Bar tracking method, system and equipment based on image recognition and storage medium

Abstract

The invention relates to the technical field of data processing, and particularly provides a bar tracking method, system and equipment based on image recognition and a storage medium, and the method comprises the steps: collecting blank numbers, equipment signals and time data of a steel rolling production line PLC and an MES system, automatically generating bar full-flow tracking information, and pre-distributing a unique code sequence; in the cooling bed link, the number of material supporting position section material groups is counted through the image recognition technology, section material group codes are distributed according to the first-in first-out principle in combination with an input roller way coding sequence, and blank numbers are bound; in the saw cutting link, the bar output roller way position sequence is recognized through images, codes and blank numbers distributed by a cooling bed are associated, and a unique sub-code is generated for each bar. The error rate and the labor intensity of traditional manual recording and cooperation are greatly reduced, and the production efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of data processing, and particularly relates to a bar tracking method, system and device based on image recognition and a storage medium. BACKGROUND

[0002] In the production process of rolled steel bars, achieving accurate tracking and quality tracing of each bar from raw material to finished product is the key to improving product quality and management level. However, in the cooling bed and sawing link, after the material is grouped and cut, the bar of multiple billet batches is easily mixed, resulting in the breakage of the corresponding relationship between the section bar and the original billet number, and the tracking is chaotic. The existing technology cannot bind a unique identity information for each final bar on a high-speed and continuous production line, which seriously restricts the improvement of accurate quality tracing and digital management level. Therefore, a new method capable of achieving accurate tracking of the whole process of the bar at the single level is urgently needed. SUMMARY

[0003] In view of the above deficiencies of the prior art, the present application provides a bar tracking method, system and device based on image recognition and a storage medium to solve the above technical problems.

[0004] In a first aspect, the present application provides a bar tracking method based on image recognition, comprising: collecting equipment data of the whole process of the rolled steel production line, wherein the equipment data includes billet numbers, equipment entering signals, equipment leaving signals and signal times from PLC and MES systems; generating tracking information of the bar at each link of the production line based on the equipment data, and pre-generating a unique coding sequence for the bar, wherein the unique coding sequence contains multiple unique codes; in the cooling bed link, identifying the number of section bar groups on the cooling bed material supporting position through image recognition technology, and based on the number of section bar groups and the pre-generated unique coding sequence of the input roller, assigning a unique code to each group of section bars according to the first-in first-out principle and binding the billet number; in the sawing link, identifying the position sequence of the bar after sawing on the output roller through image recognition technology; and based on the position sequence, the unique code assigned to the section bar group in the cooling bed link and the bound billet number, generating a unique sub-code for each bar; wherein the unique sub-code is used to realize the whole process tracking of the bar.

[0005] In an optional embodiment, collecting equipment data of the whole process of the rolled steel production line comprises: real-time collecting data including billet numbers, furnace number, steel grade, equipment entering signals, equipment leaving signals, cutting section times, transverse direction and corresponding signal times from the PLC and MES systems of the heating furnace, blooming mill, continuous rolling mill, flying shear, large cooling bed and sawing machine of the rolled steel production line.

[0006] In an optional embodiment, based on the equipment data, tracking information of the bar at each link of the production line is generated, and a unique coding sequence is pre-generated for the bar, including: Based on the equipment data, tracking information containing the billet number, the entering time and leaving time of each link equipment, the number of cutting sections, and the transverse movement direction is generated; Based on the tracking information of each bar, a unique code for the bar is generated according to a predetermined coding rule; wherein the coding rule is to combine the billet number of a piece of tracking information, the identifier of a key process link, and the sequential number in the link to form a unique code; Arrange multiple unique codes into a unique coding sequence.

[0007] In an optional embodiment, at the cooling bed link, the number of segment groups on the cooling bed material supporting position is identified through image recognition technology, and based on the number of segment groups and the pre-generated unique coding sequence of the input roller, each group of segment is assigned a unique code and bound to the billet number according to the first-in first-out principle, including: Collect segment group images through industrial cameras deployed above the input roller of the cooling bed; use a target detection and instance segmentation algorithm based on deep learning to analyze the images, identify and count the number of segment groups stably present on the current material supporting position; The number N of segment groups identified by image recognition is used as the trigger condition and quantity basis for sequentially taking N unique codes from the head of the pre-generated unique coding sequence; Bind the N unique codes taken out with the N groups of physical segments currently located on the cooling bed material supporting position to confirm the billet number information of the N groups of segments; After the binding is completed, the N unique codes are removed from the sequence, and the unique coding sequence is updated.

[0008] In an optional embodiment, the target detection and instance segmentation algorithm based on deep learning is used to analyze the images, identify and count the number of segment groups stably present on the current material supporting position, including: Preprocess the collected segment group images to eliminate image interference caused by environmental reflections, dust, or material movement; Input the preprocessed images into a trained deep learning model; the deep learning model locates and frames each segment group in the image through its target detection branch, and generates a pixel-level accurate segmentation mask for each framed segment group through its instance segmentation branch; Based on the segmentation mask, distinguish and count the segment group entities that are independent and complete in outline, thereby outputting the number of segment groups on the current material supporting position.

[0009] In an optional implementation, during the sawing process, image recognition technology is used to identify the positional order of the sawn bars on the output roller conveyor; and based on the positional order and the unique code assigned to the section group in the cooling bed stage and the bound billet number, a unique sub-code is generated for each bar, including: Images of the bars are acquired by a vision system deployed above the sawing output rollers; based on the images of the bars, the contour and spatial coordinates of each bar on the rollers are identified and determined, and the bar position order is generated according to a predetermined sorting rule. Based on the unique code of the segment group and the billet number from the cooling bed stage, and the total number of times the segment group was sawed obtained from the PLC and MES system of the sawing machine, a unique sub-code with a common parent code is generated for the M bars corresponding to the segment group. The set of unique sub-codes is bound one-to-one with the M physical bars according to the positional order of the bars.

[0010] In one optional implementation, based on the image of the bars, the contour and spatial coordinates of each bar on the roller conveyor are identified and determined, and the bar position order is generated according to a predetermined sorting rule, including: The acquired bar images are preprocessed to enhance the contrast between the bar and the background, and the contour of the bar end is extracted. Based on the extracted contours, the position of each bar in the image is determined, and its spatial coordinates in the image coordinate system are calculated. Based on the spatial coordinates, all bars are sorted in a manner parallel to the roller conveyor running direction to generate the bar position sequence.

[0011] Secondly, the present invention provides a bar tracking system based on image recognition, comprising: The data acquisition module is used to collect equipment data throughout the entire steel rolling production line. The equipment data includes billet number, equipment entry signal, equipment exit signal, and signal time from the PLC and MES systems. The coding definition module is used to generate tracking information of the bar stock at each stage of the production line based on the equipment data, and to pre-generate a unique coding sequence for the bar stock, wherein the unique coding sequence contains multiple unique codes; The coding allocation module is used to identify the number of material groups on the material support position of the cooling bed through image recognition technology in the cooling bed stage, and to assign a unique code to each material group and bind the billet number according to the first-in-first-out principle based on the number of material groups and the unique coding sequence of the pre-generated input roller. The sub-code generation module is used to identify the position sequence of the sawn bars on the output roller conveyor through image recognition technology during the sawing process; and to generate a unique sub-code for each bar based on the position sequence and the unique code assigned to the segment group in the cooling bed process and the bound billet number. The unique sub-code is used to achieve full-process tracking of the bar stock.

[0012] Thirdly, a device is provided, comprising: Memory for storing bar tracking programs based on image recognition; A processor is configured to implement the steps of the image recognition-based bar tracking method provided in the first aspect when executing the image recognition-based bar tracking program.

[0013] Fourthly, a computer-readable storage medium is provided, on which an image recognition-based bar tracking program is stored, wherein when the image recognition-based bar tracking program is executed by a processor, the steps of the image recognition-based bar tracking method provided in the first aspect are implemented.

[0014] The beneficial effects of this invention are that the image recognition-based bar tracking method, system, equipment, and storage medium provided by this invention, by integrating production line equipment data with image recognition technology, achieve precise tracking of bars at the single-bar level throughout the entire process from billet to finished product. Its beneficial effects are: it fundamentally solves the industry problem of mixed cooling bed assembly and loss of billet number information after sawing, ensuring that each bar can be accurately traced back to the original billet, providing a reliable data foundation for quality traceability. It constructs an automated tracking closed loop of "data-driven, visual verification," significantly reducing the error rate and labor intensity of traditional manual recording and collaboration, and improving production efficiency. It provides strong support for lean production, process optimization, and digital management, significantly improving the level of intelligence in steel rolling production. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic flowchart of a method according to an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram illustrating the cooling bed stage of a method according to an embodiment of the present invention.

[0018] Figure 3 This is a schematic diagram illustrating the sawing process of a method according to an embodiment of the present invention.

[0019] Figure 4 This is a schematic flowchart illustrating the entire process of tracking a method according to an embodiment of the present invention.

[0020] Figure 5 This is a schematic block diagram of a system according to an embodiment of the present invention.

[0021] Figure 6 This is a schematic diagram of the structure of a device provided in an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0024] The bar tracking method based on image recognition provided in this embodiment of the invention is executed by a computer device, and correspondingly, the bar tracking system based on image recognition runs in the computer device.

[0025] Figure 1 This is a schematic flowchart illustrating a method according to an embodiment of the present invention. Wherein, Figure 1 The executing entity can be an image recognition-based bar tracking system. Depending on different requirements, the order of steps in this flowchart can be changed, and some can be omitted.

[0026] like Figure 1 As shown, the method includes: S1. Collect equipment data throughout the entire steel rolling production line. The equipment data includes billet number, equipment entry signal, equipment exit signal, and signal time from the PLC and MES systems. S2. Based on the equipment data, generate tracking information for the bar stock at each stage of the production line, and pre-generate a unique coding sequence for the bar stock, wherein the unique coding sequence contains multiple unique codes; S3. In the cooling bed stage, the number of material segments on the cooling bed support position is identified by image recognition technology, and based on the number of material segments and the unique coding sequence of the pre-generated input roller, a unique code is assigned to each material segment according to the first-in-first-out principle and the billet number is bound. S4. In the sawing process, the position sequence of the sawn bars on the output roller conveyor is identified by image recognition technology; and based on the position sequence and the unique code assigned to the section group in the cooling bed process and the bound billet number, a unique sub-code is generated for each bar. The unique sub-code is used to achieve full-process tracking of the bar stock.

[0027] In one embodiment of the present invention, based on step S1, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0028] The PLC and MES systems of the heating furnace, billet mill, continuous rolling mill, flying shear, large cooling bed and sawing machine of the steel rolling production line collect data in real time, including billet number, furnace number, steel grade, equipment entry signal, equipment exit signal, number of cuts, lateral movement direction and corresponding signal time.

[0029] In one embodiment of the present invention, based on step S2, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0030] S201. Data Preparation and Tracking Information Generation The data processing module collects real-time data from PLCs and MES systems at key stages of the rolling mill production line (including heating furnaces, billet mills, and flying shears) via standard industrial communication protocols (such as OPC UA). After acquiring the raw data stream, the module parses, cleans, and correlates it, dynamically generating and maintaining a tracking record for each bar stock (and its subsequent segments). This tracking record is a structured dataset, with core fields including at least: Billet number: such as 123456-12345, used to identify the original steel billet.

[0031] Timestamps: Include the precise entry and exit times for each device.

[0032] Process events: such as the number of cuts, the direction of lateral movement (left / right), etc.

[0033] S202. Application of Encoding Rules and Generation of Unique Codes The system has a pre-set set of flexible and configurable coding rules. When the bar stock passes through a critical process node (such as the completion of billet opening or flying shearing), the data processing module calls this rule to extract key elements from the current tracking information of the bar stock and combine them to generate a unique code.

[0034] The core of the coding rule is to combine the blank number, the identifier of the key process step, and the sequential number within that step.

[0035] Example of internal coding: For billet numbers 123456-12345, the identifier in the billet opening stage is 2, the lateral movement direction of the offset group is right (identifier R), and the sequential number after the flying shear segment is 1. Then the generated unique internal code is: 12345612345-2R-1. This code accurately records the critical path of the material in the production line.

[0036] S203. Management of Unique Encoding Sequences At specific stages (such as before entering the cooling bed), the system arranges multiple pre-generated unique codes into an ordered unique code sequence for a batch of continuously produced material segments, according to their generation sequence. This sequence is typically maintained in system memory or a database using a queue data structure, strictly adhering to the "first-in, first-out" (FIFO) time sequence logic. This sequence serves as the data foundation for subsequent automatic matching and binding of codes with physical material groups in the cooling bed stage, ensuring consistency between information flow and physical flow.

[0037] In one embodiment of the present invention, based on step S3, the following will provide a possible embodiment and its specific implementation will be described in a non-limiting manner, referring to... Figure 2 .

[0038] S301. Acquire images of the material segments using an industrial camera deployed above the input roller conveyor of the cooling bed; analyze the images using a deep learning-based target detection and instance segmentation algorithm to identify and count the number of material segments that are stably present at the current material support position; 1. Image acquisition and preprocessing First, a high-resolution industrial area scan camera deployed above the input roller conveyor of the cooling bed acquires images of the original steel section group under the control of a specific trigger signal (such as a section arrival signal from the PLC). To address the challenges commonly encountered in steel rolling mills, such as strong glare, dust, and material movement, the system performs a series of image preprocessing algorithms before and after acquisition: Illumination normalization: An adaptive histogram equalization algorithm is used to balance the overall illumination of the image and reduce the effects of local reflections, overexposure, and shadows.

[0039] Filtering and noise reduction: Gaussian filtering or median filtering algorithms are applied to effectively suppress dust noise in the image while preserving the edge information of the material.

[0040] Motion blur compensation: If the material is in slight motion, a high-frequency pulsed LED light source is used for "freezing" photography in terms of hardware; in terms of software, an image deblurring algorithm can be used for post-processing enhancement.

[0041] After preprocessing, an image with enhanced contrast, reduced noise, and clearer contours is obtained, laying the foundation for subsequent depth analysis.

[0042] 2. Deep learning model recognition and segmentation The preprocessed image is input into a pre-trained deep learning model. This model employs an advanced architecture that integrates object detection and instance segmentation (e.g., a framework based on Mask R-CNN).

[0043] The object detection branch is responsible for quickly and accurately locating all segments in the image and selecting them with bounding boxes. This step answers the question, "Where are the segments?"

[0044] Instance segmentation branch: Within each detection box, pixel-level fine analysis is performed to generate an accurate segmentation mask for each individual segment group. This mask can accurately delineate the outline of the segment group, thereby distinguishing mutually adhered or adjacent segment groups at the pixel level.

[0045] The model, trained on a large dataset of segment images labeled with bounding boxes and pixel-level masks, has developed strong recognition capabilities and generalization ability.

[0046] 3. Statistics and Output Based on Segmentation Masks The system receives all segmentation masks returned by the model and performs post-processing analysis: Perform connected component analysis on each segmentation mask to ensure that each mask represents an independent entity.

[0047] By calculating the geometric properties of the mask (such as area and circumscribed rectangle), the tiny invalid areas caused by noise are filtered out, ensuring that only segments with complete outlines and reasonable dimensions are counted.

[0048] Finally, the number of valid segmented masks that have been verified is directly counted. This number is the number N of segment groups that are stably present at the current material support position, and it is output as key data to the subsequent encoding and binding module.

[0049] S302. The number N of material segments obtained from image recognition is used as the triggering condition and quantity basis for sequentially extracting N unique codes from the head of the pre-generated unique coding sequence.

[0050] The number N of the material segments is transmitted in real time to the core processing unit of the system. This unit maintains a queue of pre-generated unique codes based on the First-In-First-Out (FIFO) principle. The arrival of the number N immediately serves as a trigger signal and allocation basis. The system then sequentially retrieves N unique codes (e.g., in the format "Bill Number - Flying Shear Segment Number") starting from the head of the queue. This step completes the logical matching from physical world perceived information to digital codes.

[0051] S303. Bind the extracted N unique codes to the N groups of physical segments currently located on the cooling bed support position to confirm the billet number information of the N groups of segments.

[0052] The system logically binds the N unique codes retrieved to N groups of physical material sections located on the cooling bed, as confirmed by the image. Essentially, this binding involves assigning all the information carried by these N codes—including billet number, heat number, and steel grade—to the corresponding N groups of physical material sections in the system's tracking database. Thus, each group of physical material sections acquires its unique "digital identity," completing the transformation from anonymous material to traceable entity.

[0053] S304. After binding is complete, remove the N unique codes from the sequence and update the unique code sequence.

[0054] Once the binding operation is confirmed, the system immediately removes the assigned N unique codes from the previously pre-generated queue. This operation ensures that the queue remains consistent with the status of the physical material to be tracked, updates the unique code sequence, and prepares for the next round of cold bed loading and code binding, thus forming a continuous, closed-loop automated tracking process.

[0055] In one embodiment of the present invention, based on step S4, the following will provide a possible embodiment and its specific implementation will be described in a non-limiting manner, referring to... Figure 3 .

[0056] S401. Acquire images of the bars by means of a vision system deployed above the sawing output rollers; based on the images of the bars, identify and determine the contour and spatial coordinates of each bar on the rollers, and generate the bar position order according to a predetermined sorting rule.

[0057] 1. Image preprocessing and contour extraction After being sawn, the bars are conveyed to the output roller conveyor, where digital images are captured by industrial cameras deployed above the roller conveyor. Due to uneven lighting, oxide scale on the bar surface, and background interference, the system first preprocesses the original images: a contrast-limited adaptive histogram equalization algorithm is used to enhance the contrast between the bars and the background; then, a Gaussian filtering algorithm is applied to suppress image noise. After preprocessing, the system uses the Canny edge detection algorithm to extract the edge contours of all bar ends in the image, and uses a contour fitting algorithm to select closed contours that conform to the geometric characteristics of the bars.

[0058] 2. Coordinate positioning and data conversion For each extracted bar end profile, the system calculates its minimum bounding rectangle and uses the pixel coordinates of the midpoint of the bottom edge of this rectangle as the position of the bar in the image. Through pre-completed camera calibration, the system converts the pixel coordinates into spatial coordinates in a two-dimensional image coordinate system with one side of the roller conveyor as the origin. These coordinates accurately reflect the actual position of each bar in the width direction of the roller conveyor.

[0059] 3. Spatial sorting and sequence generation After collecting the spatial coordinates of all bars, the system sorts them parallel to the roller conveyor's running direction (i.e., along the roller conveyor axis). The specific sorting rule is as follows: compare the x-coordinate values ​​of each bar's spatial coordinates, and arrange all bars in ascending order from left to right (or from right to left, consistent with the actual production line layout), ultimately generating a list of bar positions with a clear sequential relationship. This list serves as the direct basis for subsequent code binding, ensuring a one-to-one correspondence between virtual codes and physical bars.

[0060] S402. Based on the unique code of the segment group and the billet number from the cooling bed stage, and the total number of times the segment group has been sawn obtained from the PLC and MES systems of the sawing machine, generate a unique sub-code with a common parent code for the M bars corresponding to the segment group.

[0061] The system retrieves the unique code (e.g., "12345612345-2R-1") and corresponding billet number of the material group bound in the cooling bed stage from the database. Simultaneously, it obtains the sawing count M of this material group on the sawing machine in real time through the MES system. Based on this data, the system automatically generates M ordered unique sub-codes, using the material group code as the parent code. The coding rule adopts the format of "parent code-sequence number," for example, generating the coding sequence ["12345612345-2R-1-1", "12345612345-2R-1-2", ..., "12345612345-2R-1-M"].

[0062] S403. Bind the set of unique sub-codes to the M physical bars according to the positional order of the bars.

[0063] A strict one-to-one mapping relationship is established between the generated M unique sub-codes and the bar position sequence obtained in S401: the first bar in the position sequence is bound to the first sub-code, the second bar is bound to the second sub-code, and so on. After binding, the system establishes a complete bar identity record in the tracking database, including the unique sub-code, billet number, binding timestamp, and final position information in the material frame, thereby realizing accurate information inheritance and full-process tracking from section material to individual bar.

[0064] Please refer to Figure 4 This embodiment takes a specific steel billet (bill number: 01234501234) as an example, combined with... Figure 1 The diagram shows the entire process of a steel rolling production line, detailing the bar stock tracking process based on image recognition and production line data: 1. From raw materials to rough rolling stage (heating furnace → billet opening → continuous rolling) When billet 01234501234 enters the heating furnace, the MES system records its basic information such as heat number and steel grade. When the billet leaves the heating furnace, the system obtains its exit timestamp through PLC signals and generates an initial tracking record. As the billet passes through the billet mill and continuous rolling mill in sequence, the system establishes a continuous time sequence chain through the entry / exit signals of each device, forming a complete material flow record of "heating furnace → billet mill → continuous rolling mill".

[0065] 2. Segmentation and Path Assignment Stage (Flying Shear → Side Group Lateral Movement) The billet is cut into multiple-length segments at the flying shear. The system generates unique codes for four segments based on the segment cutting signal: 01234501234-1, 01234501234-2, 01234501234-3, and 01234501234-4. Using the left and right movement signals of the offset group's lateral movement, the system determines that segments 1 and 2 enter the finishing rolling path, while segments 3 and 4 directly enter the sawing path, which is reflected in the coding as 01234501234-2R-1, 01234501234-2R-2, etc.

[0066] 3. Visual recognition and coding binding in the cold bed process When the material segments enter the input roller conveyor of the large cooling bed, the system has pre-generated a coding sequence [01234501234-2R-1, 01234501234-2R-2, 01234501234-4L-1, ...]. An industrial camera deployed above the cooling bed captures images of the material segments, and a deep learning algorithm identifies three material segments at the current support position. The system immediately retrieves three codes sequentially from the head of the coding sequence, binds them to the physical material segments, and updates the coding sequence.

[0067] 4. Precise positioning and sub-code generation in the sawing process The lumber group with binding code 01234501234-2R-1 enters the sawing machine and is cut into 6 standard-length bars. The system uses a vision system above the sawing output roller conveyor. Acquire images of the end of the bar, extract the contour, and calculate the spatial coordinates; The six bars are sorted according to the parallel roller conveyor direction to generate the position sequence [Pos1, Pos2, ..., Pos6]. Using the segment group code as the parent, generate 6 sub-codes: 01234501234-2R-1-1 to 01234501234-2R-1-6; By binding sub-encoders to position sequences one by one, precise tracking from group to root is achieved.

[0068] 5. Finished product collection and data closed loop After sawing, the bars are transported to the collection box via a traverse vehicle. The system displays the coding information of each group of bars in the collection box through a graphical interface to guide on-site operations. Finally, in the collection box stage, each bar has complete identity information inherited from the raw material, forming a traceable chain of "Bill No. 01234501234 → Segment Group 01234501234-2R-1 → Bar 01234501234-2R-1-3", realizing precise tracking of the entire process from the heating furnace to the collection box.

[0069] In some embodiments, the image recognition-based bar tracking system may include multiple functional modules composed of computer program segments. The computer programs for each program segment in the image recognition-based bar tracking system may be stored in the memory of a computer device and executed by at least one processor to perform (see details). Figure 1 (Description) A bar tracking function based on image recognition.

[0070] In this embodiment, the image recognition-based bar tracking system can be divided into multiple functional modules according to its functions, such as... Figure 5 As shown. The module referred to in this invention is a series of computer program segments that can be executed by at least one processor and perform a fixed function, and is stored in memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0071] The data acquisition module is used to collect equipment data throughout the entire steel rolling production line. The equipment data includes billet number, equipment entry signal, equipment exit signal, and signal time from the PLC and MES systems. The coding definition module is used to generate tracking information of the bar stock at each stage of the production line based on the equipment data, and to pre-generate a unique coding sequence for the bar stock, wherein the unique coding sequence contains multiple unique codes; The coding allocation module is used to identify the number of material groups on the material support position of the cooling bed through image recognition technology in the cooling bed stage, and to assign a unique code to each material group and bind the billet number according to the first-in-first-out principle based on the number of material groups and the unique coding sequence of the pre-generated input roller. The sub-code generation module is used to identify the position sequence of the sawn bars on the output roller conveyor through image recognition technology during the sawing process; and to generate a unique sub-code for each bar based on the position sequence and the unique code assigned to the segment group in the cooling bed process and the bound billet number. The unique sub-code is used to achieve full-process tracking of the bar stock.

[0072] Figure 6 The image recognition-based bar tracking method provided in this application embodiment can be applied to a device. Those skilled in the art will understand that the device structure involved in the embodiments of this invention does not constitute a limitation on the device. The device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. In the embodiments of this invention, the device includes, but is not limited to, laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of this application described and / or claimed herein.

[0073] The device 600 may include a processor 610, a memory 620, and a communication unit 630. These components communicate via one or more buses. Those skilled in the art will understand that the server structure shown in the figures does not constitute a limitation of the present invention. It may be a bus topology or a star topology, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0074] The present invention also provides a computer storage medium, wherein the computer storage medium may store a program, which, when executed, may include some or all of the steps provided in the embodiments of the present invention. The storage medium may be a magnetic disk, an optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0075] Those skilled in the art will clearly understand that the techniques in the embodiments of the present invention can be implemented using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, or any other medium capable of storing program code. It includes several instructions to cause a computer device (which may be a personal computer, a server, or a second device, network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0076] The same or similar parts between the various embodiments in this specification can be referred to mutually. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple, and the relevant parts can be referred to the description in the method embodiments.

[0077] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or modules may be electrical, mechanical, or other forms.

[0078] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0079] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0080] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the present invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the present invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should also be covered within the protection scope of the present invention.

Claims

1. A bar tracking method based on image recognition, characterized in that, include: Collect equipment data throughout the entire steel rolling production line. The equipment data includes billet number, equipment entry signal, equipment exit signal, and signal time from the PLC and MES systems. Based on the equipment data, tracking information of the bar stock at each stage of the production line is generated, and a unique coding sequence is pre-generated for the bar stock, the unique coding sequence containing multiple unique codes; In the cooling bed stage, the number of material segments on the cooling bed support position is identified by image recognition technology. Based on the number of material segments and the unique coding sequence of the pre-generated input roller, a unique code is assigned to each material segment according to the first-in-first-out principle and the billet number is bound. In the sawing process, image recognition technology is used to identify the positional order of the sawn bars on the output roller conveyor; and based on the positional order and the unique code assigned to the section group in the cooling bed process and the bound billet number, a unique sub-code is generated for each bar. The unique sub-code is used to achieve full-process tracking of the bar stock.

2. The method according to claim 1, characterized in that, Collect equipment data throughout the entire steel rolling production line, including: The PLC and MES systems of the heating furnace, billet mill, continuous rolling mill, flying shear, large cooling bed and sawing machine of the steel rolling production line collect data in real time, including billet number, furnace number, steel grade, equipment entry signal, equipment exit signal, number of cuts, lateral movement direction and corresponding signal time.

3. The method according to claim 1, characterized in that, Based on the equipment data, tracking information for the bar stock at each stage of the production line is generated, and a unique coding sequence is pre-generated for the bar stock, including: Based on the equipment data, tracking information including billet number, equipment entry time and exit time at each stage, number of cuts and lateral movement direction is generated; Based on the tracking information of each bar stock, a unique code for the bar stock is generated according to a predetermined coding rule; wherein, the coding rule is: to combine the billet number of a tracking information, the identifier of the key process step, and the sequence number within that step to form a unique code; Arrange multiple unique codes into a unique code sequence.

4. The method according to claim 1, characterized in that, In the cooling bed stage, image recognition technology is used to identify the number of material segments on the cooling bed support. Based on the number of material segments and the unique coding sequence of the pre-generated input roller conveyor, a unique code is assigned to each material segment according to the first-in, first-out principle, and the billet number is bound to it. This includes: Images of the material segments are captured by an industrial camera deployed above the input roller conveyor of the cooling bed; the images are analyzed using a deep learning-based target detection and instance segmentation algorithm to identify and count the number of material segments that are stably present at the current material support position. The number N of material segments obtained from image recognition is used as the triggering condition and basis for extracting N unique codes in sequence from the head of the pre-generated unique coding sequence. The extracted N unique codes are bound to the N groups of physical segments currently located at the cooling bed support position to confirm the billet number information of the N groups of segments; After binding is complete, the N unique codes are removed from the sequence, and the unique code sequence is updated.

5. The method according to claim 4, characterized in that, The image is analyzed using a deep learning-based object detection and instance segmentation algorithm to identify and count the number of stable material segments currently present at the material receiving position, including: The acquired images of the material segments are preprocessed to eliminate image interference caused by environmental reflections, dust, or material movement. The preprocessed image is input into a trained deep learning model; the deep learning model locates and selects each segment group in the image through its object detection branch, and generates a pixel-level accurate segmentation mask for each selected segment group through its instance segmentation branch. Based on the segmentation mask, the independent and complete segment groups are distinguished and counted, thereby outputting the number of segment groups on the current material support position.

6. The method according to claim 1, characterized in that, In the sawing process, image recognition technology is used to identify the positional order of the sawn bars on the output roller conveyor; and based on the positional order and the unique code assigned to the section group in the cooling bed stage and the bound billet number, a unique sub-code is generated for each bar, including: Images of the bars are acquired by a vision system deployed above the sawing output rollers; based on the images of the bars, the contour and spatial coordinates of each bar on the rollers are identified and determined, and the bar position order is generated according to a predetermined sorting rule. Based on the unique code of the segment group and the billet number from the cooling bed stage, and the total number of times the segment group was sawed obtained from the PLC and MES system of the sawing machine, a unique sub-code with a common parent code is generated for the M bars corresponding to the segment group. The set of unique sub-codes is bound one-to-one with the physical M bars according to the positional order of the bars.

7. The method according to claim 6, characterized in that, Based on the images of the bars, the contour and spatial coordinates of each bar on the roller conveyor are identified and determined, and the bar position order is generated according to a predetermined sorting rule, including: The acquired bar images are preprocessed to enhance the contrast between the bar and the background, and the contour of the bar end is extracted. Based on the extracted contours, the position of each bar in the image is determined, and its spatial coordinates in the image coordinate system are calculated. Based on the spatial coordinates, all bars are sorted in a manner parallel to the roller conveyor running direction to generate the bar position sequence.

8. A bar tracking system based on image recognition, characterized in that, include: The data acquisition module is used to collect equipment data throughout the entire steel rolling production line. The equipment data includes billet number, equipment entry signal, equipment exit signal, and signal time from the PLC and MES systems. The coding definition module is used to generate tracking information of the bar stock at each stage of the production line based on the equipment data, and to pre-generate a unique coding sequence for the bar stock, wherein the unique coding sequence contains multiple unique codes; The coding allocation module is used to identify the number of material groups on the material support position of the cooling bed through image recognition technology in the cooling bed stage, and to assign a unique code to each material group and bind the billet number according to the first-in-first-out principle based on the number of material groups and the unique coding sequence of the pre-generated input roller. The sub-code generation module is used to identify the position sequence of the sawn bars on the output roller conveyor through image recognition technology during the sawing process; and to generate a unique sub-code for each bar based on the position sequence and the unique code assigned to the segment group in the cooling bed process and the bound billet number. The unique sub-code is used to achieve full-process tracking of the bar stock.

9. A bar tracking device based on image recognition, characterized in that, include: Memory for storing bar tracking programs based on image recognition; A processor, configured to implement the steps of the image recognition-based bar tracking method as described in any one of claims 1-7 when executing the image recognition-based bar tracking program.

10. A computer-readable storage medium storing a computer program, characterized in that, The readable storage medium stores an image recognition-based bar tracking program, which, when executed by a processor, implements the steps of the image recognition-based bar tracking method as described in any one of claims 1-7.

Images