Scrap steel traceability and risk early warning management method based on image recognition and block chain

Abstract

The invention discloses a waste steel traceability and risk early warning management method based on image recognition and a block chain. According to the method, firstly, a waste steel recognition model is constructed, image segmentation and target detection are carried out, real-time video streams of a waste steel storage yard are processed in combination with OpenCV to carry out material type classification, and when a transport vehicle arrives at an unloading area, order information is automatically associated through license plate recognition, material types and weight data are synchronously obtained, and the transport vehicle is guided to a corresponding warehouse area in real time. And in the unloading process, the scrap steel recognition model is called to recognize the scrap steel in real time, and accurate classification and storage of the scrap steel are ensured. And meanwhile, a block chain technology is adopted to link related information, so that tampering-free full-life-cycle recording is realized. The accuracy and efficiency of waste steel identification are improved, full-process traceability and risk early warning are achieved, the bin pipe and charging process is optimized, the production cost is reduced, and the production efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of image recognition technology, and in particular to a method for traceability and risk warning management of scrap steel based on image recognition and blockchain. Background Technology

[0002] In the metallurgical industry, scrap steel management and quality control are crucial aspects of the production process. Traditional scrap steel management suffers from numerous problems, such as the ease with which scrap steel testing data can be lost and difficult to compile; an incomplete traceability system making it difficult to assign responsibility; and economic losses caused by scrap steel quality issues during the smelting process.

[0003] Existing intelligent scrap steel quality inspection technologies have failed to establish a feedback loop of "slag analysis → scrap steel characteristics," leading to the recurrence of the same quality problems. Furthermore, existing technologies also have shortcomings in the storage and feeding stages of scrap steel, such as non-standard material storage and low accuracy in batching. Summary of the Invention

[0004] Based on this, the embodiments of this application provide a scrap steel traceability and risk warning management method based on image recognition and blockchain. This method can comprehensively and efficiently carry out scrap steel traceability and risk warning management, so as to improve the intelligence and efficiency of scrap steel processing, ensure scrap steel quality, reduce production costs, and improve production efficiency.

[0005] Firstly, a method for scrap steel traceability and risk warning management based on image recognition and blockchain is provided, which includes:

[0006] A scrap steel identification model is constructed and used to identify scrap steel in the unloading area. Specifically, the scrap steel identification model uses Mask R-CNN for image segmentation and target detection, combines OpenCV to process real-time video streams of the scrap steel yard, and uses a ResNet deep learning model for material type classification. The training data includes manually labeled scrap steel images.

[0007] When the transport vehicle arrives at the unloading area, the order information is automatically linked through license plate recognition, and the material type and weight data are obtained simultaneously. The vehicle is guided to the corresponding storage area in real time. During the unloading process, the constructed scrap steel identification model is called to identify the scrap steel in real time, ensuring the accurate classification and storage of scrap steel.

[0008] During the unloading of scrap steel, real-time photos of the scrap steel are taken and saved to the server management system. Blockchain technology is used to put the scrap steel photos, vehicle information, time-series data, composition data, and supplier information on the chain.

[0009] The constructed scrap steel identification model is invoked to analyze the collected scrap steel images and accurately inspect the quality of the captured scrap steel portions. When foreign objects or abnormalities are detected in the scrap steel inside the vehicle, the on-site signal lights remind the crane operator to conduct a secondary on-site confirmation, thereby achieving full-process quality inspection tracking and control.

[0010] Optionally, the method further includes:

[0011] When a scrap steel blending task is received, a reasonable blending plan with the lowest cost is generated by combining the purchase price, market price, and finished product selling price. The plan is then sent to the feeding crane to form a blending task. At the same time, a blacklist area and electronic fence are generated to prevent the feeding crane from entering irrelevant areas unnecessarily.

[0012] The constructed scrap steel identification model is invoked to identify scrap steel in real time during the feeding process, generate scrap steel type and grade information, and generate a batching progress bar based on the read weight information.

[0013] Optionally, constructing a scrap steel identification model also includes:

[0014] The training data is preprocessed, including grayscale conversion, binarization, and edge detection, to enhance image features;

[0015] Data augmentation techniques are used to expand the training dataset; these techniques include random cropping, rotation, and flipping.

[0016] Cross-validation was used to evaluate the model performance.

[0017] Optionally, automatically linking order information via license plate recognition also includes:

[0018] Verify the license plate recognition results to ensure the accuracy of the license plate number;

[0019] Add vehicle arrival timestamp to order information;

[0020] When license plate recognition fails, a manual intervention process is automatically triggered to ensure the correct association of order information.

[0021] Optionally, the step of taking real-time photos of scrap steel also includes:

[0022] Use cameras at multiple angles to take pictures to ensure comprehensive coverage of the scrap steel;

[0023] The system transmits and stores captured images in real time. During the photo-taking process, it automatically detects lighting conditions and adjusts camera parameters automatically when the lighting is insufficient to ensure image quality.

[0024] Optionally, the analysis using the scrap steel identification model also includes:

[0025] The collected scrap steel images are preprocessed, including background noise removal and image size standardization;

[0026] Use deep learning models to extract features and classify preprocessed images;

[0027] When an anomaly is detected, the anomaly information is automatically recorded and an anomaly report is generated.

[0028] Secondly, a scrap steel traceability and risk warning management system based on image recognition and blockchain is provided. This system includes:

[0029] The module is used to build a scrap steel identification model, which is used to identify scrap steel in the unloading area. Specifically, the scrap steel identification model uses Mask R-CNN for image segmentation and target detection, combines OpenCV to process real-time video streams of the scrap steel yard, and uses a ResNet deep learning model for material type classification. The training data includes manually labeled scrap steel images.

[0030] The identification module is used to automatically associate order information through license plate recognition when the transport vehicle arrives at the unloading area, synchronously obtain material type and weight data, and guide the corresponding storage area in real time. During the unloading process, the constructed scrap steel identification model is called to identify the scrap steel in real time, ensuring the accurate classification and storage of scrap steel.

[0031] The data acquisition module is used to take real-time photos of the scrap steel during the unloading process and save them to the server management system. It uses blockchain technology to put the scrap steel photos, vehicle information, time-series data, composition data, and supplier information on the chain.

[0032] The confirmation module is used to call the constructed scrap steel identification model, analyze the collected scrap steel images, and accurately verify the quality of the captured scrap steel. When foreign objects or abnormalities are detected in the scrap steel inside the vehicle, the on-site signal lights remind the crane operator to conduct a secondary on-site confirmation, realizing full-process quality verification tracking and control.

[0033] Thirdly, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any of the methods described in the first aspect above.

[0034] Fourthly, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements any of the methods described in the first aspect above.

[0035] Fifthly, a computer program product is provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements any of the methods described in the first aspect above.

[0036] The beneficial effects of the technical solutions provided in this application include at least the following:

[0037] (1) By using Mask R-CNN for image segmentation and target detection, combined with OpenCV for processing real-time video streams of scrap steel yards, and using ResNet deep learning model for material type classification, this method can achieve high-precision scrap steel identification. The detection rate of individual scrap steel is as high as 95% or more, and the accuracy of individual scrap steel is as high as 93%, which significantly improves the efficiency and accuracy of scrap steel processing.

[0038] (2) By using blockchain technology, scrap steel photos, vehicle information, time-series data, composition data, and supplier information are uploaded to the chain to ensure an immutable full lifecycle record. This not only enables full-process traceability of scrap steel but also provides timely warnings when anomalies are detected, reducing risks in the production process.

[0039] (3) By automatically linking order information through license plate recognition, vehicles are guided to the corresponding warehouse area in real time, and the movement of the overhead crane is monitored in real time during the unloading process to ensure the accurate classification and storage of scrap steel. The dynamic feeding module combines the purchase price, market price, and finished product selling price to generate a reasonable ratio scheme with the lowest cost, which improves the automation level of warehouse management and feeding and reduces labor costs. Attached Figure Description

[0040] To more clearly illustrate the embodiments of this application or the technical solutions in 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 merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0041] Figure 1 A flowchart illustrating the steps of a scrap steel traceability and risk warning management method based on image recognition and blockchain, provided in this application embodiment;

[0042] Figure 2 A block diagram of a scrap steel traceability and risk warning management system based on image recognition and blockchain provided in this application embodiment;

[0043] Figure 3 This is a schematic diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0045] In the description of this application, the terms "comprising," "having," and any variations thereof are intended to cover non-exclusive inclusion, such as a process, method, system, product, or apparatus that includes a series of steps or units, not necessarily limited to those steps or units that are expressly listed, but may also include other steps or units that are not expressly listed but are inherent to these processes, methods, products, or apparatuses, or steps or units added based on further optimizations conceived in this application.

[0046] When scrap steel is used as a raw material in steelmaking plants, hazardous materials (such as oil, paint, and plastics), harmful metals (such as lead, cadmium, mercury, and zinc), or other impurities (such as rubber and electronic waste) may be mixed in with the scrap steel, posing serious risks to the smelting process, equipment lifespan, product quality, and environmental and human health. Steel mills will need to strengthen control over the source and composition testing reports of scrap steel. This patent application aims to address the issue that when similar scrap steel sources are discovered during the steelmaking process, the source and supplier of the scrap steel can be traced back through a system using scrap steel photographs. This allows for stricter control during the acceptance of scrap steel deliveries from suppliers, and also improves the marking, warning, and inclusion of similar scrap steel during the grading process, preventing its acceptance and avoiding subsequent economic losses during the smelting process.

[0047] The intelligent scrap steel quality inspection system takes real-time photos of the scrap steel during the unloading process and saves the image set to the server management system. Using blockchain technology, the scrap steel photos, vehicle information, time-series data, composition data, and supplier information are recorded on the chain to ensure an immutable full lifecycle record.

[0048] The shortcomings of existing technology include:

[0049] The existing system has not established a feedback loop from "slag analysis to scrap steel characteristics", resulting in the same quality problems occurring repeatedly.

[0050] Data from scrap steel testing is prone to loss and difficult to compile; the traceability system is incomplete, making it difficult to trace responsibility; process control is disconnected and lacks feedback optimization.

[0051] This application aims to address the issue of identifying similar scrap steel sources during the steelmaking process. By using photos of the scrap steel, the system can trace its origin and supplier, thereby strengthening control over the acceptance of scrap steel deliveries. Furthermore, it enhances the marking, warning, and exclusion of similar scrap steel during the grading process, preventing its acceptance and avoiding subsequent economic losses during the smelting process.

[0052] Please refer to Figure 1The document illustrates a flowchart of a scrap steel traceability and risk warning management method based on image recognition and blockchain, as provided in an embodiment of this application. The method may include the following steps:

[0053] S1. Construct a scrap steel identification model and use the scrap steel identification model to identify scrap steel in the unloading area.

[0054] Specifically, the scrap steel identification model uses Mask R-CNN for image segmentation and target detection, combines OpenCV to process real-time video streams from scrap steel yards, and utilizes a ResNet deep learning model for material type classification. The training data includes manually labeled scrap steel images.

[0055] The specific implementation process for this step is as follows:

[0056] Image segmentation and object detection were performed using the Mask R-CNN algorithm. Mask R-CNN is an advanced object detection algorithm that can accurately locate and segment the target region of scrap steel in an image, providing a foundation for subsequent recognition.

[0057] This paper describes how to process real-time video streams from a scrap yard using the OpenCV library. OpenCV is a powerful computer vision library capable of real-time video stream processing, including image reading, preprocessing, and feature extraction, providing high-quality input data for Mask R-CNN.

[0058] The ResNet deep learning model is used for scrap steel type classification. ResNet is a classic deep learning network architecture with strong feature learning capabilities. By training on a large number of manually labeled scrap steel images, the ResNet model can learn the characteristics of different scrap steel types, thereby achieving accurate classification of scrap steel. The training data includes tens of thousands of scrap steel images classified and labeled according to the GB / T4223 standard. These images cover various types of scrap steel, ensuring that the model has good generalization ability and can accurately identify scrap steel of different thicknesses (from 0.5mm to over 20mm) and shapes.

[0059] S2: When the transport vehicle arrives at the unloading area, the order information is automatically linked through license plate recognition, and the material type and weight data are obtained simultaneously. The vehicle is guided to the corresponding storage area in real time. During the unloading process, the constructed scrap steel recognition model is called to identify the scrap steel in real time, ensuring the accurate classification and storage of scrap steel.

[0060] The specific implementation process for this step is as follows:

[0061] The system automatically links order information using license plate recognition technology. Equipped with high-precision license plate recognition equipment, the system quickly and accurately identifies the license plate number of a vehicle entering the unloading area and retrieves the corresponding order information from the database. This order information includes crucial data such as material type and weight, which is essential for subsequent scrap steel processing and management.

[0062] The system synchronously acquires material type and weight data and guides vehicles to the corresponding storage area in real time. Based on the material type and weight data in the order information, the system automatically calculates the storage area the vehicle should go to and guides the vehicle to the designated storage area in real time through voice prompts or displays. For example, if the vehicle is loaded with briquetted scrap steel, the system will prompt "Briquettes, please go to B3 Light Material Area" to ensure that the vehicle can reach the designated location quickly and accurately.

[0063] During the unloading process, a pre-built scrap steel identification model is used to identify the scrap steel in real time. Multiple cameras are installed in the unloading area to capture images of the scrap steel in real time, and the image data is transmitted to the scrap steel identification model. The model identifies and classifies the scrap steel in the images in real time, ensuring that the scrap steel is accurately stored in the corresponding storage area. If a mismatch is found between the type of scrap steel and the storage area (e.g., briquetted materials stored in the pig iron block area), the system will immediately intercept the operation and send a manual review reminder, preventing illegal storage from occurring at the source.

[0064] S3 takes real-time photos of the scrap steel during the unloading process and saves them to the server management system.

[0065] Among them, blockchain technology is used to put scrap steel photos, vehicle information, time-series data, composition data, and supplier information on the chain.

[0066] The specific implementation process for this step is as follows:

[0067] The system captures real-time images of the scrap steel. Multiple high-resolution cameras are installed in the unloading area to take real-time photos of the scrap steel from different angles during the unloading process. These cameras can capture detailed image information of the scrap steel, including its shape, size, color, and other characteristics.

[0068] The collected scrap steel images are saved to the server management system. The server management system will centrally manage and store the collected scrap steel images for easy subsequent querying and analysis. Simultaneously, the system will tag each image, recording metadata such as the image capture time, location, and corresponding vehicle information, enabling accurate traceability of the source of each image.

[0069] By employing blockchain technology, scrap steel photos, vehicle information, time-series data, composition data, and supplier information are uploaded to the blockchain. Blockchain technology, with its immutable and decentralized characteristics, ensures the authenticity and integrity of the data. The system packages the collected scrap steel photos, along with related vehicle information, time-series data, composition data, and supplier information, into a data block and uploads this block to the blockchain network. Each data block has a unique hash value, which allows for quick location and verification of the block's content. Thus, once the data is on the blockchain, it cannot be tampered with, achieving an immutable record of the entire lifecycle of scrap steel and providing reliable data support for subsequent quality traceability and liability determination.

[0070] S4 calls the constructed scrap steel identification model to analyze the collected scrap steel images and accurately inspect the quality of the captured scrap steel portions. When foreign objects or abnormalities are detected in the scrap steel inside the vehicle, the on-site signal lights remind the crane operator to conduct a secondary on-site confirmation, realizing full-process quality inspection tracking and control.

[0071] Finally, the system will call the constructed scrap steel recognition model to analyze the collected scrap steel images and perform the following operations:

[0072] Accurate quality inspection is performed on the collected scrap steel portions. The scrap steel identification model performs in-depth analysis of the collected scrap steel images, extracts the scrap steel's characteristic information, and evaluates the quality of the scrap steel based on a pre-trained classification model. The evaluation includes multiple aspects such as the type, size, weight, and composition of the scrap steel, ensuring that the quality of the scrap steel meets production requirements.

[0073] When foreign objects or abnormalities are detected in the scrap steel inside the crane, the system alerts the crane operator via on-site signal lights to conduct a secondary on-site verification. The system monitors the quality of the scrap steel in real time. If foreign objects (such as non-metallic substances or harmful impurities) are found, or if the scrap steel's size, weight, or other parameters are abnormal, the system will immediately trigger an alarm and alert the crane operator via on-site signal lights. Upon receiving the alert, the crane operator will pause operation and conduct a secondary on-site verification of the scrap steel to ensure that any quality issues are addressed promptly.

[0074] The system enables end-to-end quality tracking and control. It tracks and records the entire scrap steel processing process in real time, meticulously recording and monitoring each step from unloading and storage to feeding. By utilizing a scrap steel identification model to analyze and evaluate the scrap steel at each stage, the system ensures effective quality control throughout the entire process, thereby reducing production risks caused by scrap steel quality issues and improving production efficiency and product quality.

[0075] In optional embodiments of this application, it further includes:

[0076] When a scrap steel blending task is received, a reasonable blending plan with the lowest cost is generated by combining the purchase price, market price, and finished product selling price. The plan is then sent to the feeding crane to form a blending task. At the same time, a blacklist area and electronic fence are generated to prevent the feeding crane from entering irrelevant areas unnecessarily.

[0077] The constructed scrap steel identification model is invoked to identify scrap steel in real time during the feeding process, generate scrap steel type and grade information, and generate a batching progress bar based on the read weight information.

[0078] The following describes the specific implementation process of the above method, which includes:

[0079] Scrap steel identification model:

[0080] Multi-model feature extraction and fusion technology is used to identify scrap steel material types. Based on the existing scrap steel identification model, Mask R-CNN is used for image segmentation and target detection, and OpenCV is used to process the real-time video stream of the scrap steel yard.

[0081] Material type classification: Using ResNet or EfficientNet deep learning models, the training data contains tens of thousands of labeled scrap steel images (classified according to GB / T 4223 standard).

[0082] The integration of multi-model feature extraction technology can improve the accuracy of scrap steel type identification, providing advantages for the intelligent and efficient development of the scrap steel processing industry. The detection rate of individual scrap steel is as high as 95%, and the accuracy rate is as high as 93%. The model can identify thicknesses ranging from 0.5mm to over 20mm.

[0083] Warehouse Management System: When transport vehicles arrive at the unloading area, the system automatically associates order information through license plate recognition, synchronously obtains data such as material type and weight, and provides real-time guidance to the corresponding storage area (e.g., voice prompt "Please proceed to B3 Light Material Area for briquettes"). During unloading, the system monitors the crane's movement position in real time, combines it with preset storage area information, and intelligently judges whether it meets the requirements through visual technology. Abnormal situations immediately trigger audible and visual alarms to ensure accurate material placement. If a mismatch occurs (e.g., briquettes are stored in the pig iron block area), the system immediately intercepts the operation and sends a manual review reminder, preventing unauthorized storage from the source.

[0084] The system automatically generates inbound and outbound data archives, and all data is synchronized to the central database in real time. Core inventory information is presented in real time through data dashboards and status dashboards.

[0085] The intelligent warehouse management system achieves precise guidance of overhead cranes and reduces unnecessary movement time through the deep integration of technology and processes.

[0086] Dynamic Feeding Module: Upon receiving a scrap steel feeding task, the precise batching model combines purchase price, market price, and finished product selling price to generate a reasonable batching scheme with the lowest cost. This scheme is then sent to the feeding crane to form a feeding task. Simultaneously, a blacklist of feeding areas and electronic fences are generated to prohibit the feeding crane from unnecessarily entering irrelevant areas, forming the first line of control. Engineers can manually plan blacklist areas based on the on-site slag and reclaimed scrap steel storage areas.

[0087] The overhead crane operator operates according to the system-generated batching table and recommended feeding areas. Each time the crane adds material, the system identifies the scrap steel and generates information such as material type and grade. Simultaneously, it generates a batching progress bar based on the read weight information, enabling task sharing among multiple cranes. When other cranes have vacant tasks, they can use the progress bar to assist in batching, shortening the batching time. The system also identifies any suspicious materials entering the crane and verifies the matching degree between the added material and the batching table by combining the crane / weighbridge metering information. If mismatched materials are detected, the system issues an audible and visual alarm, prompting the crane operator to adjust. Once the system detects that the current scrap steel addition meets the required ratio, it prompts the crane operator to add the next piece of scrap steel. Once the batching is completed according to the batching table, the system automatically locks the scrap steel trough, prohibiting other cranes from adding any scrap steel to it, forming a second layer of control. At the same time, the system uploads the scrap steel trough's addition information to the smart steelmaking system and the crane intelligent management system.

[0088] The overhead crane operation records are linked to orders and responsible persons, enabling traceability of the entire operation process and supporting anomaly review and evaluation. Under a multi-condition, dual-control system, the accuracy of ingredient batching is improved, and the impact of abnormal factors on the quality of finished products is reduced.

[0089] In summary, the technical effects of this application include:

[0090] 1. Conduct more comprehensive sampling of scrap steel.

[0091] 2. The lack of dust detection during the overall scrap steel quality inspection process is a problem, as no large amount of dust is generated during the grabbing process; the overall inspection process cannot be improved because it is impossible to inspect the already grabbed scrap steel.

[0092] 3. Quality inspection inside the vehicle may result in some items being missed.

[0093] 4. Full-process quality inspection tracking and control.

[0094] The specific technical points of this application include:

[0095] 1. This application is for monitoring scrap steel yards and stockpiles, and providing real-time video display.

[0096] 2. This application aims to stitch together all the cameras to create a real-time view of the entire material yard.

[0097] 3. The portion of the scrap steel quality inspection process that is optimized in this application.

[0098] 4. This application applies to material yards that store scrap steel stockpiles, but are not limited to them.

[0099] Please refer to Figure 2 The diagram illustrates a block diagram of a scrap steel traceability and risk warning management system based on image recognition and blockchain, provided in an embodiment of this application. The system may include:

[0100] The module is used to build a scrap steel identification model, which is used to identify scrap steel in the unloading area. Specifically, the scrap steel identification model uses Mask R-CNN for image segmentation and target detection, combines OpenCV to process real-time video streams of the scrap steel yard, and uses a ResNet deep learning model for material type classification. The training data includes manually labeled scrap steel images.

[0101] The identification module is used to automatically associate order information through license plate recognition when the transport vehicle arrives at the unloading area, synchronously obtain material type and weight data, and guide the corresponding storage area in real time. During the unloading process, the constructed scrap steel identification model is called to identify the scrap steel in real time, ensuring the accurate classification and storage of scrap steel.

[0102] The data acquisition module is used to take real-time photos of the scrap steel during the unloading process and save them to the server management system. It uses blockchain technology to put the scrap steel photos, vehicle information, time-series data, composition data, and supplier information on the chain.

[0103] The confirmation module is used to call the constructed scrap steel identification model, analyze the collected scrap steel images, and accurately verify the quality of the captured scrap steel. When foreign objects or abnormalities are detected in the scrap steel inside the vehicle, the on-site signal lights remind the crane operator to conduct a secondary on-site confirmation, realizing full-process quality verification tracking and control.

[0104] Specific limitations regarding the scrap steel traceability and risk warning management system based on image recognition and blockchain can be found in the limitations of the scrap steel traceability and risk warning management method based on image recognition and blockchain mentioned above, and will not be repeated here. Each module in the aforementioned scrap steel traceability and risk warning management system based on image recognition and blockchain can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0105] In one embodiment, an electronic device is provided, which may be a computer, and its internal structure diagram may be as follows: Figure 3As shown, the electronic device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used for scrap steel traceability and risk warning management data based on image recognition and blockchain. The network interface of the computer device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a scrap steel traceability and risk warning management method based on image recognition and blockchain.

[0106] Those skilled in the art will understand that, Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0107] In one embodiment of this application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the above-described method for traceability and risk warning management of scrap steel based on image recognition and blockchain.

[0108] In one embodiment of this application, a computer program product is provided, including a computer program / instructions, which, when executed by a processor, implements the steps of the above-described method for scrap steel traceability and risk warning management based on image recognition and blockchain.

[0109] The computer-readable storage medium and computer program product provided in this embodiment are similar in implementation principle and technical effect to the above method embodiments, and will not be repeated here.

[0110] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in M ​​forms, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Memory Bus Direct RAM (RDRAM), Direct Memory Bus Dynamic RAM (DRDRAM), and Memory Bus Dynamic RAM (RDRAM), etc.

[0111] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0112] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for traceability and risk warning management of scrap steel based on image recognition and blockchain, characterized in that, The method includes: A scrap steel identification model is constructed and used to identify scrap steel in the unloading area. Specifically, the scrap steel identification model uses Mask R-CNN for image segmentation and target detection, combines OpenCV to process real-time video streams of the scrap steel yard, and uses a ResNet deep learning model for material type classification. The training data includes manually labeled scrap steel images. When the transport vehicle arrives at the unloading area, the order information is automatically linked through license plate recognition, and the material type and weight data are obtained simultaneously. The vehicle is guided to the corresponding storage area in real time. During the unloading process, the constructed scrap steel identification model is called to identify the scrap steel in real time, ensuring the accurate classification and storage of scrap steel. During the unloading of scrap steel, real-time photos of the scrap steel are taken and saved to the server management system. Blockchain technology is used to put the scrap steel photos, vehicle information, time-series data, composition data, and supplier information on the chain. The constructed scrap steel identification model is invoked to analyze the collected scrap steel images and accurately inspect the quality of the captured scrap steel portions. When foreign objects or abnormalities are detected in the scrap steel inside the vehicle, the on-site signal lights remind the crane operator to conduct a secondary on-site confirmation, thereby achieving full-process quality inspection tracking and control.

2. The method according to claim 1, characterized in that, The method further includes: When a scrap steel blending task is received, a reasonable blending plan with the lowest cost is generated by combining the purchase price, market price, and finished product selling price. The plan is then sent to the feeding crane to form a blending task. At the same time, a blacklist area and electronic fence are generated to prevent the feeding crane from entering irrelevant areas unnecessarily. The constructed scrap steel identification model is invoked to identify scrap steel in real time during the feeding process, generate scrap steel type and grade information, and generate a batching progress bar based on the read weight information.

3. The method according to claim 1, characterized in that, Building a scrap steel identification model also includes: The training data is preprocessed, including grayscale conversion, binarization, and edge detection, to enhance image features; Data augmentation techniques are used to expand the training dataset; these techniques include random cropping, rotation, and flipping. Cross-validation was used to evaluate the model performance.

4. The method according to claim 1, characterized in that, Automatically linking order information through license plate recognition also includes: Verify the license plate recognition results to ensure the accuracy of the license plate number; Add vehicle arrival timestamp to order information; When license plate recognition fails, a manual intervention process is automatically triggered to ensure the correct association of order information.

5. The method according to claim 1, characterized in that, The steps for real-time photo capture of scrap steel also include: Use cameras at multiple angles to take pictures to ensure comprehensive coverage of the scrap steel; The system transmits and stores captured images in real time. During the photo-taking process, it automatically detects lighting conditions and adjusts camera parameters automatically when the lighting is insufficient to ensure image quality.

6. The method according to claim 1, characterized in that, The analysis, which utilizes the scrap steel identification model, also includes: The collected scrap steel images are preprocessed, including background noise removal and image size standardization; Use deep learning models to extract features and classify preprocessed images; When an anomaly is detected, the anomaly information is automatically recorded and an anomaly report is generated.

7. A scrap steel traceability and risk warning management system based on image recognition and blockchain, characterized in that, The system includes: The module is used to build a scrap steel identification model, which is used to identify scrap steel in the unloading area. Specifically, the scrap steel identification model uses Mask R-CNN for image segmentation and target detection, combines OpenCV to process real-time video streams of the scrap steel yard, and uses a ResNet deep learning model for material type classification. The training data includes manually labeled scrap steel images. The identification module is used to automatically associate order information through license plate recognition when the transport vehicle arrives at the unloading area, synchronously obtain material type and weight data, and guide the corresponding storage area in real time. During the unloading process, the constructed scrap steel identification model is called to identify the scrap steel in real time, ensuring the accurate classification and storage of scrap steel. The data acquisition module is used to take real-time photos of the scrap steel during the unloading process and save them to the server management system. It uses blockchain technology to put the scrap steel photos, vehicle information, time-series data, composition data, and supplier information on the chain. The confirmation module is used to call the constructed scrap steel identification model, analyze the collected scrap steel images, and accurately verify the quality of the captured scrap steel. When foreign objects or abnormalities are detected in the scrap steel inside the vehicle, the on-site signal lights remind the crane operator to conduct a secondary on-site confirmation, realizing full-process quality verification tracking and control.

8. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, implements the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 6.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the method described in any one of claims 1 to 6.

Images