Precise identification method and device for automobile part machining

Abstract

The invention relates to the field of part machining recognition, in particular to a precise recognition method and device for automobile part machining. The method comprises the following steps: obtaining a part processing monitoring video, and carrying out dynamic video imaging parameter adjustment and full video delay synchronization processing to obtain a delay synchronization optimization video; according to the delay synchronization optimization video, carrying out visual identification on the micro structure of the part, and extracting a three-dimensional morphological parameter of each feature point; performing frame-by-frame optical flow change tracking on the delay synchronous optimization video, and performing morphological parameter change calculation on the three-dimensional morphological parameters to obtain an overall processing change state of the part; and processing path change logic mining is carried out according to the delay synchronous optimization video, three-dimensional time sequence change modeling is carried out on the overall processing change state of the part, and a real-time processing change model of the part is constructed. The machining precision and production efficiency of parts are improved, the rejection rate is remarkably reduced, and the product quality is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of parts processing and identification, and in particular to a precise identification method and device for automobile parts processing. Background Art

[0002] With the continuous development of the manufacturing industry, especially the rapid advancement of the automotive industry, the demand for precision machining of automotive parts is increasing year by year. In modern automobile production, the quality of parts directly impacts the performance, safety, and durability of the entire vehicle. Therefore, ensuring that every part meets precise dimensional requirements, surface quality, and structural strength during machining has become a key issue in automotive manufacturing. Traditional automotive parts machining relies on manual operations and traditional inspection methods, but these methods often fail to meet the requirements of efficiency, precision, and automation, easily leading to machining errors and unstable product quality.

[0003] Especially with the large-scale production of automotive parts and their stringent requirements for safety, durability, and performance, the efficient and precise monitoring and analysis of component processing has become a pressing challenge in the manufacturing industry. Traditional quality control methods rely primarily on manual inspection and simple automated equipment, such as optical measuring instruments and coordinate measuring machines. These methods are often limited by operator skill, the equipment's detection range, and variability in the working environment, making it difficult to achieve fully automated, all-weather, and high-precision process control.

[0004] Therefore, how to use a more intelligent technology to comprehensively and accurately identify various changes in automotive parts during the precision machining process, especially real-time monitoring under high-precision requirements and dynamic environments, has become a core issue that the manufacturing industry urgently needs to solve. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention proposes a precision identification method and device for automobile parts processing to solve at least one of the above technical problems.

[0006] To achieve the above object, the present invention provides a method for precision identification of automobile parts processing, comprising the following steps: Step S1: Obtaining a component processing monitoring video, and performing dynamic video imaging parameter adjustment and full video delay synchronization processing to obtain a delay synchronization optimized video; Step S2: Visually recognize the microstructure of the component based on the delayed synchronous optimization video, perform three-dimensional morphological geometry calculation, and extract the three-dimensional morphological parameters of each feature point; Step S3: tracking optical flow changes frame by frame on the delayed synchronization optimization video, and calculating morphological parameter changes on the three-dimensional morphological parameters to obtain the overall processing change state of the component; Step S4: mining the processing path change logic based on the delayed synchronization optimization video, and then performing three-dimensional time series change modeling on the overall processing change state of the component to build a real-time processing change model of the component; Step S5: Perform potential processing deviation detection and processing error calculation on the global delay optimization video to obtain component processing error parameters; Step S6: Perform final processing simulation on the real-time processing change model of the component based on the component processing error parameters, and perform real-time processing parameter correction to perform precision processing identification optimization operations.

[0007] This invention uses a high-resolution industrial camera to capture real-time video data of component processing, ensuring high-quality images with sufficient detail. This step ensures the accuracy of the underlying data for subsequent analysis. Dynamic adjustment of video imaging parameters such as exposure, white balance, and gain addresses the challenges of varying lighting and reflections during processing. This real-time adjustment ensures that every frame of the component in the video is clearly presented, avoiding misjudgments caused by poor image quality. In multi-device or multi-sensor systems, data transmission can result in delayed asynchrony. Delayed synchronization technology eliminates this delay in the video data, ensuring temporal consistency and minimizing data deviations caused by delays. This step effectively improves the synchronization between the video and the processing, enabling subsequent analysis to more accurately capture dynamic changes in the component. Leveraging image processing and deep learning techniques (such as convolutional neural networks), it is possible to identify micron-level structures and detailed features on components, even tiny defects during processing. By calculating the three-dimensional coordinates and geometric parameters (such as curvature, angle, and depth) of each feature point on the component, a comprehensive understanding of the component's spatial shape is achieved. Compared to traditional two-dimensional image analysis, three-dimensional computing captures more spatial information and provides a more three-dimensional description of processing features. By extracting 3D morphological parameters, the spatial position of each feature point can be accurately determined, providing precise data for subsequent machining error detection and time-series change modeling. Frame-by-frame optical flow tracking technology precisely captures component displacement and changes during machining, making it suitable for dynamic monitoring of component deformation and motion during machining. It can track the displacement of tiny surface areas of a component, identify subtle changes, and help track errors or changes during machining in real time. By calculating changes in 3D morphological parameters, deformation or errors in components during machining can be accurately identified. Compared to traditional machining monitoring methods, optical flow-based change tracking captures real-time component changes more comprehensively and offers higher accuracy. This step provides a comprehensive picture of the machining state, enabling real-time assessment of component machining changes from microscopic to macroscopic perspectives, identifying instabilities or potential problems during machining. By comprehensively analyzing optical flow data and machining trajectories in the video, path changes at each stage of the machining process can be extracted, revealing the underlying logic of machining behavior. This step enables modeling of the entire machining process and further exploring potential machining patterns. Combining 3D change information at each stage with the time series creates a dynamic, real-time model of component machining changes. This model can be continuously updated to reflect component changes in real time, significantly improving the visualization and controllability of the machining process. By establishing a 3D temporal change model, machining trajectories, time progression, and process behavior can be precisely and synchronously managed, providing a theoretical basis for subsequent process optimization. Through in-depth analysis of delayed synchronization optimization videos, various deviations (such as dimensional and shape errors) arising during component machining can be detected in real time.Compared to traditional deviation detection methods, video-based detection is more real-time and dynamic, enabling timely identification and correction of problems. Precise image analysis quantifies machining errors and calculates their specific values. This provides precise input for subsequent compensation calculations, preventing quality failures caused by machining errors. Combined with real-time video streaming, the component machining process can be continuously monitored, ensuring that errors throughout the manufacturing process are promptly detected and corrected, guaranteeing product precision and quality. Machining simulations based on error parameters and timely correction of machining process parameters effectively reduce production deviations and ensure that components meet design standards. This process provides a closed-loop system for process optimization, enabling real-time corrections in actual production. The optimized parameters derived from simulations can be used to instantly adjust the operating status of machining equipment, automatically executing precision machining operations and achieving automated error compensation. This step significantly improves production accuracy and efficiency. Real-time dynamic adjustment of machining parameters ensures maximum component machining precision, significantly reducing scrap rates and improving product quality, especially in high-precision manufacturing processes.

[0008] In this specification, a precision identification device for automobile parts processing is provided, which is used to perform the precision identification method for automobile parts processing as described above, comprising: The video optimization module is used to obtain component processing monitoring videos, adjust dynamic video imaging parameters, and perform full video delay synchronization processing to obtain delayed synchronization optimized videos; The 3D morphology calculation module is used to visually identify the microstructure of components based on the delayed synchronous optimization video, perform 3D morphology geometry calculations, and extract the 3D morphology parameters of each feature point; An optical flow change tracking module is used to track the optical flow changes of the delayed synchronization optimized video frame by frame, and calculate the morphological parameter changes of the three-dimensional morphological parameters to obtain the overall processing change state of the component; The 3D modeling module is used to mine the logic of machining path changes based on the delayed synchronization optimization video, and then perform 3D time series change modeling on the overall machining change state of the component to build a real-time machining change model for the component; Deviation detection module, used to detect potential processing deviations and calculate processing errors for the global delay optimization video, and obtain component processing error parameters; The parameter correction module is used to perform final processing simulation on the real-time processing change model of the parts based on the processing error parameters of the parts, and to perform instant processing process parameter correction to perform precision processing identification and optimization operations.

[0009] This invention uses a high-resolution camera to capture processing video, ensuring clear and detailed image information. The optimized video ensures precise capture of the component surface and the processing process. This module dynamically adjusts video imaging parameters (such as exposure, gain, and white balance) to account for various lighting and reflection variations, ensuring that the video maintains excellent visual quality under various working conditions. The optimized image quality is enhanced, reducing image blur or errors caused by lighting variations. Synchronous processing eliminates time delays in the video, enabling precise alignment of data from multiple cameras and sensors. This process effectively reduces timing inconsistencies caused by video delays, making subsequent processing analysis more accurate, which is particularly crucial in high-speed production lines. By combining image depth information with geometric calculation methods, 3D morphological parameters (such as depth, curvature, and angle) are extracted for each feature point. This 3D data provides accurate spatial information for subsequent precision analysis and error detection, clearly depicting the spatial variations of the component. Accurate calculation of the component's 3D morphology provides an accurate reference model for subsequent detection and compensation of processing errors. This process is particularly important for precision parts, ensuring they meet design requirements. Optical flow algorithms can track minute displacements of component surfaces or structures frame by frame, accurately capturing subtle deformations during machining. This is crucial for dynamically monitoring real-time changes during machining, enabling the identification of small deformations that are difficult to detect using static images. Combined with 3D morphological data, the module calculates morphological changes at each feature point of the component, such as positional offset and dimensional deformation, thereby reflecting the overall machining status of the component. In-depth analysis of machining path changes reveals the individual steps and path logic involved in the component machining process. This provides fundamental data for process optimization, helping engineers identify potential anomalies and irregularities during production. By combining 3D data from each machining stage with time series data, a dynamic 3D machining change model is constructed, reflecting the real-time changes in component morphology during machining. This makes the machining process more predictable and controllable. 3D temporal change modeling enables real-time monitoring of component machining processes, identifying potential deviations and enabling adjustments to ensure accuracy and consistency at each machining step. By analyzing globally optimized videos, potential deviations in the machining process, such as dimensional and shape deviations, can be detected in real time. Early detection of problems can prevent subsequent quality issues and reduce scrap rates. Image analysis technology accurately calculates machining errors and provides quantitative data for subsequent corrections. Compared to traditional manual inspection, this calculation method is more accurate and reduces human error. This module continuously monitors the entire machining process, ensuring that all machining deviations remain within the predetermined tolerance range, thereby improving the overall quality stability of the component. Real-time feedback on error parameters enables automatic adjustment of machining parameters, enabling real-time process optimization.Ensure that the component processing process can be dynamically adjusted based on feedback errors to ensure accuracy. Real-time compensation calculations for processing errors can significantly reduce quality issues caused by error accumulation. Adjustment of processing parameters ensures that the size and shape of the component meet preset standards. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 This is a schematic flow chart of the steps of a precision identification method for automobile parts processing according to the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Schematic diagram of the detailed implementation steps of step S3. DETAILED DESCRIPTION

[0011] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0012] This application provides a precision identification method and apparatus for automotive parts processing. The execution entities of this precision identification method and apparatus include, but are not limited to, the following: mechanical equipment, data processing platforms, cloud server nodes, network upload devices, etc., which can be considered general computing nodes of this application. The data processing platform includes, but is not limited to, at least one of an audio and image management system, an information management system, and a cloud data management system.

[0013] See also Figures 1 to 4 The present invention provides a precision identification method for automobile parts processing, which comprises the following steps: Step S1: Obtaining a component processing monitoring video, and performing dynamic video imaging parameter adjustment and full video delay synchronization processing to obtain a delay synchronization optimized video; Step S2: Visually recognize the microstructure of the component based on the delayed synchronous optimization video, perform three-dimensional morphological geometry calculation, and extract the three-dimensional morphological parameters of each feature point; Step S3: tracking optical flow changes frame by frame on the delayed synchronization optimization video, and calculating morphological parameter changes on the three-dimensional morphological parameters to obtain the overall processing change state of the component; Step S4: mining the processing path change logic based on the delayed synchronization optimization video, and then performing three-dimensional time series change modeling on the overall processing change state of the component to build a real-time processing change model of the component; Step S5: Perform potential processing deviation detection and processing error calculation on the global delay optimization video to obtain component processing error parameters; Step S6: Perform final processing simulation on the real-time processing change model of the component based on the component processing error parameters, and perform real-time processing parameter correction to perform precision processing identification optimization operations.

[0014] This invention uses a high-resolution industrial camera to capture real-time video data of component processing, ensuring high-quality images with sufficient detail. This step ensures the accuracy of the underlying data for subsequent analysis. Dynamic adjustment of video imaging parameters such as exposure, white balance, and gain addresses the challenges of varying lighting and reflections during processing. This real-time adjustment ensures that every frame of the component in the video is clearly presented, avoiding misjudgments caused by poor image quality. In multi-device or multi-sensor systems, data transmission can result in delayed asynchrony. Delayed synchronization technology eliminates this delay in the video data, ensuring temporal consistency and minimizing data deviations caused by delays. This step effectively improves the synchronization between the video and the processing, enabling subsequent analysis to more accurately capture dynamic changes in the component. Leveraging image processing and deep learning techniques (such as convolutional neural networks), it is possible to identify micron-level structures and detailed features on components, even tiny defects during processing. By calculating the three-dimensional coordinates and geometric parameters (such as curvature, angle, and depth) of each feature point on the component, a comprehensive understanding of the component's spatial shape is achieved. Compared to traditional two-dimensional image analysis, three-dimensional computing captures more spatial information and provides a more three-dimensional description of processing features. By extracting 3D morphological parameters, the spatial position of each feature point can be accurately determined, providing precise data for subsequent machining error detection and time-series change modeling. Frame-by-frame optical flow tracking technology precisely captures component displacement and changes during machining, making it suitable for dynamic monitoring of component deformation and motion during machining. It can track the displacement of tiny surface areas of a component, identify subtle changes, and help track errors or changes during machining in real time. By calculating changes in 3D morphological parameters, deformation or errors in components during machining can be accurately identified. Compared to traditional machining monitoring methods, optical flow-based change tracking captures real-time component changes more comprehensively and offers higher accuracy. This step provides a comprehensive picture of the machining state, enabling real-time assessment of component machining changes from microscopic to macroscopic perspectives, identifying instabilities or potential problems during machining. By comprehensively analyzing optical flow data and machining trajectories in the video, path changes at each stage of the machining process can be extracted, revealing the underlying logic of machining behavior. This step enables modeling of the entire machining process and further exploring potential machining patterns. Combining 3D change information at each stage with the time series creates a dynamic, real-time model of component machining changes. This model can be continuously updated to reflect component changes in real time, significantly improving the visualization and controllability of the machining process. By establishing a 3D temporal change model, machining trajectories, time progression, and process behavior can be precisely and synchronously managed, providing a theoretical basis for subsequent process optimization. Through in-depth analysis of delayed synchronization optimization videos, various deviations (such as dimensional and shape errors) arising during component machining can be detected in real time.Compared to traditional deviation detection methods, video-based detection is more real-time and dynamic, enabling timely identification and correction of problems. Precise image analysis quantifies machining errors and calculates their specific values. This provides precise input for subsequent compensation calculations, preventing quality failures caused by machining errors. Combined with real-time video streaming, the component machining process can be continuously monitored, ensuring that errors throughout the manufacturing process are promptly detected and corrected, guaranteeing product precision and quality. Machining simulations based on error parameters and timely correction of machining process parameters effectively reduce production deviations and ensure that components meet design standards. This process provides a closed-loop system for process optimization, enabling real-time corrections in actual production. The optimized parameters derived from simulations can be used to instantly adjust the operating status of machining equipment, automatically executing precision machining operations and achieving automated error compensation. This step significantly improves production accuracy and efficiency. Real-time dynamic adjustment of machining parameters ensures maximum component machining precision, significantly reducing scrap rates and improving product quality, especially in high-precision manufacturing processes.

[0015] In the embodiment of the present invention, see Figure 1 , is a schematic flow chart of the steps of a precise identification method for automobile parts processing according to the present invention. In this example, the precise identification method for automobile parts processing includes the following steps: Step S1: Obtaining a component processing monitoring video, and performing dynamic video imaging parameter adjustment and full video delay synchronization processing to obtain a delay synchronization optimized video; In this example, select appropriate monitoring equipment (such as an HD camera or industrial camera) that supports high frame rates and high resolution to capture details during the component processing. Generally, a resolution of 1920x1080 or higher (such as 4K) is selected, with a frame rate of at least 30 frames per second to ensure smooth video. Select an industrial camera that supports 1080p resolution and 60 frames per second, paired with an appropriate lens, to obtain clear processing images. Ensure adequate lighting in the processing environment to reduce shadows and reflections. This can be achieved by using uniform LED lighting or ring lights to ensure even illumination across the entire processing area and avoid image quality degradation caused by uneven lighting. Set up four LED lights, positioned in four directions around the processing area, to achieve uniform lighting and eliminate shadows. Start video recording to record the entire component processing process. Ensure the camera is stable throughout the entire process to avoid shake. Use a tripod or fixed mount. Set the recording time to the entire processing cycle to ensure that the entire process is captured from start to finish. The video length may exceed 10 minutes, depending on the processing time. Perform a preliminary analysis of the recorded video to observe the image clarity, contrast, and exposure. Determine whether there is overexposure or underexposure, and whether the target component is clearly visible. This step is crucial for ensuring the effectiveness of subsequent analysis. Use video analysis software (such as Adobe Premiere or Final Cut Pro) to observe the video's brightness histogram and confirm the image's distribution across the dynamic range. Based on the initial analysis results, adjust the camera's dynamic imaging parameters. Key adjustments include exposure time, gain (ISO), white balance, and contrast to optimize image quality. If the video is overexposed, adjust the exposure time to 1 / 120 second and the gain to ISO 200 to ensure moderate brightness and clear component details. During the actual recording process, monitor the video in real time and make dynamic adjustments accordingly. This can be done by viewing the live feed on a monitor or screen to ensure optimal imaging throughout the recording. If excessive brightness is detected in a frame during recording, immediately adjust the camera settings to ensure consistent image quality in subsequent frames. In the acquired raw video, identify time synchronization issues caused by device response delays or motion blur. Frame-to-frame time delays may occur, especially during fast processing. During the analysis process, we discovered significant motion blur in certain frames, making it difficult to discern processed details. We then processed the video using delayed synchronization algorithms (such as timestamp synchronization and frame interpolation). By analyzing the time differences between frames, we adjusted the playback order to achieve synchronization. This processing can be performed using specialized software (such as After Effects). Frame interpolation smoothed the transitions between frames, reducing motion blur and improving image clarity.After completing the delay synchronization process, output the optimized video to ensure smooth video playback and high image quality. High-quality codecs (such as H.264) can be selected as the output format to preserve video details. The generated delay synchronization optimized video should have a resolution of 1920x1080 and a frame rate of 60 fps to ensure clear images for subsequent analysis and facilitate defect detection and feature extraction.

[0016] Step S2: Visually recognize the microstructure of the component based on the delayed synchronous optimization video, perform three-dimensional morphological geometry calculation, and extract the three-dimensional morphological parameters of each feature point; In this embodiment, key frames are extracted from a delayed synchronization optimized video to facilitate visual recognition of small structures. The selected key frames should represent different states of the component during the machining process. Key frames are typically extracted every few frames to ensure coverage of the entire machining process. One frame is extracted from every 10 frames. For a 60-second video with a 30 fps frame rate, approximately 180 frames are extracted to ensure critical moments are captured. The extracted frames should be high-quality images (e.g., PNG format) to preserve detail. The extracted key frames are labeled with small structural feature points. Feature points may include holes, edges, and concave and convex surfaces, which serve as the basis for subsequent analysis. Feature point detection is performed using computer vision tools (such as OpenCV or MATLAB). Harris corner detection or the SIFT (Scale-Invariant Feature Transform) algorithm can be used to automatically identify and label key feature points. If 10 feature points are detected in a frame, their image locations (x, y coordinates) and feature descriptors are recorded for subsequent analysis. The labeled feature points are screened to remove noise and unreliable features, ensuring consistency and reliability across frames. By comparing the changes in the same feature points in adjacent frames, the RANSAC (Random Sample Consensus) algorithm is used to verify the stability of the selected feature points throughout the video. If the position of a feature point varies by less than 1 pixel between consecutive frames, it is considered stable and meets the requirements of subsequent analysis. Three-dimensional spatial reconstruction is performed using the coordinates of the feature points from multiple viewpoints. Using stereo vision technology, the 2D feature point coordinates are converted to 3D spatial coordinates using known camera parameters (such as focal length, sensor size, and camera position). If the positions of the feature points in two frames are P1 (x1, y1) and P2 (x2, y2), their 3D coordinates Z are calculated based on the camera's intrinsic and extrinsic parameters. Triangulation is then used to obtain the 3D coordinates (X, Y, Z) of each feature point. Geometric parameters are extracted from the reconstructed 3D feature points, including diameter, depth, height, and width. These parameters reflect the geometric characteristics of the feature points and facilitate subsequent analysis. For a hole feature point, its diameter D and depth H are calculated. If D is calculated to be 5 mm and H is 2 mm, these parameters are recorded as the 3D morphological parameters of the feature point. Organize the extracted 3D morphological parameters into a structured data table. The 3D coordinates and corresponding geometric parameters of each feature point should be clearly recorded to facilitate subsequent analysis and comparison. Construct a data table that lists each feature point's number, 3D coordinates (X, Y, Z), diameter, depth, and other information to ensure data integrity and easy access. Use 3D visualization software (such as Blender or MATLAB) to generate a 3D model from the extracted 3D feature points and morphological parameters for intuitive analysis and verification. Draw a 3D model based on the extracted data, observing whether features such as holes and protrusions meet design requirements and verifying the model's completeness and accuracy.

[0017] Step S3: tracking optical flow changes frame by frame on the delayed synchronization optimization video, and calculating morphological parameter changes on the three-dimensional morphological parameters to obtain the overall processing change state of the component; In this embodiment, a suitable optical flow calculation algorithm (such as the Lucas-Kanade method or the Horn-Schunck method) is selected to track optical flow changes in the delayed synchronization optimized video. Optical flow algorithms can effectively capture pixel movement between consecutive frames, helping to identify dynamic changes in parts during the manufacturing process. The Lucas-Kanade method is selected because it is suitable for small displacement scenarios, has low computational complexity, and can process high-resolution video in real time. Specifically, optical flow is calculated between every two frames to ensure that subtle changes are captured. Images are extracted frame by frame from the delayed synchronization optimized video, and the optical flow between each frame and the previous frame is calculated. By analyzing the pixel motion vectors, the motion trajectory of the part during the manufacturing process is obtained. For a 60-second video with a frame rate of 30 fps, there are a total of 1800 frames. By processing each frame, the optical flow is calculated frame by frame, and the motion vector of each feature point is recorded. If the position change of a feature point in consecutive frames is (Δx, Δy), the optical flow vector is V = (Δx, Δy). The calculated optical flow vectors are recorded as the dynamic optical flow change trajectory, forming a series of continuous motion vectors. These trajectories describe the motion pattern of the component during machining, facilitating subsequent analysis. If the optical flow trajectory of a feature point within the first five frames is [(x1, y1), (x2, y2), (x3, y3), (x4, y4), (x5, y5)], the changes in these points are recorded to form an optical flow trajectory dataset. Based on the previously extracted 3D morphological parameters, the component's geometric features, such as diameter, depth, and height, are defined. Based on this, the changes in these parameters during machining are calculated. For example, if the initial diameter of a feature point is D0 = 5 mm, and frame-by-frame optical flow tracking reveals that the feature point gradually changes to D1 = 4.5 mm during machining, the change in diameter is calculated as ΔD = D1 - D0 = -0.5 mm. The changes in morphological parameters from different frames are integrated to form a complete change dataset. By comparing the morphological parameters of each frame, the overall change state of the component throughout the machining process can be identified. The diameter change of each feature point in each frame is recorded, forming a change record such as "Feature Point 1: D0 = 5 mm, D1 = 4.8 mm, D2 = 4.6 mm" for subsequent analysis. Based on the calculated morphological parameter changes, the overall processing change status of the component is summarized. This includes the changes in all feature points and their impact on component performance. If the diameter change of most feature points is within an acceptable range (e.g., ±0.2 mm), the component can be judged to be in good condition. If the change of some feature points exceeds the threshold, further analysis and processing are required. The morphological parameter change results and dynamic optical flow trajectories are visualized to help engineers intuitively understand the changes in the component during the processing. Dynamic charts can be generated using data visualization tools (such as MATLAB or Python libraries).A 3D scatter plot showing the locations and variations of feature points is generated, allowing users to visualize the overall dynamic behavior of feature points during the machining process. The calculated variations are compared with standard machining requirements to verify component quality. This critical step helps identify potential issues during machining. The variation results for each feature point are compared with the design requirements. If a feature point's variation exceeds the standard range (e.g., ±0.5 mm), it is recorded as a potential issue for subsequent resolution.

[0018] Step S4: mining the processing path change logic based on the delayed synchronization optimization video, and then performing three-dimensional time series change modeling on the overall processing change state of the component to build a real-time processing change model of the component; In this embodiment, the optical flow change trajectory and three-dimensional morphological parameter data extracted from the delayed synchronization optimization video are collated and cleaned. The data is ensured to contain no missing or outliers to ensure the accuracy of subsequent analysis. The motion vector of each feature point in the optical flow trajectory is checked to confirm that valid data is recorded in each frame. For example, position changes should not exceed a set reasonable range (e.g., ±1 pixel). Based on the extracted optical flow data, path changes during component processing are analyzed. Feature extraction can identify significant path changes by comparing the motion trajectories at different processing stages. A threshold is set. If the optical flow velocity of a feature point suddenly increases at a certain stage (e.g., exceeding a set threshold of 0.3 mm / s), it is recorded as a key event for path change. These events are used for subsequent logic mining. A change logic model is established to describe the causal relationship of path changes during processing. A causal inference model can be used to identify the relationship between different processing parameters and path changes. The relationship between processing parameters (e.g., tool feed rate and cutting depth) and path changes is analyzed, and a model is established using statistical methods (e.g., regression analysis). If an increase in tool feed rate is found to be positively correlated with path deviation, this logical relationship is recorded. The path change information obtained from optical flow change logic mining is integrated with the previously extracted 3D morphological parameter data. This creates a 3D dataset containing a time series for dynamic modeling. A data structure is created to record the location, morphological parameters (such as diameter and height) of the feature points at each time point, along with the corresponding machining path change information. This data is recorded as follows: "Time point 1: Feature point A location, diameter, height, path change." A real-time machining change model of the component is constructed using a 3D modeling tool (such as MATLAB or Blender). This model should dynamically reflect the changes in each feature point during machining. The integrated time series data is input into the modeling tool to generate a dynamic 3D model that displays the real-time changes in the component's feature points during machining. The model should include animation effects to demonstrate how the feature points change over time. The constructed real-time machining change model is validated to ensure its consistency with the actual machining process. The model's accuracy can be verified by comparing it with actual measurement data. During machining, the feature points are measured, their locations and morphological parameters recorded, and then compared with the model's predictions. If the predicted changes are within ±0.1 mm, the model is considered accurate.

[0019] Step S5: Perform potential processing deviation detection and processing error calculation on the global delay optimization video to obtain component processing error parameters; In this embodiment, keyframes are extracted from the global delay optimization video to detect potential processing deviations. Keyframes should cover different stages of the entire processing process to ensure that all features of the component are captured. A keyframe is extracted every 5 frames, resulting in approximately 120 frames for the entire processing process (assuming a 10-second video length and a frame rate of 30 fps). Each frame should be high-quality (e.g., in PNG format) to preserve details. Image processing algorithms (e.g., edge detection and template matching) are used to identify and mark feature points in each keyframe. These feature points may include holes, edges, protrusions, etc. The Canny edge detection algorithm is used to extract feature points and compare them to a preset standard model. The distance and deviation between the actual feature points and the standard feature points are calculated. Based on the deviation between the actual feature points and the standard model, a deviation detection algorithm (e.g., threshold determination, statistical analysis, etc.) is applied to determine whether there is potential processing deviation. A reasonable deviation threshold (e.g., ±0.2 mm) is set to identify deviation areas. If the actual dimension of a feature point is 9.5 mm, while the standard dimension is 10 mm, the calculated deviation is -0.5 mm, exceeding the threshold, and the area is marked as a potential machining deviation area. Define machining error parameters, including positional error, dimensional error, and morphological error. Record the specific error values ​​for each potential deviation area for subsequent analysis. By comparing the actual hole diameter with the standard hole diameter, record the actual dimension of each hole and its deviation from the design, forming a record such as "Hole A: Actual dimension 9.5 mm, designed dimension 10 mm, error -0.5 mm." Summarize the error parameters of all potential machining deviation areas to calculate the overall machining error. This can be achieved through simple weighted averaging or statistical analysis to assess the overall machining quality of the component. If the errors of five feature points on a part are -0.5 mm, 0.3 mm, -0.2 mm, 0.1 mm, and -0.4 mm, respectively, the overall machining error can be calculated as their average: (-0.5 + 0.3 - 0.2 + 0.1 - 0.4) / 5 = -0.14 mm. Calculated machining error parameters are visualized, presenting the error distribution using charts or heat maps. This provides a more intuitive understanding of component machining quality. Heat maps depict the error value for each feature point, with color representing the magnitude of the error, making it easier to identify problem areas. Areas with significant error are marked as requiring special attention.

[0020] Step S6: Perform final processing simulation on the real-time processing change model of the component based on the component processing error parameters, and perform real-time processing parameter correction to perform precision processing identification optimization operations.

[0021] In this embodiment, the component machining error parameters obtained from the previous step are collated and analyzed. These error parameters, including positional error, dimensional error, and morphological error, provide a basis for subsequent machining simulations. If the recorded error of a feature point is -0.5 mm (aperture diameter) or +0.3 mm (edge), these parameters will be used as inputs to influence the final machining simulation. Based on the collected machining error parameters, the component's real-time machining variation model is updated. Finite element analysis (FEA) can be used to consider the impact of these errors on the component's morphology. By inputting the feature point error values ​​into the model, numerical simulation techniques are used to adjust the model's geometry to ensure it reflects the actual state during machining, such as updating the aperture from the designed 10 mm to 9.5 mm. Based on the updated real-time machining variation model, a final machining simulation is performed. This simulation should dynamically reflect potential changes during machining to ensure model accuracy. A simulation model is generated using 3D modeling software (such as MATLAB or SolidWorks) to demonstrate the actual changes in the component during machining, including the impact of the errors on its morphology. The results of the final machining simulation are analyzed to identify potential problem areas during machining. By comparing simulation results with design requirements, it's determined whether parameter corrections are necessary. If the simulation shows that the error at a feature point exceeds a set threshold (e.g., ±0.2 mm), immediate correction of machining parameters is required. Based on the analysis results, immediate corrections are made to the component's machining parameters. These parameters may include tool feed rate, depth of cut, and tool path, aiming to reduce errors during machining. If a feature point's diameter is insufficient, the tool feed rate can be adjusted from 100 mm / min to 80 mm / min, and the depth of cut increased to ensure the final machining result meets design standards. During the actual machining process, real-time monitoring and feedback mechanisms are implemented to ensure that modified process parameters are applied in real time. Adjustments are made through the control interface of the CNC system or machining equipment, ensuring machining flexibility. Tool status and machining parameters are monitored in real time through the CNC machine's control panel, ensuring that new settings take effect immediately and can be dynamically adjusted during machining. Based on the corrected machining parameters, precision machining identification and optimization operations are performed. During machining, the component's machining status is continuously monitored to ensure that the shape and size of each feature point meet standards. During the actual machining process, the diameters of feature points are measured in real time to ensure they remain within the designed range. If deviations are detected, machining parameters are immediately adjusted based on the feedback. While the optimization process is in progress, every step of the machining process is recorded and analyzed. This includes real-time measurement data and machining parameters for each feature point to evaluate the optimization results. The actual dimensions of each feature point after machining are recorded and compared with the initial design to analyze whether the optimized machining errors have been improved. A final machining report is generated based on the optimization results.The report should detail the processing status, error parameters, and optimization results for each feature point to facilitate subsequent quality control and improvement. The generated report might include information such as "Feature point A: Actual diameter 9.8 mm, designed diameter 10 mm, error -0.2 mm. After optimization, the error improved to ±0.1 mm," providing data support for subsequent production.

[0022] In this embodiment, refer to Figure 2 , is a flowchart of the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Acquire component processing monitoring videos using high-resolution industrial cameras; Adjust dynamic video imaging parameters of parts processing monitoring videos to construct imaging optimization monitoring videos; Extracting the timestamp of each frame of the imaging optimization monitoring video; Performing a global frame delay average calculation according to the timestamp to obtain an inter-frame average delay value; Performing frame-by-frame delay correction calculation on the timestamp of each frame according to the average delay value between frames to generate a multi-frame correction time difference; Full video delay synchronization processing is performed based on the multi-frame correction time difference, thereby obtaining a delay synchronization optimized video.

[0023] In this example, a high-resolution industrial camera is selected for component processing monitoring. Ensure that the camera has a sufficient frame rate (e.g., 30 fps or higher) and resolution (e.g., 1920x1080 or higher) to capture clear processing details. Configure the camera's aperture, shutter speed, and ISO value to suit the lighting conditions in the processing environment to ensure optimal video quality. During filming, ensure uniform lighting in the processing environment, without strong shadows or reflections. Use a soft light or ring light to improve lighting conditions and avoid image distortion caused by uneven lighting. Before filming, perform environmental testing to ensure that the camera can operate stably under the specified lighting conditions. Adjust the light intensity to 450 Lux to ensure clear images. Dynamically adjust the video imaging parameters, including exposure time, focus, and gain, to accommodate dynamic changes in the component processing process. Monitor video quality in real time to ensure that the image is free of blur or overexposure at all times. During actual processing, if the image is too dark, adjust the exposure time in real time from 1 / 60 second to 1 / 30 second to increase image brightness. Record the adjusted parameters and apply them to the video recording to generate an optimized imaging monitoring video. Ensure that the video maintains good image quality throughout the processing process, providing a clear data foundation for subsequent analysis. The optimized video stream maintains a stable frame rate and clear image details, allowing each processing step to be clearly identified. Use video processing tools to extract the timestamp of each frame in the optimized imaging monitoring video. Timestamps should be in milliseconds to ensure the precise time of each frame. If the total video length is 10 seconds and the frame rate is 30 fps, 300 timestamps should be generated to record the specific time of each frame. Store the extracted timestamps in the dataset as a timestamp list for subsequent calculations and analysis. Ensure that the order of the timestamps matches the order of the video frames. The timestamp format in the recorded timestamp list is: [0 ms, 33.33 ms, 66.67 ms, ..., 10000 ms] to ensure the accuracy of each time point. Calculate the time delay between adjacent frames and obtain the delay value for each pair of adjacent frames. The delay value is calculated as: Delay = Current frame timestamp - Previous frame timestamp. If the timestamp of the second frame is 33.33 ms and the timestamp of the first frame is 0 ms, the delay is calculated to be 33.33 ms. The delay values ​​of all adjacent frames are averaged to obtain the global inter-frame average delay value. This value is used for subsequent frame-by-frame delay correction. If the calculated delays of the first 10 frames are 33.33 ms, 33.67 ms, and so on, the global inter-frame average delay value is the arithmetic average of these delays, assuming it is 33.50 ms. Based on the global inter-frame average delay value, the timestamp of each frame is corrected to calculate the corrected timestamp. The correction method is: Corrected timestamp = Original timestamp - (Frame number - 1) × Average delay value.If the original timestamp of the third frame is 66.67 ms, the corrected timestamp is 66.67 ms - 2 × 33.50 ms = -0.33 ms (needs to be adjusted to 0 ms depending on the actual situation). The corrected timestamps for each frame are calculated to obtain the multi-frame corrected time difference, forming a new timestamp dataset for subsequent delay synchronization. The new timestamp list will be adjusted to: [0 ms, 33.50 ms, 67.00 ms, ..., 10000 ms], ensuring temporal consistency between frames. Based on the multi-frame corrected time differences, the entire video is delay-synchronized to ensure that the display time of each frame is consistent with the corrected timestamps. Use video editing software to apply the corrected timestamps to the original video to ensure that the display of each frame is synchronized with the timestamps. After synchronization is complete, export the delay-synchronized video to ensure that the timing of each frame matches the corrected timestamps during playback, providing a smoother viewing experience. The exported video should maintain the original resolution and frame rate to ensure that every detail can be clearly identified during processing monitoring.

[0024] In this embodiment, the specific steps of adjusting the dynamic video imaging parameters of the component processing monitoring video and constructing the imaging optimization monitoring video are as follows: Extracting an initial frame image of the video based on the component processing monitoring video; Calculate the surface reflectivity of the components on the initial frame of the video and extract the surface reflectivity; Analyze the material properties of parts based on the initial frame of the video Identify light intensity changes based on the initial frame image of the video to obtain light change features Dynamically adjusting video imaging parameters based on surface light reflectivity, material properties of the components, and light change characteristics to obtain dynamically optimized imaging parameters; Based on the dynamic optimization of imaging parameters, the global video parameters of the component processing monitoring video are optimized to construct the imaging optimized monitoring video.

[0025] In this example, an appropriate time point is selected from the component processing monitoring video, and the initial frame image at that moment is extracted. Typically, a stable moment during the processing process is selected to ensure a clear and representative image. Images are extracted midway through the process, when the processing speed and state are relatively stable, facilitating subsequent analysis. Frame images at the selected time point are extracted using video processing software or tools (such as OpenCV or FFmpeg). Ensure that the extracted image format is high-quality (such as PNG or TIFF) to preserve more detail. The frame image is extracted at a resolution of 1920x1080 to ensure image clarity and detail integrity. Light reflectance refers to the ability of a material surface to reflect incident light and is typically calculated by calculating the ratio of reflected light intensity to incident light intensity. High reflectance generally indicates a smooth or glossy surface. For the extracted initial frame image, the intensity of the incident light source (e.g., 300 Lux) is measured using a photometer or illuminometer. The reflected light intensity is recorded, and a region of interest (ROI) is selected within the image for analysis using an appropriate tool (such as a camera or photometer). Assuming the incident light intensity is 300 Lux and the reflected light intensity is 150 Lux, the reflectance is calculated as: Reflectance = Reflected Light Intensity / Incident Light Intensity = 150 Lux / 300 Lux = 0.5 (i.e., 50%). The calculated illumination reflectance is recorded as a key feature and used for subsequent dynamic imaging parameter adjustment and optimization. Recording a reflectance value of 50% will influence the parameter settings for subsequent image optimization. Analyze the component's material properties based on the initial frame image. Determine the material type (e.g., aluminum, steel, plastic, etc.) based on image features such as texture, color, and gloss. A smooth surface with a metallic luster suggests aluminum alloy. Compare the extracted reflectance with known material characteristics to further confirm the material's identity. Aluminum alloy's reflectance typically ranges from 30% to 80%. A measured value of 50% is consistent with this characteristic. Record these characteristics as baseline data for subsequent imaging parameter adjustments. In the initial frame image, the characteristics of light intensity changes are identified by analyzing the brightness histogram or grayscale value changes. Dynamic monitoring of light changes is crucial for subsequent optimization. Analyze the brightness distribution in the image. If the brightness changes in certain areas exceed the set threshold (such as 15%), it can be regarded as a change in light intensity. Record the identified light change characteristics, including the amplitude and frequency of the changes and possible influencing factors (such as movement of the processing tool or changes in ambient light). Record the light intensity fluctuation of ±20 Lux over a period of time and mark it as a light change feature. Based on the extracted surface light reflectivity, material properties and light change characteristics, preliminarily set the dynamic imaging parameters. These parameters may include exposure time, gain, white balance and contrast. If the reflectivity is 50%, the exposure time can be set to 1 / 100 second to avoid overexposure.During the actual monitoring process, the imaging parameters are continuously adjusted based on real-time feedback and analysis results. The gain value is adjusted according to the characteristics of the lighting changes to ensure that the image remains clear when the light intensity changes. If the light is detected to be weakened, the gain can be increased from 1.5 to 2.0 to increase the image brightness. The adjusted dynamic optimization imaging parameters are applied to the entire component processing monitoring video. Ensure that all frame images maintain consistent imaging quality. Use video processing software to batch apply settings to ensure that all frames in the video are processed according to the new dynamic parameters. Export the optimized monitoring video to ensure that the video shows the best image quality during playback and can clearly show the details of the component processing process. Set the output video format to high quality (such as H.264 encoding) to ensure that the processing features can be clearly identified in subsequent analysis.

[0026] In this embodiment, refer to Figure 3 , is a flowchart of the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Perform visual recognition of component microstructures based on delayed synchronously optimized videos and mark multiple microstructure feature points; Calculating the spatial positioning coordinates of each of the micro-structural feature points; Perform microstructure feature type analysis on multiple tiny structural feature points to obtain the type of each feature point; The three-dimensional morphological geometry calculation is performed according to the type of each feature point, and the three-dimensional morphological parameters of each feature point are extracted.

[0027] In this embodiment, keyframes are selected from the delayed synchronization optimized video to ensure that microstructural features are clearly visible in the image. Keyframes are typically selected at moments during the machining process when microstructural changes are noticeable. Keyframes are extracted every five frames of the video to ensure that changes at different machining stages are captured for subsequent analysis. Image processing software (such as OpenCV) or computer vision tools are used to analyze each keyframe and manually or automatically mark multiple microstructural feature points. These feature points can include important surface features such as bumps, holes, and edges. For aluminum alloy parts, all visible microholes and protrusions are marked in the keyframes, ensuring that these feature points are consistent across frames. For each marked microstructural feature point, its coordinates in three-dimensional space are calculated. This process typically requires spatial reconstruction using the camera's intrinsic and extrinsic parameters. If the camera has a focal length of 5 mm and a sensor size of 1 / 3 inch, the corresponding three-dimensional coordinates (X, Y, Z) are calculated using a perspective projection model based on the pixel positions (e.g., (x, y) coordinates) of the marked feature points in the image. If the video contains depth information (e.g., acquired through stereo vision or a depth camera), the Z coordinate of each feature point can be directly extracted from the depth map. If depth information is unavailable, it must be acquired through other methods (such as laser ranging or structured light). Assuming the depth of a feature point is 3 mm, the 3D coordinates of that point can be calculated by combining known camera parameters, forming a complete spatial positioning dataset. Feature types are analyzed for multiple feature points, typically based on the geometry and material properties of the microstructure. A preliminary classification can be performed based on the feature point's shape (e.g., circular, square, concave, convex, etc.). Among the marked feature points, if some are clearly circular, they can be classified as "holes," while feature points that exhibit planar convexity are labeled "concave-convex." The type of each feature point and its corresponding spatial coordinates are recorded in a data table to form a systematic dataset for subsequent analysis and calculations. Feature point A is recorded as "hole" with coordinates (X_A, Y_A, Z_A), and feature point B is recorded as "concave-convex" with coordinates (X_B, Y_B, Z_B). Select the appropriate geometric calculation method based on the type of each feature point. The calculation method for geometric parameters may be different for different types of feature points. For circular holes, the diameter is calculated, while for concave-convex structures, the height and width need to be measured. If feature point A is a hole, use image analysis technology (such as edge detection) to calculate its diameter; if feature point B is concave-convex, measure its height change. Calculate the three-dimensional morphological parameters of each feature point, including diameter, depth, height, width, etc. These parameters can reflect the role and influence of the feature point in the surface structure of the component. Assuming that the diameter of feature point A is 1 mm and the height of feature point B is 0.5 mm, record these parameters as basic data for the analysis of the microstructure of the component.

[0028] In this embodiment, refer to Figure 4 , is a flowchart of the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Track the optical flow changes frame by frame in the delayed synchronization optimized video and extract the dynamic pixel optical flow change trajectory; Identify the spatial position change of the spatial positioning coordinates according to the dynamic pixel optical flow change trajectory, and extract the temporal displacement of each feature point; Calculating the morphological parameter changes of the three-dimensional morphological parameters according to the dynamic pixel optical flow change trajectory to obtain the parameter change value of each feature point; The processing change is inferred based on the time sequence displacement and the parameter change value to obtain the overall processing change state of the component.

[0029] In this embodiment, a suitable optical flow algorithm (such as the Lucas-Kanade method or the Horn-Schunck method) is selected to extract dynamic pixel optical flow changes from the delayed synchronization optimized video. These algorithms can effectively calculate the pixel movement between consecutive frames. The Lucas-Kanade method is used. This method is suitable for scenes with small displacements and has low computational complexity, making it suitable for real-time processing. The delayed synchronization optimized video is processed frame by frame, and the optical flow between each frame and the previous frame is calculated. By analyzing the motion vector of the pixel, the change trajectory of the dynamic pixel is identified. If in a certain frame, the feature point moves from position (x1, y1) to (x2, y2), the optical flow vector is calculated as V = (x2 - x1, y2 - y1), and this vector is recorded for subsequent analysis. The calculated optical flow vector is recorded as the dynamic pixel optical flow change trajectory, forming a series of continuous motion vectors. These trajectories will describe the motion pattern of the feature point in the video. If the optical flow trajectory of a feature point moving within five frames is [(x1, y1), (x2, y2), (x3, y3), (x4, y4)], then the changes in these points constitute the optical flow trajectory of the feature point. Based on the dynamic pixel optical flow change trajectory, the spatial coordinates of each feature point are updated. The new spatial position is calculated by applying the optical flow vector to the feature point's initial coordinates. If the feature point's initial coordinates are (X, Y, Z), its coordinates are updated to (X', Y', Z') based on the optical flow change, where X' = X + Δx, Y' = Y + Δy, and Z' = Z + Δz. The temporal displacement of each feature point between frames is calculated. This displacement reflects the spatial change of the feature point during the processing process. If the position of the feature point in the first and second frames is (X, Y) and (X', Y'), respectively, the temporal displacement can be calculated as D = √((X' - X)² + (Y' - Y)²). Based on the dynamic pixel optical flow change trajectory, the changes in the three-dimensional morphological parameters of each feature point during the processing are calculated. This process requires the integration of the feature point's temporal displacement. If a feature point's initial height is H and its height changes to H' after processing, then its height change ΔH = H' - H. The parameter change values ​​of all feature points are recorded. The temporal displacement of each feature point is associated with its morphological parameter change to form a parameter change record table. This table will be used to evaluate the overall processing status of the component. If the temporal displacement of feature point A is D_A and the morphological parameter change is ΔH_A, it will be recorded as "Feature point A: temporal displacement D_A, parameter change ΔH_A." Based on the extracted temporal displacement and parameter change values, the overall processing change state is inferred. By analyzing the changes in each feature point, the overall performance of the component during the processing can be judged.If the temporal displacement of most feature points is within an acceptable range (e.g., less than 0.5 mm), the component's machining condition is considered good. Conversely, if the displacement of some feature points exceeds a set threshold, inspection is required. The change information for all feature points is summarized to form a machining change report. This report details the status of each feature point and the overall machining quality of the component. If analysis reveals abnormal changes in feature points A and B, the report will indicate that these feature points require further inspection or adjustment to ensure component production quality.

[0030] In this embodiment, step S4 includes the following steps: Analyze the parts processing trajectory based on the dynamic pixel optical flow change trajectory and extract the processing trajectory; Mining tool motion paths based on machining process trajectories to extract machining tool motion paths; Performing multi-stage process behavior analysis on the motion path of the machining tool to extract dynamic process behavior characteristics of the multiple stages; Conduct processing path change logic mining on dynamic process behavior characteristics at multiple stages to obtain component processing behavior rules; According to the processing behavior rules of parts, the three-dimensional time series change model of the overall processing change state of parts is carried out to build a real-time processing change model of parts.

[0031] In this embodiment, the process trajectory of component processing is extracted from the dynamic pixel optical flow trajectory obtained in the previous step. This trajectory depicts the relative motion between the tool and the workpiece during the machining process. The motion of the characteristic points in the optical flow trajectory is integrated to form a clear machining trajectory, describing how the tool moves along the workpiece surface during machining. The extracted machining process trajectory is converted into an analyzable data format, typically including parameters such as timestamps, position coordinates, and velocity. This data is used for subsequent path analysis. The process trajectory data is organized in chronological order, and the coordinates and corresponding time information of each key location are recorded to facilitate analysis of the tool's motion characteristics. Based on the extracted machining process trajectory, the tool's motion path is defined. This path reflects how the tool moves along the workpiece surface during machining. If the tool moves along a specific path during machining, information such as the tool's motion direction, velocity, and acceleration are recorded at each stage. Data mining algorithms (such as cluster analysis or path analysis) are used to further explore the tool's motion path. By analyzing changes in the tool path, motion patterns at different machining stages can be identified. Using the K-means clustering algorithm, the tool's motion path during machining is classified into multiple categories, identifying different machining modes (such as cutting and grinding). Based on changes in the tool's motion path, the machining process is divided into multiple stages. Each stage should represent different tool machining behaviors and process characteristics. The machining process is divided into initial cutting, finishing, and post-processing stages. The tool motion characteristics within each stage may vary. Key dynamic process behavior characteristics are extracted from each stage, including tool speed, acceleration, and cutting depth. These characteristics help understand the tool's operating state at each stage. The average tool speed during the initial cutting stage is recorded as 100 mm / min, acceleration is 5 m / s², and cutting depth is 2 mm. Based on the dynamic process behavior characteristics of multiple stages, the logic behind the machining path changes is analyzed. Correlations between different stages are identified to understand how they affect tool motion and machining results. Characteristics of the initial cutting and finishing stages are analyzed. Excessively high cutting speeds in the initial stage can lead to increased workpiece surface roughness in the subsequent finishing stage. The discovered changes are summarized as component machining behavior patterns, forming a theoretical framework to help analyze and predict potential problems and optimization directions during the machining process. The pattern of "workpiece surface quality deteriorating at excessively high cutting speeds" was noted, which will be addressed in future machining. A three-dimensional time-series variation model was constructed based on the component machining behavior patterns. This model should dynamically reflect changes during the machining process, including time, position, and state. Using 3D modeling software (such as MATLAB or Blender), a model was constructed based on the extracted motion paths and behavioral characteristics, reflecting each stage of the machining process and its changes. The time-series variation model was combined with real-time data to construct a real-time component machining variation model.The model should be able to be updated in real time to reflect the current processing status. A dynamic system with real-time data input should be established to automatically adjust the model parameters according to changes in the processing process, forming a real-time updated processing change status.

[0032] In this embodiment, step S5 includes the following steps: Based on the global delay optimization video, the key processing area is identified and divided into local equal-size areas to extract multiple area enlargement videos; Perform visual identification of component defects on enlarged videos of multiple regions, and mark the defective regions of the components; Detect potential processing deviations based on multiple region-enlarged videos and extract potential processing deviation areas; The machining error is calculated for the defective area and potential machining deviation area of ​​the parts, and the machining error parameters of the parts are obtained.

[0033] In this embodiment, critical processing areas are identified from the global delay optimization video. Critical processing areas are typically important parts of a component that are subject to significant stress or change during processing. Image processing algorithms (such as edge detection and region growing) can be used to identify these areas. The Canny edge detection algorithm analyzes edge information in the video frames to identify possible processing areas. These areas may include holes, cutting edges, etc. The identified critical processing areas are then locally divided into multiple sub-areas of equal size. Each sub-area should have similar features to facilitate subsequent processing. For example, if a critical area of ​​50 mm x 50 mm is identified, it can be divided into 10 5 mm x 5 mm sub-areas. Each sub-area will be used for subsequent defect detection and analysis. Based on the multiple sub-areas, corresponding enlarged videos are generated. The video frames of each sub-area can be cropped and enlarged to ensure that the details of each area are clearly visible during analysis. For each 5 mm x 5 mm sub-area, a corresponding enlarged video is generated, with a resolution of at least twice that of the original video to better identify small defects. Each enlarged video is used to perform visual identification of component defects region by region. Use computer vision techniques (such as deep learning models and image segmentation algorithms) to automatically detect and identify defects. Use the YOLO (You Only Look Once) model for object detection, trained to identify possible defect types such as cracks, dents, or surface scratches. Mark identified defect areas and record their location and type. Use rectangular boxes or other markings in the enlarged video to highlight defect areas. If a crack is found in a region, mark it with a red rectangular box and record its coordinates and dimensions for subsequent analysis. Analyze machining deviations in each region based on the enlarged video. These deviations may be caused by tool wear, workpiece deformation, or other factors during machining. Define machining deviation as the difference between actual and designed dimensions, and set a threshold (such as ±0.1 mm) as the criterion for potential deviations. Use image processing techniques (such as template matching or morphological analysis) to detect potential machining deviations in each region. Compare the actual image with the designed model to identify areas of potential deviation. If the designed aperture is 10 mm, but the aperture measured in the actual image is 9.8 mm, mark this area as a potential machining deviation area and record the deviation value. Extract and record the detected potential machining deviation areas. Ensure that the location information and corresponding deviation value of each deviation area are fully preserved. Organize the coordinates and deviation values ​​of all deviation areas into a data table to facilitate subsequent analysis and reporting. Based on the marked defective areas and potential machining deviation areas of the component, calculate the machining error. Machining error is generally defined as the difference between the actual machining state of the component and the design standard.The machining error is defined as the actual dimension of a component minus the designed dimension. For example, if the actual dimension is 9.5 mm and the designed dimension is 10 mm, the machining error is -0.5 mm. Calculate specific error parameters for each defective area and potential deviation area. This includes positional error, dimensional error, and other factors to comprehensively assess the machining quality of the component. Record the error value for each defective area and calculate the overall machining error distribution to create a machining error parameter report. Summarize all calculated component machining error parameters to generate a detailed machining error report. The report should include statistical information on defective areas, potential deviation areas, and their corresponding error values.

[0034] In this embodiment, step S6 includes the following steps: Perform final machining simulation on the real-time machining change model of the component based on the component machining error parameters to obtain a final machining simulation model; Calculate the morphological deviation of the final processing simulation model based on the preset standard processing model and identify the morphological deviation parameters; Performing deviation compensation calculation on the morphological deviation parameter to obtain a feedforward deviation compensation value; Based on the feedforward deviation compensation value, the processing parameters of the automobile parts are corrected in real time to perform precision processing identification optimization operations.

[0035] In this embodiment, the component machining error parameters extracted from the previous step are collected. These parameters include positional error, dimensional error, and morphological error, and can reflect the actual state of the component during machining. Assume that the recorded error parameters include: aperture error of -0.5 mm, edge straightness error of 0.2 mm, etc. Based on the collected machining error parameters, the component's real-time machining variation model is updated. This model should be able to reflect the morphological changes caused by errors during machining. Using finite element analysis (FEA), the collected error parameters are input into the model, and the component's geometry is updated in real time to generate a final machining simulation model. The real-time machining variation model is simulated to generate the final machining simulation model. This model should fully reflect the actual machining state, facilitating subsequent deviation calculation and compensation. In the final machining simulation model, the aperture size is adjusted to 9.5 mm to reflect the actual machining results and is visualized. The ideal shape of the component is defined based on a pre-set standard machining model. This model is typically based on a design drawing or CAD model and contains all ideal parameters. Assume that in the standard machining model, the designed aperture is 10 mm, and the edges are expected to be straight lines. Compare the final machining simulation model with the standard machining model to calculate morphological deviations. This calculation method typically subtracts the designed dimensions from the actual dimensions, identifying the deviation parameters for each feature. If the aperture in the final model is 9.5 mm, the morphological deviation is 9.5 mm - 10 mm = -0.5 mm. Record this deviation value. Record all calculated morphological deviation parameters, including aperture error and edge error. This data will be used for subsequent compensation calculations. Record all deviation parameters, such as -0.5 mm aperture deviation and 0.2 mm edge deviation, and organize them into a table. Based on the morphological deviation parameters, apply compensation calculation theory to determine the feedforward deviation compensation value. Feedforward compensation involves adjusting parameters based on predicted deviations before machining to reduce the final error. If the aperture deviation is -0.5 mm, the compensation value should be set to +0.5 mm to ensure that the final aperture meets the design requirements. Calculate the compensation value for each deviation parameter to ensure that the compensated machining parameters offset the original deviation. If the edge deviation is 0.2 mm, the corresponding compensation value should be -0.2 mm, and the machining tool path should be adjusted to achieve this compensation. All calculated feedforward deviation compensation values ​​are recorded to form a compensation parameter table. These compensation values ​​will serve as the basis for subsequent machining parameter corrections. The compensation values ​​are recorded as +0.5 mm for the aperture and -0.2 mm for the edge, ensuring that each parameter has a clear compensation basis. Based on the feedforward deviation compensation values, the machining process parameters of automotive parts are immediately corrected. These parameters may include tool feed rate, cutting depth, and tool path. If the aperture compensation value is +0.5 mm, the tool feed rate needs to be adjusted during the machining process to ensure that the final aperture reaches 10 mm.During the actual machining process, the corrected process parameters are applied in real time. This can be adjusted through the control interface of the CNC system or machining equipment. The CNC machine tool's tool feed rate is set to the actual required value, ensuring real-time response during the machining process and adjusting based on the compensation value. Using the corrected process parameters, precision machining identification and optimization are performed to ensure that the machining quality of the part meets the design requirements. After the adjustments, a complete machining process is carried out, and the final part's dimensional and morphological parameters are measured to ensure that all features meet the standards.

[0036] In this embodiment, a precision identification device for automobile parts processing is provided, which is used to perform the precision identification method for automobile parts processing as described above, including: The video optimization module is used to obtain component processing monitoring videos, adjust dynamic video imaging parameters, and perform full video delay synchronization processing to obtain delayed synchronization optimized videos; The 3D morphology calculation module is used to visually identify the microstructure of components based on the delayed synchronous optimization video, perform 3D morphology geometry calculations, and extract the 3D morphology parameters of each feature point; An optical flow change tracking module is used to track the optical flow changes of the delayed synchronization optimized video frame by frame, and calculate the morphological parameter changes of the three-dimensional morphological parameters to obtain the overall processing change state of the component; The 3D modeling module is used to mine the logic of machining path changes based on the delayed synchronization optimization video, and then perform 3D time series change modeling on the overall machining change state of the component to build a real-time machining change model for the component; Deviation detection module, used to detect potential processing deviations and calculate processing errors for the global delay optimization video, and obtain component processing error parameters; The parameter correction module is used to perform final processing simulation on the real-time processing change model of the parts based on the processing error parameters of the parts, and to perform instant processing process parameter correction to perform precision processing identification and optimization operations.

[0037] This invention uses a high-resolution camera to capture processing video, ensuring clear and detailed image information. The optimized video ensures precise capture of the component surface and the processing process. This module dynamically adjusts video imaging parameters (such as exposure, gain, and white balance) to account for various lighting and reflection variations, ensuring that the video maintains excellent visual quality under various working conditions. The optimized image quality is enhanced, reducing image blur or errors caused by lighting variations. Synchronous processing eliminates time delays in the video, enabling precise alignment of data from multiple cameras and sensors. This process effectively reduces timing inconsistencies caused by video delays, making subsequent processing analysis more accurate, which is particularly crucial in high-speed production lines. By combining image depth information with geometric calculation methods, 3D morphological parameters (such as depth, curvature, and angle) are extracted for each feature point. This 3D data provides accurate spatial information for subsequent precision analysis and error detection, clearly depicting the spatial variations of the component. Accurate calculation of the component's 3D morphology provides an accurate reference model for subsequent detection and compensation of processing errors. This process is particularly important for precision parts, ensuring they meet design requirements. Optical flow algorithms can track minute displacements of component surfaces or structures frame by frame, accurately capturing subtle deformations during machining. This is crucial for dynamically monitoring real-time changes during machining, enabling the identification of small deformations that are difficult to detect using static images. Combined with 3D morphological data, the module calculates morphological changes at each feature point of the component, such as positional offset and dimensional deformation, thereby reflecting the overall machining status of the component. In-depth analysis of machining path changes reveals the individual steps and path logic involved in the component machining process. This provides fundamental data for process optimization, helping engineers identify potential anomalies and irregularities during production. By combining 3D data from each machining stage with time series data, a dynamic 3D machining change model is constructed, reflecting the real-time changes in component morphology during machining. This makes the machining process more predictable and controllable. 3D temporal change modeling enables real-time monitoring of component machining processes, identifying potential deviations and enabling adjustments to ensure accuracy and consistency at each machining step. By analyzing globally optimized videos, potential deviations in the machining process, such as dimensional and shape deviations, can be detected in real time. Early detection of problems can prevent subsequent quality issues and reduce scrap rates. Image analysis technology accurately calculates machining errors and provides quantitative data for subsequent corrections. Compared to traditional manual inspection, this calculation method is more accurate and reduces human error. This module continuously monitors the entire machining process, ensuring that all machining deviations remain within the predetermined tolerance range, thereby improving the overall quality stability of the component. Real-time feedback on error parameters enables automatic adjustment of machining parameters, enabling real-time process optimization.Ensure that the component processing process can be dynamically adjusted based on feedback errors to ensure accuracy. Real-time compensation calculations for processing errors can significantly reduce quality issues caused by error accumulation. Adjustment of processing parameters ensures that the size and shape of the component meet preset standards.

[0038] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0039] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A precision identification method for automobile parts processing, characterized in that: The following steps are involved: Step S1: Obtaining a component processing monitoring video, and performing dynamic video imaging parameter adjustment and full video delay synchronization processing to obtain a delay synchronization optimized video; Step S2: Visually recognize the microstructure of the component based on the delayed synchronous optimization video, perform three-dimensional morphological geometry calculation, and extract the three-dimensional morphological parameters of each feature point; Step S3: tracking optical flow changes frame by frame on the delayed synchronization optimization video, and calculating morphological parameter changes on the three-dimensional morphological parameters to obtain the overall processing change state of the component; Step S4: mining the logic of machining path changes based on the delayed synchronization optimization video, and then modeling the three-dimensional temporal changes of the overall machining change state of the component to build a real-time machining change model of the component; Step S5: Perform potential processing deviation detection and processing error calculation on the global delay optimization video to obtain component processing error parameters; Step S6: Perform final processing simulation on the real-time processing change model of the component based on the component processing error parameters, and perform real-time processing parameter correction to perform precision processing identification optimization operations.

2. The precise identification method for automobile parts processing according to claim 1, characterized in that: The specific steps of step S1 are: Acquire component processing monitoring videos using high-resolution industrial cameras; Adjust dynamic video imaging parameters of component processing monitoring videos to construct imaging-optimized monitoring videos; Extracting the timestamp of each frame of the imaging optimization monitoring video; Performing a global frame delay average calculation according to the timestamp to obtain an inter-frame average delay value; Performing frame-by-frame delay correction calculation on the timestamp of each frame according to the average delay value between frames to generate a multi-frame correction time difference; Full video delay synchronization processing is performed based on the multi-frame correction time difference, thereby obtaining a delay synchronization optimized video.

3. The precise identification method for automobile parts processing according to claim 2, characterized in that: The specific steps of adjusting the dynamic video imaging parameters of the component processing monitoring video and constructing the imaging optimization monitoring video are as follows: Extracting an initial frame image of the video based on the component processing monitoring video; Calculate the surface reflectivity of the components on the initial frame of the video and extract the surface reflectivity; Analyze the material properties of parts based on the initial frame of the video Identify light intensity changes based on the initial frame image of the video to obtain light change features Dynamically adjusting video imaging parameters based on surface light reflectivity, material properties of the components, and light change characteristics to obtain dynamically optimized imaging parameters; Based on the dynamic optimization of imaging parameters, the global video parameters of the component processing monitoring video are optimized to construct the imaging optimized monitoring video.

4. The precise identification method for automobile parts processing according to claim 1, characterized in that: The specific steps of step S2 are: Perform visual recognition of component microstructures based on delayed synchronously optimized videos and mark multiple microstructure feature points; Calculating the spatial positioning coordinates of each of the micro-structural feature points; Perform microstructure feature type analysis on multiple tiny structural feature points to obtain the type of each feature point; The three-dimensional morphological geometry calculation is performed according to the type of each feature point, and the three-dimensional morphological parameters of each feature point are extracted.

5. The precise identification method for automobile parts processing according to claim 1, characterized in that: The specific steps of step S3 are: Track the optical flow changes frame by frame in the delayed synchronization optimized video and extract the dynamic pixel optical flow change trajectory; Identify the spatial position change of the spatial positioning coordinates according to the dynamic pixel optical flow change trajectory, and extract the temporal displacement of each feature point; Calculating the morphological parameter changes of the three-dimensional morphological parameters according to the dynamic pixel optical flow change trajectory to obtain the parameter change value of each feature point; The processing change is inferred based on the time sequence displacement and the parameter change value to obtain the overall processing change state of the component.

6. The precise identification method for automobile parts processing according to claim 1, characterized in that: The specific steps of step S4 are: Analyze the parts processing trajectory based on the dynamic pixel optical flow change trajectory and extract the processing trajectory; Mining tool motion paths based on machining process trajectories to extract machining tool motion paths; Performing multi-stage process behavior analysis on the motion path of the machining tool to extract dynamic process behavior characteristics of the multiple stages; Conduct processing path change logic mining on dynamic process behavior characteristics at multiple stages to obtain component processing behavior rules; According to the processing behavior rules of parts, the three-dimensional time series change model of the overall processing change state of parts is carried out to build a real-time processing change model of parts.

7. The precise identification method for automobile parts processing according to claim 1, characterized in that: The specific steps of step S5 are: Based on the global delay optimization video, the key processing area is identified and divided into local equal-size areas to extract multiple area enlargement videos; Perform visual identification of component defects on each region of enlarged videos of multiple regions and mark the defective regions of the components; Detect potential processing deviations based on expanded videos of multiple regions and extract potential processing deviation areas; The machining error is calculated for the defective area and potential machining deviation area of ​​the parts, and the machining error parameters of the parts are obtained.

8. The precise identification method for automobile parts processing according to claim 1, characterized in that: The specific steps of step S6 are: Perform final machining simulation on the real-time machining change model of the component based on the component machining error parameters to obtain a final machining simulation model; Calculate the morphological deviation of the final processing simulation model based on the preset standard processing model and identify the morphological deviation parameters; Performing deviation compensation calculation on the morphological deviation parameter to obtain a feedforward deviation compensation value; Based on the feedforward deviation compensation value, the processing parameters of the automobile parts are corrected in real time to perform precision processing identification optimization operations.

9. A precision identification device for automobile parts processing, characterized in that: The method for performing the precision identification of automobile parts processing according to claim 1 comprises: The video optimization module is used to obtain component processing monitoring videos, adjust dynamic video imaging parameters, and perform full video delay synchronization processing to obtain delayed synchronization optimized videos; The 3D morphology calculation module is used to visually identify the microstructure of components based on the delayed synchronous optimization video, perform 3D morphology geometry calculations, and extract the 3D morphology parameters of each feature point; An optical flow change tracking module is used to track the optical flow changes of the delayed synchronization optimized video frame by frame, and calculate the morphological parameter changes of the three-dimensional morphological parameters to obtain the overall processing change state of the component; The 3D modeling module is used to mine the logic of machining path changes based on the delayed synchronization optimization video, and then perform 3D time series change modeling on the overall machining change state of the component to build a real-time machining change model for the component; Deviation detection module, used to detect potential processing deviations and calculate processing errors for the global delay optimization video, and obtain component processing error parameters; The parameter correction module is used to perform final processing simulation on the real-time processing change model of the parts based on the processing error parameters of the parts, and to perform instant processing process parameter correction to perform precision processing identification and optimization operations.

Images