Electronic scale control method and device based on image recognition and reverse payment, and medium

Abstract

The invention discloses an electronic scale control method and device based on image recognition and reverse payment, and a medium, and relates to the technical field of electronic scales. The method comprises the following steps: acquiring an image of an article placed on a scale pan, and obtaining weight data of the article; performing forward reasoning on the article image, extracting image features, and outputting a preliminary identification result of the article and a corresponding confidence coefficient; fusing the preliminary recognition result and the weight data on a feature level to generate an article identifier; in response to a reverse payment instruction initiated by a merchant, a dynamic collection two-dimensional code is rendered and generated on a terminal display screen, and a user is prompted to show a payment code image; a payment code image shown by a user is collected, and the payment code image is analyzed to obtain an encrypted collection account identifier. According to the method, the object type is automatically obtained through image recognition, fusion verification is carried out on the object type and the weighing data, the recognition accuracy and reliability are greatly improved, and a foundation is laid for automatic pricing.

Description

Technical Field

[0001] This application relates to the field of electronic scale technology, and in particular to an electronic scale control method, device and medium for image recognition and reverse payment. Background Technology

[0002] Existing smart electronic scales mainly integrate weighing, pricing, and traditional merchant payment functions, achieving a preliminary level of electronic payment. However, in scenarios where payment to the seller is required, such as agricultural product procurement, valuable material transactions, or tea markets, existing technology has significant shortcomings.

[0003] First, the payment process is one-way, making it impossible for merchants to directly make reverse payments to sellers' personal accounts via electronic scale terminals. The process is cumbersome, and the flow of funds is difficult to track electronically. Second, the transaction data of existing equipment is isolated, and the identification of goods, such as agricultural products and material categories, relies heavily on manual input, which is inefficient and prone to errors. This prevents transaction information, item information, and payment information from being automatically linked to form an effective electronic traceability chain. This poses significant challenges to merchants' inventory management, financial reconciliation, and supply chain traceability.

[0004] Based on the above analysis, the problems and shortcomings of the existing technology are as follows: Existing electronic scales cannot achieve automatic, secure, and traceable reverse payments, further fragmenting transaction information. Summary of the Invention

[0005] This application provides an electronic scale control method, device, and medium for image recognition and reverse payment, which can solve the problems of electronic scales being unable to achieve automatic, secure, and traceable reverse payment and the fragmentation of transaction information in the prior art.

[0006] In a first aspect, embodiments of this application provide an electronic scale control method for image recognition and reverse payment. The method includes: acquiring an image of an item placed on the scale pan and obtaining the weight data of the item; performing forward inference on the item image, extracting image features and outputting a preliminary identification result of the item and the corresponding confidence level; fusing the preliminary identification result and the weight data at the feature level to generate an item identifier; responding to a reverse payment instruction initiated by a merchant, rendering and generating a dynamic payment QR code on the terminal display screen and prompting the user to present a payment code image; acquiring the payment code image presented by the user and parsing the payment code image to obtain an encrypted payment account identifier.

[0007] In one implementation of this application, the preliminary identification result and weight data are fused at the feature level to generate an item identifier. Specifically, this includes: comparing the weight data with the upper and lower thresholds of the standard weight range corresponding to the preliminary identification result; if the weight data falls within the standard weight range, confirming the preliminary identification result as an item identifier; if the weight data falls outside the standard weight range, activating an auxiliary identification mechanism.

[0008] In one implementation of this application, after performing forward inference on the object image, extracting image features, and outputting the preliminary recognition result of the object, the method further includes: if the confidence level is lower than a confidence threshold, determining whether the illumination of the object image is lower than a preset threshold; if it is lower than the preset threshold, activating an image enhancement algorithm, adjusting the contrast and gamma value of the object image, and detecting whether there are overexposed highlight areas in the object image caused by object reflection; if there are overexposed highlight areas, using an image inpainting-based algorithm to fill and repair the overexposed highlight areas based on surrounding pixel information, generating an optimized object image; and inputting the optimized object image into a convolutional neural network model for recognition.

[0009] In one implementation of this application, if the weight data falls outside the standard weight range, an auxiliary recognition mechanism is activated, which specifically includes: receiving the user's selection command via a touch screen or voice; verifying the selection command; and if the verification fails, activating a second auxiliary recognition mechanism.

[0010] In one implementation of this application, the selection command is verified. If the verification fails, a second auxiliary identification mechanism is activated, which specifically includes: activating the material feature detection module integrated in the terminal to perform feature scanning on the item and obtain the feature spectrum of the item. The material feature detection module includes a near-infrared sensor and an optical feature scanner. The feature spectrum is compared with the feature spectrum corresponding to the selection command to generate a feature matching degree. If the feature matching degree does not reach a preset threshold, the user is prompted to check whether the selected type is accurate, and staff are notified.

[0011] In one implementation of this application, the method further includes: verifying the consistency between the hardware fingerprint and the pre-registered device information, and whether there are any abnormal markers in the real-time status information of the terminal; if the hardware fingerprint does not match or there are abnormal markers, suspending the payment process and pushing an anomaly investigation instruction to the merchant terminal.

[0012] In one implementation of this application, the method further includes: embedding transaction context information when the QR code is generated, the context information including the current item type identifier, the hash value of the weight data, and the validity period; after the user scans the code, verifying whether the payment operation matches the transaction context information embedded in the QR code; if the validity period has expired or the item information has been tampered with, refusing payment and simultaneously prompting the terminal to refresh the QR code.

[0013] In one implementation of this application, after collecting the payment code image presented by the user and parsing the payment code image to obtain an encrypted payee account identifier, the method further includes: constructing a payment request data packet, the data packet including the payee account identifier, type identifier, weight data and payment amount, and a user identity session token generated by biometric authentication; decrypting and verifying the validity of the user identity session token through a two-way authentication SSL encrypted channel, and calling the payment gateway application interface to transfer funds to the account corresponding to the payee account identifier; after the funds are successfully transferred, generating a structured transaction record, associating it with the transaction timestamp, and writing it into a tamper-proof cloud database for traceability and reconciliation queries.

[0014] Secondly, embodiments of this application also provide an electronic scale control device for image recognition and reverse payment. The device includes at least one processor and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to: acquire an image of an item placed on the weighing pan and obtain the weight data of the item; perform forward inference on the item image, extract image features, and output a preliminary identification result of the item and the corresponding confidence level; fuse the preliminary identification result and the weight data at the feature level to generate an item identifier; in response to a reverse payment instruction initiated by a merchant, render and generate a dynamic payment QR code on the terminal display screen and prompt the user to present a payment code image; acquire the payment code image presented by the user and parse the payment code image to obtain an encrypted payment account identifier.

[0015] Thirdly, this application also provides a non-volatile computer storage medium for controlling an electronic scale with image recognition and reverse payment, storing computer-executable instructions. The computer-executable instructions are configured to: acquire an image of an item placed on the scale pan and obtain the weight data of the item; perform forward inference on the item image, extract image features and output a preliminary identification result of the item and the corresponding confidence level; fuse the preliminary identification result and the weight data at the feature level to generate an item identifier; in response to a reverse payment instruction initiated by a merchant, render and generate a dynamic payment QR code on the terminal display screen and prompt the user to present a payment code image; acquire the payment code image presented by the user and parse the payment code image to obtain an encrypted payment account identifier.

[0016] This application provides an electronic scale control method, device, and medium for image recognition and reverse payment. It automatically acquires the item type through image recognition and merges it with weighing data for verification, greatly improving the accuracy and reliability of recognition and laying the foundation for automatic pricing. By generating dynamic QR codes, scanning user payment codes, and linking with a cloud payment gateway, it improves transaction efficiency and security. It constructs a trusted traceability data chain, ensuring the authenticity and non-repudiation of transaction participants and data through multiple security measures such as biometric authentication, device verification, and context binding. Key elements of each transaction, such as item information, weight, amount, identities of both parties, and timestamps, are automatically associated and encrypted, and even written to the blockchain, forming a complete, reliable, and easily searchable electronic traceability record, greatly facilitating subsequent reconciliation, auditing, and supply chain management. It enhances the system's intelligence and interactive experience: when automatic identification is questionable, it introduces human-computer interaction and multi-level auxiliary identification mechanisms to ensure accurate data input even in complex scenarios, guaranteeing the system's practicality and robustness. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating an electronic scale control method for image recognition and reverse payment provided in this application embodiment; Figure 2 This is a schematic diagram of the internal structure of an electronic scale control device for image recognition and reverse payment provided in an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in 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] This application provides an electronic scale control method, device, and medium for image recognition and reverse payment, which solves the problems in the prior art where electronic scales cannot achieve automatic, secure, and traceable reverse payment and where transaction information is fragmented.

[0020] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0021] Figure 1This is a flowchart illustrating an electronic scale control method for image recognition and reverse payment, provided as an embodiment of this application. Figure 1 As shown in the figure, the electronic scale control method for image recognition and reverse payment provided in this application embodiment specifically includes the following steps: Step 10: Capture an image of the item placed on the weighing pan and obtain the weight data of the item; Step 20: Perform forward inference on the object image, extract image features, and output the preliminary identification results of the object, as well as the corresponding confidence level.

[0022] In this step, the electronic scale terminal adopts a locally deployed lightweight convolutional neural network model. Taking tea as an example, the model training set covers more than 600 common tea varieties, including image samples of each tea in different states and forms, including loose tea, compressed tea, tea bags, and under different lighting conditions, and is labeled with key feature tags, such as Longjing tea - flat shape - yellow-green background with white frost, and Pu'er ripe tea - tightly rolled strips - brownish-red.

[0023] Includes: Morphological features: leaf curl, strip tightness, and tea cake pressing texture; Color features: distribution of the main color tone under the RGB channel, such as the proportion of gold, yellow, and black in Jin Jun Mei; Microscopic features: tea hair density, leaf vein distribution, and fragmentation; Output result format: preliminary recognition results output after model inference.

[0024] As an optional embodiment, after performing forward inference on the object image, extracting image features, and outputting the preliminary recognition result of the object, the method may further include: Step 201: If the confidence level is lower than the confidence threshold, determine whether the illumination of the object image is lower than a preset threshold; Step 202: If it is lower than the preset threshold, start the image enhancement algorithm, adjust the contrast and gamma value of the object image, and detect whether there are overexposed highlight areas in the object image caused by the reflection of the object; Step 203: If there are overexposed highlight areas, use an image inpainting-based algorithm to fill and repair the overexposed highlight areas according to the surrounding pixel information to generate an optimized object image; Step 204: Input the optimized object image into a convolutional neural network model for recognition.

[0025] In this step, the Retinex decomposition algorithm is used to separate the illumination and reflection components of the image, improving the contrast of the tea texture area and enhancing the edge contour of Longjing tea leaves. Simultaneously, the gamma value is dynamically adjusted (γ ranges from 1.2 to 1.8), adaptively adjusting based on the proportion of dark areas to improve the color representation of tea leaves in shadow areas, avoiding misidentification of the reddish-brown color of Zhengshan Xiaozhong as Qimen black tea. A threshold-based highlight detection algorithm is used to identify overexposed areas on the tea surface caused by reflections from plastic bags or direct illumination from supplementary lights, as well as reflective points of tea hairs in loose tea that are directly illuminated by lights. For the detected overexposed highlight areas, an image inpainting algorithm based on generative adversarial networks is used for filling and repair: using the effective pixels at the edge of the overexposed area—the normal color and texture of the tea leaves surrounding the highlight—as a reference, the model generates filling content consistent with the surrounding features. Considering the specific characteristics of tea leaves, the inpainting algorithm focuses on preserving the directionality of the tea hairs, such as the prominent white hairs of Biluochun, and the continuity of the tea strands, such as the robust strands of Da Hong Pao, to avoid the loss of key features after repair. The repaired tea image is re-input into the convolutional neural network model, and the same inference process as in step 20 is executed to output the secondary recognition result and confidence level. If the confidence level of the secondary recognition increases above the threshold, the result is used as the preliminary recognition result; if it is still below the threshold, further calibration is performed by combining weight data.

[0026] The above process can effectively address the complex imaging environment of the tea market, reduce recognition errors caused by light and reflection, and significantly improve the accuracy of tea identification, especially for tea varieties that rely on color and texture features.

[0027] Step 30: Fuse the preliminary identification results with the weight data at the feature level to generate an item identifier; In this step, the inherent relationship between tea type and weight is utilized to fuse the image features of the preliminary identification results with the weight data, calibrate the identification results, and finally generate accurate tea type labels, such as West Lake Longjing (loose tea) and Pu'er cake tea (357g).

[0028] As an optional embodiment, the preliminary identification results and weight data are fused at the feature level to generate an item identifier, which may specifically include: Step 301: Compare the weight data with the upper and lower thresholds of the standard weight range corresponding to the preliminary identification results; In this step, the system pre-stores a database of tea categories and standard weight ranges, retrieves the standard range corresponding to the preliminary identification results, and compares the actual weight data with the upper and lower thresholds of the range.

[0029] Step 302: If the weight data falls within the standard weight range, confirm the preliminary identification result as an item identification; Step 303: If the weight data falls outside the standard weight range, activate the auxiliary identification mechanism.

[0030] In this step, if there is a slight deviation, the weight will be slightly off, and the preliminary result will be retained for auxiliary recognition; if there is a significant deviation, auxiliary recognition will be started directly, and the terminal will display a candidate list including the preliminary judgment result, similar categories with matching weight, and frequently traded categories of merchants. Merchants can select and confirm via touch screen or voice, and the final selection result will be used as the item identifier.

[0031] As an optional embodiment, if the weight data falls outside the standard weight range, an auxiliary recognition mechanism is activated, which may specifically include: Step 3031: Receiving the user's selection instruction via touch screen or voice; In this step, when the weight data deviates from the standard range, the terminal display screen can show 3-5 candidate tea categories. Merchants can select them by clicking on the touch screen or by entering a voice command to confirm. The system receives and parses the command in real time to generate the tea type to be verified.

[0032] Step 3032: Verify the selected instruction. If the verification fails, activate the second auxiliary identification mechanism.

[0033] As an optional embodiment, the selection command is verified. If the verification fails, a second auxiliary identification mechanism is activated, which may include: Step 30321: Activating the material feature detection module integrated in the terminal to perform feature scanning on the item and obtain the feature spectrum of the item. The material feature detection module includes a near-infrared sensor and an optical feature scanner; Step 30322: Comparing the feature spectrum with the feature spectrum corresponding to the selection command to generate a feature matching degree; Step 30323: If the feature matching degree does not reach a preset threshold, prompting the user to check whether the selected type is accurate and notifying the staff.

[0034] In this step, the near-infrared sensor and optical feature scanner integrated into the terminal are activated to perform a dual feature scan on the tea leaves on the weighing pan: The scan acquires the component feature spectrum of the tea leaves, focusing on the characteristic peaks of marker components such as tea polyphenols, amino acids, and caffeine. For example, Longjing tea typically has a high amino acid content, corresponding to specific wavelength absorption peaks. It also captures high-resolution spectra of the leaf microstructure, such as the dense white down of Biluochun and the dragonfly-head shape of Tieguanyin. The scanned spectra are compared with the system's pre-stored standard feature spectrum library for Biluochun, and the feature matching degree (value 0-1) is calculated using a cosine similarity algorithm. If the matching degree is greater than or equal to a preset threshold, the merchant's selection is confirmed as correct, and Biluochun is used as the final identifier. If the matching degree is less than the threshold, the terminal immediately displays a feature mismatch warning indicating that the tea polyphenol content is too low, potentially inconsistent with Biluochun. Simultaneously, a notification is pushed to market management personnel via the merchant's app for on-site verification, preventing transaction disputes caused by misclassification.

[0035] Step 40: In response to the merchant's reverse payment instruction, a dynamic payment QR code is generated on the terminal display screen, and the user is prompted to show the payment code image.

[0036] In this step, you can choose to actively scan to pay or passively scan to pay.

[0037] Step 50: Collect the payment code image presented by the user and parse the payment code image to obtain the encrypted receiving account identifier.

[0038] As an optional embodiment, the method may further include: verifying the consistency between the hardware fingerprint and the pre-registered device information, and whether there are any abnormal markers in the real-time status information of the terminal; if the hardware fingerprint does not match or there are abnormal markers, suspending the payment process and pushing an anomaly investigation instruction to the merchant terminal.

[0039] In this step, the terminal hardware fingerprint is generated by combining the device's unique MAC address, the hash value of the weighing sensor calibration parameters, and the camera module serial number. It is then used to verify the consistency with the device information submitted by the merchant during pre-registration, such as checking the registration fingerprint of scale number 10 at the tea stall. The system also detects in real time whether there are any abnormal markings on the device, including: the weighing sensor has not been calibrated within 30 days, the camera is obstructed, or the near-infrared module is faulty, all of which may affect the accuracy of tea weight or category identification.

[0040] As an optional embodiment, the method may further include: embedding transaction context information when the QR code is generated, the context information including the current item type identifier, the hash value of the weight data, and the validity period; after the user scans the code, verifying whether the payment operation matches the transaction context information embedded in the QR code; if the validity period has expired or the item information has been tampered with, refusing payment and simultaneously prompting the terminal to refresh the QR code.

[0041] In this step, the QR code is generated by embedding key information for the current transaction: tea type identifier, SHA-256 hash value of weight data to prevent tampering with weighing data, and a 60-second validity period to accommodate the rapid transaction pace after tea weighing. After the user scans the code, the payment gateway automatically verifies: whether the payment operation is within the validity period, and whether the tea type / weight hash at the time of scanning matches the information embedded in the QR code, preventing merchants from secretly changing the tea category after weighing. If the validity period has expired or the information does not match, the payment gateway rejects the transaction, the terminal simultaneously prompts that the QR code has expired / the information is abnormal, and automatically refreshes the QR code.

[0042] As an optional embodiment, after collecting the payment code image presented by the user and parsing the payment code image to obtain the encrypted payee account identifier, the method may further include: constructing a payment request data packet, the data packet including the payee account identifier, type identifier, weight data and payment amount, and a user identity session token generated by biometric authentication; decrypting and verifying the validity of the user identity session token through a two-way authentication SSL encrypted channel, and calling the payment gateway application interface to transfer funds to the account corresponding to the payee account identifier; after the funds are successfully transferred, generating a structured transaction record, associating it with the transaction timestamp, and writing it into a tamper-proof cloud database for traceability and reconciliation queries.

[0043] This step involves encrypting the user's payment account, tea type identifier, weight data, calculated payment amount, and user biometric authentication token. The data packet is transmitted through a two-way authentication SSL channel. The terminal and the cloud mutually verify the certificates. After decryption, the cloud first verifies the validity of the biometric token to confirm that it is the user's own operation, and then calls the payment gateway interface to complete the fund transfer, deducting 50g × 800 yuan / jin = 80 yuan from the user's account to the merchant's account. After successful payment, a structured record is generated, including tea type, weight, amount, identity of buyer and seller, and timestamp, and written to the tamper-proof cloud database for subsequent tea traceability and merchant reconciliation.

[0044] The above are embodiments of the method proposed in this application. Based on the same inventive concept, embodiments of this application also provide an electronic scale control device for image recognition and reverse payment, the structure of which is as follows: Figure 2 As shown.

[0045] Figure 2 This is a schematic diagram of the internal structure of an electronic scale control device for image recognition and reverse payment, provided as an embodiment of this application. Figure 2 As shown, the device includes: At least one processor 201; And a memory 202 that is communicatively connected to at least one processor; The memory 202 stores instructions executable by at least one processor. These instructions are executed by at least one processor 201 to enable the processor 201 to: acquire an image of an item placed on a weighing pan and obtain the weight data of the item; perform forward inference on the item image, extract image features, and output a preliminary identification result of the item and the corresponding confidence level; fuse the preliminary identification result with the weight data at the feature level to generate an item identifier; in response to a reverse payment instruction initiated by a merchant, render and generate a dynamic payment QR code on the terminal display screen and prompt the user to present a payment code image; acquire the payment code image presented by the user and parse the payment code image to obtain an encrypted payment account identifier.

[0046] Some embodiments of this application provide corresponding to Figure 1 A non-volatile computer storage medium for controlling an electronic scale with image recognition and reverse payment stores computer-executable instructions. These instructions are configured to: acquire an image of an item placed on the scale pan and obtain the item's weight data; perform forward inference on the item image, extract image features, and output a preliminary identification result of the item and its corresponding confidence level; fuse the preliminary identification result with the weight data at the feature level to generate an item identifier; respond to a reverse payment instruction initiated by a merchant, render and generate a dynamic payment QR code on the terminal display screen, and prompt the user to present a payment code image; acquire the payment code image presented by the user, and parse the payment code image to obtain an encrypted payment account identifier.

[0047] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for IoT devices and media are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0048] The systems, media, and methods provided in this application are one-to-one correspondences. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.

[0049] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0050] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0051] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0052] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0053] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0054] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0055] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0056] It should also be noted that 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 limitation, 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.

[0057] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An electronic scale control method of image recognition and reverse payment, characterized by, The method includes: Capture an image of an item placed on a weighing pan and obtain the weight data of that item; Forward reasoning is performed on the object image to extract image features and output the preliminary identification result of the object, as well as the corresponding confidence level; The preliminary identification results are fused with the weight data at the feature level to generate an item identifier; In response to a merchant's reverse payment instruction, a dynamic QR code for receiving payment is generated on the terminal display screen, and the user is prompted to present the payment code image. The payment code image presented by the user is collected, and the payment code image is parsed to obtain an encrypted receiving account identifier.

2. The electronic scale control method of claim 1, wherein, The step of fusing the preliminary identification result with the weight data at the feature level to generate an item identifier specifically includes: The weight data is compared with the upper and lower thresholds of the standard weight range corresponding to the preliminary identification result; If the weight data falls within the standard weight range, the preliminary identification result is confirmed as the item identifier; If the weight data falls outside the standard weight range, an auxiliary identification mechanism is activated.

3. The electronic scale control method of claim 1, wherein, After performing forward reasoning on the object image, extracting image features, and outputting preliminary object recognition results, the method further includes: If the confidence level is lower than the confidence threshold, determine whether the illuminance of the item image is lower than a preset threshold; If the image is below the preset threshold, the image enhancement algorithm is activated to adjust the contrast and gamma value of the image of the item, and to detect whether there are overexposed highlight areas in the image of the item caused by the reflection of the item. If the overexposed highlight area exists, an image inpainting algorithm is used to fill and repair the overexposed highlight area based on the surrounding pixel information to generate an optimized image of the item. The optimized image of the item is input into a convolutional neural network model for recognition.

4. The electronic scale control method of claim 2, wherein, If the weight data falls outside the standard weight range, an auxiliary identification mechanism is activated, specifically including: Receive user selection commands via touchscreen or voice; The selection command is verified. If the verification fails, a second auxiliary identification mechanism is activated.

5. The electronic scale control method of claim 4, wherein, The step of verifying the selection instruction, and if the verification fails, activating the second auxiliary recognition mechanism, specifically includes: The integrated material feature detection module of the terminal is activated to perform feature scanning on the item and obtain the feature spectrum of the item. The material feature detection module includes a near-infrared sensor and an optical feature scanner. The feature spectrum is compared with the feature spectrum corresponding to the selection instruction to generate a feature matching degree; If the feature matching degree does not reach the preset threshold, the user is prompted to check whether the selected type is accurate, and staff are notified.

6. The electronic scale control method of claim 1, wherein, The method further includes: Verify the consistency between the hardware fingerprint and the pre-registered device information, and check for any abnormal markers in the terminal's real-time status information; If the hardware fingerprint does not match or there is an abnormal mark, the payment process is suspended and an anomaly investigation instruction is pushed to the merchant terminal.

7. The electronic scale control method of claim 1, wherein, The method further includes: When the QR code is generated, transaction context information is embedded, including the current item type identifier, the hash value of the weight data, and the validity period. After the user scans the code, the system verifies whether the payment operation matches the transaction context information embedded in the QR code. If the validity period has expired or the item information has been altered, payment will be refused, and the terminal will be prompted to refresh the QR code.

8. The electronic scale control method of claim 1, wherein, After collecting the payment code image presented by the user and parsing the payment code image to obtain an encrypted payee account identifier, the method further includes: Construct a payment request data packet, the data packet including the receiving account identifier, type identifier, weight data and payment amount, and a user identity session token generated by biometric authentication; The validity of the user identity session token is decrypted and verified through a two-way authentication SSL encrypted channel, and the payment gateway application interface is called to transfer funds to the account corresponding to the receiving account identifier. After the funds are successfully transferred, a structured transaction record is generated, which is linked to the transaction timestamp and written into a tamper-proof cloud database for traceability and reconciliation.

9. An electronic scale control device for image recognition and reverse payment, characterized by, The device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to: Capture an image of an item placed on a weighing pan and obtain the weight data of that item; Forward reasoning is performed on the object image to extract image features and output the preliminary identification result of the object, as well as the corresponding confidence level; The preliminary identification results are fused with the weight data at the feature level to generate an item identifier; In response to a merchant's reverse payment instruction, a dynamic QR code for receiving payment is generated on the terminal display screen, and the user is prompted to present the payment code image. The payment code image presented by the user is collected, and the payment code image is parsed to obtain an encrypted receiving account identifier.

10. A non-transitory computer storage medium storing computer-executable instructions of an electronic scale control of image recognition and reverse payment, characterized in that, The computer-executable instructions are set as follows: Capture an image of an item placed on a weighing pan and obtain the weight data of that item; Forward reasoning is performed on the object image to extract image features and output the preliminary identification result of the object, as well as the corresponding confidence level; The preliminary identification results are fused with the weight data at the feature level to generate an item identifier; In response to a merchant's reverse payment instruction, a dynamic QR code for receiving payment is generated on the terminal display screen, and the user is prompted to present the payment code image. The payment code image presented by the user is collected, and the payment code image is parsed to obtain an encrypted receiving account identifier.

Images