includes computer operation instructions. More specifically, the memory 600 may include read-only memory (ROM), other types of static storage devices capable of storing static information and instructions, random access memory (RAM), other types of dynamic storage devices capable of storing information and instructions, disk storage, flash memory, etc.

[0138] Input device 630 may include a device for receiving user input data and information, such as a keyboard, mouse, camera, scanner, light pen, voice input device, touch screen, pedometer, or gravity sensor.

[0139] Output device 640 may include devices that allow information to be output to a user, such as a display screen, printer, speaker, etc.

[0140] The communication interface 620 may include a device that uses any transceiver to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0141] The processor 610 executes the program stored in the memory 600 and calls other devices, which can be used to implement the various steps of the image review method provided in the above embodiments of this application.

[0142] Exemplary computer program products and storage media In addition to the methods and devices described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the image review method described in the embodiments of this application.

[0143] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0144] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0145] Furthermore, embodiments of this application may also be storage media storing a computer program, which is executed by a processor of the steps in the image review method described in the embodiments of this application.

[0146] In addition, embodiments of this application may also be chips, which include processors and data interfaces. The processor reads instructions stored in the memory through the data interface to execute the steps in the image review method described in the embodiments of this application.

[0147] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0148] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For apparatus embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0149] The steps in the methods of the various embodiments of this application can be adjusted, merged, or deleted in order according to actual needs, and the technical features described in each embodiment can be replaced or combined.

[0150] The modules and sub-modules in the devices and terminals in the various embodiments of this application can be merged, divided, and deleted according to actual needs.

[0151] It should be understood that the disclosed terminals, devices, and methods can be implemented in other ways, given the several embodiments provided in this application. For example, the terminal embodiments described above are merely illustrative. For instance, the division of modules or sub-modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple sub-modules or modules may be combined or integrated into another module, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0152] The modules or submodules described as separate components may or may not be physically separate. The components that constitute a module or submodule may or may not be physical modules or submodules; that is, they may be located in one place or distributed across multiple network modules or submodules. Some or all of the modules or submodules can be selected to achieve the purpose of this embodiment's solution, depending on actual needs.

[0153] Furthermore, the functional modules or sub-modules in the various embodiments of this application can be integrated into one processing module, or each module or sub-module can exist physically separately, or two or more modules or sub-modules can be integrated into one module. The integrated modules or sub-modules described above can be implemented in hardware or in the form of software functional modules or sub-modules.

[0154] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0155] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software unit executed by a processor, or a combination of both. The software unit can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0156] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, 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. Without further limitations, 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 said element.

Claims

1. An image verification method, characterized in that, The method includes: The first review process is executed on the first image to be reviewed; wherein the first image to be reviewed is the outer packaging image of the target product collected in the logistics process, and the outer packaging image includes the coding information of the target product; If the first image to be reviewed passes the review, the target attribute information of the target product in the first image to be reviewed is obtained; wherein, the target attribute information includes at least one of the following: the brand information of the target product and the type information of the target product; The first review item corresponding to the second image to be reviewed is determined based on the target attribute information; wherein, the second image to be reviewed is an image of the target product obtained after the first image to be reviewed has passed the review, and the second image to be reviewed includes a product image of the target product; the first review item refers to the review item used to review the target product; The first review process is executed on the second image to be reviewed based on the first review item to obtain the review result.

2. The image verification method according to claim 1, characterized in that, If the first image to be reviewed passes the first review process, the method further includes: A second review process is executed on the first image to be reviewed; wherein, the second review process has a higher review accuracy but a longer review time compared to the first review process; If the first image to be reviewed fails the second review process, a first prompt message is sent to the first terminal; wherein, the first prompt message is used to remind the user to re-acquire the first image to be reviewed; the first terminal refers to a terminal device used to acquire the first image to be reviewed; If the first image to be reviewed passes the second review process, the first image to be reviewed is sent to the second terminal; wherein, the second terminal refers to the terminal device associated with the user who requests the first image to be reviewed.

3. The image verification method according to claim 1, characterized in that, If the target image to be reviewed passes the first review process, the method further includes: A second review process is executed on the target image to be reviewed; wherein the target image to be reviewed includes the first image to be reviewed and the second image to be reviewed; the second review process has a higher review accuracy but a longer review time compared to the first review process. If the target image to be reviewed fails the second review process, a first prompt message is sent to the first terminal; wherein, the first prompt message is used to remind the user to re-acquire the target image to be reviewed; the first terminal refers to a terminal device used to acquire the target image to be reviewed; If the target image to be reviewed passes the second review process, the target image to be reviewed is sent to a second terminal; wherein, the second terminal refers to a terminal device associated with the user who requests the target image to be reviewed.

4. The image verification method according to claim 2, characterized in that, The second review process for the first image to be reviewed includes: The encoded information in the first image to be reviewed is identified using optical character recognition algorithm and regular expression matching algorithm. If the encoded information in the first image to be reviewed cannot be identified by optical character recognition algorithm and regular expression matching algorithm, the encoded information in the first image to be reviewed is identified by named entity recognition algorithm. If the named entity recognition algorithm fails to identify the encoded information in the first image to be reviewed, multiple pre-trained large models are used to identify the encoded information in the first image to be reviewed. The same recognition result in the multiple large models is used as the recognition result of the encoded information.

5. The image verification method according to claim 3, characterized in that, The second image to be reviewed includes: a first image and a second image, wherein the first image is a product image of the target product, and the second image is a composite image of the target product and its outer packaging.

6. The image verification method according to claim 5, characterized in that, During the execution of the first review process or the second review process on the second image, the method further includes: Using a preset segmentation algorithm, the target product and its outer packaging in the second image are identified. Based on the positions of the target product and its outer packaging in the second image, the second image is segmented to obtain a first sub-image and a second sub-image. The first sub-image includes the target product, and the second sub-image includes the outer packaging. The first sub-image and the second sub-image are reviewed separately.

7. The image verification method according to claim 1, characterized in that, The first review process also includes a second review item, which includes at least one of the following: review of random shooting, review of re-shooting, and review of image clarity.

8. The image verification method according to claim 1, characterized in that, During the first review process on the first image to be reviewed, the method further includes: If the first image to be reviewed fails the review, the first image to be reviewed is retrieved again, and the first review process is re-executed until the first image to be reviewed passes the review or the number of reviews reaches the preset number.

9. An electronic device, characterized in that, include: Memory and processor; The memory is connected to the processor and is used to store programs; The processor is used to implement the image review method as described in any one of claims 1 to 8 by running a program in the memory.

10. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the image review method as described in any one of claims 1 to 8.

Images