Saw chain processing method and system combined with visual inspection

Abstract

The present application provides a saw chain processing method and system combined with visual detection, which relates to the field of processing data processing technology, including: obtaining real-time processing technology information through intelligent sensing equipment; obtaining a real-time saw chain image set through dynamic acquisition of intelligent image acquisition equipment; introducing a predetermined modeling strategy to analyze the first image and construct a three-dimensional model; analyzing the comparison result between the three-dimensional model and the predetermined three-dimensional model to obtain a real-time processing anomaly index; when the real-time processing anomaly index reaches a predetermined anomaly index threshold, issuing a processing anomaly warning signal; calling a predetermined processing control mechanism, and dynamically processing and controlling the real-time processing technology information. The present application can solve the technical problems in the prior art that the processing accuracy is unstable and the product quality is affected by the complex structure of the saw chain and is easily affected by various factors. Through real-time monitoring and anomaly detection, problems in the processing process can be dynamically adjusted, thereby improving the processing accuracy and product quality of the saw chain.

Description

Technical Field

[0001] The present application relates to the technical field of processing data, and in particular to a saw chain processing method and system combined with visual inspection. Background Art

[0002] The saw chain is composed of multiple ring links, chains and teeth, each of which is responsible for cutting. Its structure is complex, involving multiple teeth and fine geometric shapes. The angle, shape and size of each tooth have a crucial impact on the cutting performance and service life of the saw chain. The processing of the saw chain usually includes multiple links such as the manufacture of links, grinding of teeth, trimming of tooth shape, and assembly of chains. However, in the actual processing process, it is often affected by many factors, such as equipment wear, unstable process parameters (such as feed speed, cutting speed, tool pressure, etc.), etc., which can easily lead to unstable saw chain processing accuracy, abnormal tooth shape, excessive surface roughness, broken teeth and other problems, thus affecting product quality.

[0003] In summary, the prior art has a technical problem in that the saw chain has a complex structure and is susceptible to various factors, resulting in unstable processing accuracy, which further affects product quality. Summary of the invention

[0004] The purpose of this application is to provide a saw chain processing method and system combined with visual inspection, so as to solve the technical problem in the prior art that the saw chain has a complex structure and is susceptible to various factors, resulting in unstable processing accuracy, which further affects product quality.

[0005] In order to achieve the above objectives, the present application provides a saw chain processing method and system combined with visual inspection.

[0006] In a first aspect, the present application provides a saw chain processing method combined with visual detection, wherein the saw chain processing method combined with visual detection is implemented by a saw chain processing system combined with visual detection, wherein the saw chain processing method combined with visual detection comprises: dynamically monitoring the saw chain processing process by means of an intelligent sensing device to obtain real-time processing technology information; dynamically acquiring a real-time saw chain image set by means of an intelligent image acquisition device arranged on the saw chain, wherein the real-time saw chain image set comprises a first image, wherein the first image refers to a saw chain image acquired by means of a first collector in the intelligent image acquisition device; introducing a predetermined modeling strategy to analyze the first image, and constructing a three-dimensional model of the saw chain; analyzing the comparison result between the three-dimensional model and the predetermined three-dimensional model of the saw chain to obtain a real-time processing abnormality index of the saw chain; when the real-time processing abnormality index reaches a predetermined abnormality index threshold, issuing a processing abnormality warning signal; based on the processing abnormality warning signal, calling a predetermined processing control mechanism, and dynamically controlling the real-time processing technology information.

[0007] Optionally, extract a subject-background slicing strategy from the predetermined modeling strategy; slice the first image according to the subject-background slicing strategy to obtain a first slicing result, and extract a first subject slice from the first slicing result; obtain a first acquisition control parameter of the first acquisition device; perform a structural analysis on the first subject slice in combination with the first acquisition control parameter to obtain a three-dimensional structure of the saw chain; and align the first point cloud data of the first image to the three-dimensional structure to obtain the three-dimensional model.

[0008] Optionally, the first image is rasterized to obtain a first raster group; the first raster in the first raster group is locally binary-converted to obtain a first binary code, and the first binary code group is formed; the first binary code group is screened with a preset saw chain binary code as a constraint to obtain a first target binary code group; the first main body slice is calibrated based on the first target raster group obtained by reverse matching the first target binary code group.

[0009] Optionally, multi-dimensional features of the three-dimensional structure are collected to obtain structural feature parameters; the structural feature parameters are compared with predetermined structural feature parameters of the predetermined three-dimensional model to obtain a structural comparison result; it is determined and analyzed that the structural abnormality index obtained by the structural comparison result is within the predetermined abnormality index threshold; if so, a structural abnormality prediction model is activated to analyze the structural feature parameters to obtain a first predicted abnormality category; a first control indicator corresponding to the first predicted abnormality category is matched in a predetermined processing control mapping list stored in the predetermined processing control mechanism; and parameter information corresponding to the first control indicator in the real-time processing technology information is controlled.

[0010] Optionally, if it is not, multi-dimensional features of the three-dimensional model are collected to obtain model feature parameters; the model feature parameters are compared with predetermined model feature parameters of the predetermined three-dimensional model to obtain model comparison results; the model comparison results are determined and analyzed to determine whether the model abnormality index is within the predetermined abnormality index threshold; if it is, the model abnormality prediction model is activated to analyze the model feature parameters to obtain a second predicted abnormality category; the second control indicator corresponding to the second predicted abnormality category is matched in the predetermined processing control mapping list; and the parameter information corresponding to the second control indicator in the real-time processing technology information is controlled.

[0011] Optionally, an exception category set is formed, and any exception category in the exception category set is extracted; a process indicator set is formed, and a correlation analysis is performed between the arbitrary exception category and the process indicator set to obtain any correlation result; the arbitrary correlation result is analyzed and significantly correlated process indicators are screened to form an arbitrary process indicator set of the arbitrary exception category; the predetermined processing control mapping list is formed according to the mapping relationship between the arbitrary exception category and the arbitrary process indicator set.

[0012] Optionally, the abnormality category set includes a structural abnormality category set and a model abnormality category set, and the structural abnormality category set includes at least tooth shape abnormality, broken tooth abnormality and deformation abnormality, and the model abnormality category set includes at least fracture abnormality and surface roughness abnormality.

[0013] In a second aspect, the present application further provides a saw chain processing system combined with visual detection, which is used to execute the saw chain processing method combined with visual detection as described in the first aspect, wherein the saw chain processing system combined with visual detection includes: a real-time information acquisition module, the real-time information acquisition module is used to dynamically monitor the saw chain processing process through an intelligent sensing device to obtain real-time processing technology information; an image acquisition module, the image acquisition module is used to dynamically acquire a real-time saw chain image set through an intelligent image acquisition device arranged on the saw chain, the real-time saw chain image set includes a first image, wherein the first image refers to an image acquired by a first collector in the intelligent image acquisition device. A saw chain image; a model building module, the model building module is used to introduce a predetermined modeling strategy to analyze the first image, and construct a three-dimensional model of the saw chain; an abnormality comparison module, the abnormality comparison module is used to analyze the comparison result of the three-dimensional model with the predetermined three-dimensional model of the saw chain to obtain the real-time processing abnormality index of the saw chain; an abnormality warning module, the abnormality warning module is used to issue a processing abnormality warning signal when the real-time processing abnormality index reaches a predetermined abnormality index threshold; a processing control module, the processing control module is used to call a predetermined processing control mechanism based on the processing abnormality warning signal, and dynamically process and control the real-time processing process information.

[0014] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0015] The saw chain processing process is dynamically monitored by intelligent sensing equipment to obtain real-time processing information; the real-time saw chain image set is dynamically acquired by an intelligent image acquisition device arranged on the saw chain, and the real-time saw chain image set includes a first image, wherein the first image refers to the saw chain image acquired by the first acquisition device in the intelligent image acquisition device; a predetermined modeling strategy is introduced to analyze the first image, and a three-dimensional model of the saw chain is constructed; the comparison result between the three-dimensional model and the predetermined three-dimensional model of the saw chain is analyzed to obtain the real-time processing abnormality index of the saw chain; when the real-time processing abnormality index reaches the predetermined abnormality index threshold, a processing abnormality warning signal is issued; based on the processing abnormality warning signal, a predetermined processing control mechanism is called, and the real-time processing information is dynamically processed and controlled. That is to say, through the dynamic monitoring of intelligent sensing equipment and image acquisition equipment, various process data in the saw chain processing process are obtained in real time, a three-dimensional model of the saw chain is constructed, the actual processing results are compared with the predetermined three-dimensional model, any morphological or dimensional deviations in the processing process are accurately identified, the processing parameters are dynamically adjusted, the processing technology is optimized, and the processing accuracy and product quality of the saw chain are improved.

[0016] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0018] Figure 1 A schematic diagram of the process of the saw chain processing method combined with visual inspection in the present application;

[0019] Figure 2 This is a schematic diagram of the structure of the saw chain processing system combined with visual inspection in this application.

[0020] Explanation of the accompanying drawings: real-time information acquisition module 11, image acquisition module 12, model building module 13, abnormality comparison module 14, abnormality warning module 15, processing control module 16. DETAILED DESCRIPTION

[0021] This application solves the technical problem in the prior art that the saw chain has a complex structure and is susceptible to various factors, resulting in unstable processing accuracy, which further affects product quality, by providing a saw chain processing method and system combined with visual inspection. Through dynamic monitoring of intelligent sensing equipment and image acquisition equipment, various process data in the saw chain processing process are obtained in real time, a three-dimensional model of the saw chain is constructed, and the actual processing results are compared with the predetermined three-dimensional model. Any morphological or dimensional deviations in the processing process are accurately identified, and processing parameters are dynamically adjusted to optimize the processing technology, thereby improving the processing accuracy and product quality of the saw chain.

[0022] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0023] For example, please refer to the attached Figure 1 The present application provides a saw chain processing method combined with visual detection, wherein the saw chain processing method combined with visual detection is applied to a saw chain processing system combined with visual detection, and the saw chain processing method combined with visual detection specifically includes the following steps:

[0024] Step 1: Use intelligent sensing equipment to dynamically monitor the saw chain processing process to obtain real-time processing information.

[0025] Specifically, intelligent sensing equipment is used to monitor each link of the processing process in real time and obtain real-time processing information, such as processing speed, cutting temperature, cutting force, tool wear, machine vibration, etc. Intelligent sensing equipment is a device that integrates sensors, processing units and communication modules to monitor changes in physical or chemical processes in real time, obtain relevant data and transmit and analyze it through the network or other means. Sensors include temperature sensors, pressure sensors, acceleration sensors, etc. In saw chain processing, it is possible to monitor parameters such as temperature, pressure, vibration, speed, etc. during the processing and convert them into readable signals.

[0026] The saw chain processing refers to the production and manufacturing process of the saw chain (the cutting part on the chain saw, composed of multiple links with teeth), including cutting, grinding, heat treatment and other steps. Dynamic monitoring refers to the real-time acquisition of relevant data throughout the entire processing process. It is not a single data at a static moment, but continuous tracking and recording of data changes as the processing proceeds, capturing abnormal fluctuations, trend changes and other information during the processing. By continuously collecting and monitoring processing technology information in real time, it ensures that the processing accuracy of each link meets the requirements, avoids out-of-tolerance problems, and reduces the chances of later corrections and rework.

[0027] Step 2: Dynamically acquire a real-time saw chain image set through an intelligent image acquisition device arranged on the saw chain, wherein the real-time saw chain image set includes a first image, wherein the first image refers to a saw chain image acquired by a first collector in the intelligent image acquisition device.

[0028] Specifically, during the saw chain processing, the intelligent image acquisition device is placed at an appropriate position of the saw chain processing equipment, usually near the processing machine or above the processing area, such as the end of the processing line or the inspection station. Select a suitable image acquisition device, such as a high-resolution camera or an industrial camera, and regularly collect images of the saw chain according to the predetermined monitoring area and time sequence. The saw chain image set is dynamically collected by the intelligent image acquisition device, and the real-time saw chain image set is continuously collected in chronological order to reflect the real-time status of the saw chain during the processing process, including the size, shape, surface quality, etc. of the saw chain.

[0029] The real-time saw chain image set includes but is not limited to the first image, which is collected by the first collector in the device and represents a specific perspective or state of the saw chain. It may contain important information such as the wear, shape change, and surface defects of the saw chain. The first collector refers to an image sensor or camera in the intelligent image acquisition device, which is used to collect specific image data. For example, assume that during the tooth shape processing of the saw chain, the intelligent image acquisition device is arranged in front of the processing line. The first collector is responsible for photographing the preliminary tooth shape of each saw chain and marking the collected image as the first image. When the saw chain passes through the subsequent processing link, other collectors in the device will also collect new images to form subsequent image data sets. Assume that the first image shows the preliminary shape and size of the saw chain, and the subsequent images show the fine details after processing. By comparing these images, it is possible to detect whether there is a tooth shape deviation (such as a blunt tooth angle or a large size) during the processing process and issue an alarm in time. By collecting saw chain images in real time, abnormal conditions in the processing process, such as wear and breakage, can be quickly discovered.

[0030] Step three: introducing a predetermined modeling strategy to analyze the first image and construct a three-dimensional model of the saw chain.

[0031] Specifically, the predetermined modeling strategy is an analysis method based on image features, which is used to extract representative structures or features from the collected image data. The predetermined modeling strategy can be a specific algorithm or process that helps to extract meaningful information from the image, such as edge, surface or morphological features. The collected first image is analyzed by the predetermined modeling strategy, including image segmentation, feature extraction, structure analysis and other methods, with the purpose of extracting useful geometric information and structural features from the image. For example, using the image processing algorithm in OpenCV, the first image is firstly edge detected to extract the contour information of the saw chain.

[0032] The first image is preprocessed, including denoising, contrast enhancement, and image segmentation, and the image is analyzed using a predetermined modeling strategy, which usually includes a series of image processing, feature extraction, and geometric derivation methods. The subject-background slicing strategy in the predetermined modeling strategy is extracted to slice the first image, separate the subject from the background, and extract the subject part of the first image. Afterwards, the extracted subject part is compared with the standard model, and the subject part in the image is adjusted to match the standard form more accurately.

[0033] The first acquisition control parameters of the first acquisition device that takes the first image are obtained, including angle, distance, focal length, etc., and the main part of the extracted first image is deeply analyzed, including determining the outline of the saw chain, identifying the key points on the saw chain, etc., and the geometric model of the saw chain, that is, the three-dimensional structure, is constructed according to the extracted features. Point cloud data is extracted from the first image, and each point represents the position of a pixel in the image in three-dimensional space. The three-dimensional structure is aligned according to the layout point cloud data to create an accurate three-dimensional model. The image is analyzed and a three-dimensional model is constructed through a predetermined modeling strategy, the geometric information of the saw chain is accurately obtained, an accurate three-dimensional model is established, the design of the saw chain is optimized, the occurrence of unqualified products is reduced, and the processing accuracy and product performance are improved.

[0034] Step 4: Analyze the comparison result between the three-dimensional model and the predetermined three-dimensional model of the saw chain to obtain the real-time processing abnormality index of the saw chain.

[0035] Specifically, the actual three-dimensional model is compared with the predetermined three-dimensional model to obtain the comparison result. The predetermined three-dimensional model refers to the ideal three-dimensional model of the saw chain defined based on design requirements or standards. It represents the ideal shape and size of the saw chain and is usually generated by the design team or CAD system according to the standard design of the saw chain. The predetermined three-dimensional model is used as the quality standard of the processing process, and all processed saw chains should be as consistent as possible with the model. The core of the comparison process is to compare the actual three-dimensional model with the predetermined three-dimensional model point by point, face by face, or part by part, including point cloud matching, shape difference, volume error, etc. Match the point cloud data of the actual three-dimensional model and the predetermined three-dimensional model, and calculate the distance error between each corresponding point; compare the shape differences of the model surface or contour to identify the deviations of the model in size, shape, etc.; calculate the difference in volume between the two models to evaluate whether there is too much or insufficient material in the processing process.

[0036] The comparison result refers to the deviation or difference obtained by comparing the actual 3D model with the predetermined 3D model, such as dimensional error, shape error, surface defect, etc., which reflects the gap between the actual saw chain and the ideal saw chain. The real-time processing anomaly index is calculated using error analysis methods, such as root mean square error (the mean of the sum of squares of each error in the comparison result, and taking the square root), maximum deviation (the maximum error value in the comparison result), weighted error (different weights are given to errors according to the importance of different parts, and the weighted processing anomaly index is obtained). The real-time processing anomaly index is a quantitative indicator used to evaluate the deviation between the actual saw chain and the predetermined design model during the current processing. By comparing the geometric differences of the 3D models, the real-time processing anomaly index is calculated to reflect any deviation that does not meet the design requirements during the processing. By comparing the real-time 3D model with the predetermined model, the deviation in the processing is detected in real time, the processing anomaly is discovered in time, and defective products are avoided in the production process. The real-time processing anomaly index can be calculated to accurately understand the degree of processing deviation, help adjust the processing parameters, and ensure the accuracy of the final product.

[0037] Step 5: When the real-time processing abnormality index reaches a predetermined abnormality index threshold, a processing abnormality warning signal is issued.

[0038] Specifically, the predetermined abnormal index threshold is a pre-set threshold used to determine whether there is an abnormality in the processing process. When the actual calculated real-time processing abnormality index reaches the predetermined abnormal index threshold, it is considered that there is an abnormality in the processing process and an alarm is automatically triggered. The predetermined abnormal index threshold is usually set based on factors such as product quality requirements, processing accuracy or equipment capabilities.

[0039] When the real-time processing abnormality index exceeds the predetermined abnormality index threshold, it is determined that there is an abnormality in the current processing process, triggering an early warning signal to remind the operator or processing equipment to adjust the processing parameters or check the processing equipment. The early warning signal can be issued in a variety of ways, such as sound alarm, visual alarm, screen prompt or automatic notification to the operator. For example, when the threshold value of the processing abnormality index is determined to be 0.05, when the calculated real-time processing abnormality index reaches 0.15, it exceeds the set threshold value (0.05), and the processing abnormality is automatically judged to occur. The red warning light is displayed on the operation panel through the visual alarm, and the sound alarm is issued at the same time to remind the operator to check the processing process. By monitoring and comparing the real-time processing abnormality index, the deviation or quality problems that may occur in the processing process can be quickly detected, thereby ensuring the stability of the processing process.

[0040] Step six: based on the abnormal processing warning signal, a predetermined processing control mechanism is retrieved, and dynamic processing control is performed on the real-time processing technology information.

[0041] Specifically, the processing abnormality warning signal refers to the abnormal situation or potential risk that occurs during the processing process, which is detected by monitoring and analyzing the real-time data during the processing. When the processing abnormality index reaches the predetermined abnormality index threshold, an abnormality warning signal is generated. The predetermined processing control mechanism refers to a set of pre-set processing measures to deal with processing abnormalities, which triggers corresponding control measures according to different types of abnormalities (such as tooth shape abnormalities, surface roughness abnormalities, etc.), such as adjusting the cutting speed, feed speed, tool angle, etc. The predetermined processing control mechanism includes a predetermined processing control mapping list, that is, a list consisting of each abnormality category and its corresponding process indicator set.

[0042] During the processing, the processing data is collected in real time and the comparison results between the three-dimensional model and the predetermined three-dimensional model are analyzed to determine whether there are processing anomalies. Whether there are significant deviations between the structural characteristic parameters (such as tooth shape, size, etc.) of the three-dimensional model and the structural characteristic parameters of the predetermined three-dimensional model, if the structural anomaly index reaches the predetermined anomaly index threshold, the structural anomaly prediction model is activated to analyze whether there are problems such as tooth shape abnormality, broken teeth, deformation, etc., and generate corresponding control measures to dynamically control the corresponding parameter information in the real-time processing technology information. If the predetermined anomaly index threshold is not reached, other features of the three-dimensional model (such as surface quality, surface roughness, etc.) are compared with the predetermined three-dimensional model to obtain the model comparison results. If the model anomaly index exceeds the predetermined threshold, the model anomaly prediction model is activated for further analysis to determine whether there are problems such as fracture and excessive surface roughness, and control measures are generated based on the prediction results.

[0043] Based on the analysis results of the structural anomaly prediction model or the model anomaly prediction model, select the corresponding control indicators from the predetermined processing control mapping list. These indicators usually include cutting speed, feed speed, tool selection, temperature control, etc. Adjust the processing parameters in real time according to the control indicators to optimize the processing process, improve product quality and avoid anomalies. By monitoring the process status in real time during the processing and responding quickly to abnormal situations to avoid further deterioration of the problem, automatically adjust the process parameters such as cutting speed and feed speed, improve production efficiency without affecting product quality, reduce processing cycle, reduce scrap rate, improve processing quality, and maximize production efficiency.

[0044] Further, step three of this application includes:

[0045] Extracting the subject-background slicing strategy in the predetermined modeling strategy; slicing the first image according to the subject-background slicing strategy to obtain a first slicing result, and extracting a first subject slice in the first slicing result; acquiring a first acquisition control parameter of the first collector; performing a structural analysis on the first subject slice in combination with the first acquisition control parameter to obtain the three-dimensional structure of the saw chain; aligning the first point cloud data of the first image to the three-dimensional structure to obtain the three-dimensional model.

[0046] Specifically, the subject-background slicing strategy in the predetermined modeling strategy is extracted to divide the image into two parts: the subject and the background. The subject refers to the saw chain itself, while the background refers to the processing environment or surrounding debris. By slicing, the image is decomposed into multiple levels, each layer represents a part of the image, and the saw chain part in the image can be extracted, thereby eliminating background noise and improving the accuracy of image analysis. For example, the saw chain is identified using threshold segmentation or deep learning methods, and the image is sliced ​​into several levels, focusing on extracting key parts of the saw chain, such as chain tooth shape, metal parts, etc.

[0047] The first image is sliced ​​using the subject-background slicing strategy to divide the image into multiple parts (slices), and the first slicing result is extracted, which is usually one or more image slices, each slice containing a part of the image. The first slicing result usually refers to the slice part related to the subject, which may contain the key feature part of the saw chain. The first subject slice is extracted from the first slicing result, which contains the main object (saw chain) and the key geometric information of the saw chain. The slicing process includes image preprocessing (denoising, contrast enhancement, etc.), subject recognition, slice segmentation, etc., to separate the saw chain (subject) from the background and obtain the first subject slice.

[0048] Get the first acquisition control parameters of the first collector, that is, the parameters that affect the image acquisition quality and viewing angle, such as angle, distance, focal length, etc. Angle refers to the viewing angle between the camera or sensor and the saw chain object; distance refers to the distance between the camera and the object (such as the saw chain), and different distances will affect the field of view and detail clarity of the image; focal length determines the focusing range of the image, affecting the depth and details of the image. For example, if the distance between the camera and the saw chain is close, the acquired image will have a higher detail resolution; if the angle between the camera and the saw chain is large, it may cause a distortion of the viewing angle, affecting the accuracy of the model.

[0049] The first main body slice is structurally analyzed by using the first acquisition control parameter, and the slice is converted into three-dimensional geometric information by computer vision algorithm or geometric analysis method, so as to obtain the three-dimensional structure of the saw chain. The structural analysis not only considers the two-dimensional features of the image, but also combines the acquisition angle, distance and other information to ensure the accuracy of the three-dimensional modeling, which helps to reveal the geometric properties of the object such as the size, shape, and spatial distribution. Geometric features such as the width, depth, angle, and contour curve of the tooth profile are extracted from the first main body slice. The geometric analysis tool in the image processing software (such as matlab or OpenCV) is used to analyze the specific area in the image to obtain the size, shape and other structural features of the tooth profile. The geometric features are converted into three-dimensional data by combining the control parameters of the collector (such as angle, distance, etc.). The extracted two-dimensional geometric features (such as tooth profile) are reconstructed in three dimensions according to the control parameters of the collector by reverse modeling technology. Alternatively, a modeling technology based on point cloud data (such as multi-view geometric reconstruction) is used to analyze the image data at different perspectives to construct a complete three-dimensional model of the saw chain.

[0050] After structural analysis and 3D modeling, the complete 3D structure of the saw chain is obtained, including the spatial distribution of each part of the saw chain, the precise size of the tooth shape, the geometric shape of the chain, etc. Point cloud data is a set of data points in a 3D space collected by laser scanning, structured light scanning or other 3D scanning technologies, and usually contains the 3D coordinates of each point on the surface of an object. The first point cloud data of the first image, that is, the point coordinate data of the first image, is obtained by 3D scanning, and is registered to the 3D structure to obtain a complete 3D model.

[0051] Point cloud registration is the process of aligning multiple point cloud data sources to ensure that the various parts of the generated 3D model are accurately spliced. The goal of registration is to accurately align the point cloud data with the 3D structure model so that each point in the point cloud can be accurately mapped to the corresponding position of the 3D structure. It is usually divided into two steps: roughly aligning the point cloud data with the 3D structure through preliminary estimates (such as known landmarks or preliminary alignment), that is, coarse registration; further refining the alignment process through algorithms (such as ICP algorithms) so that each point of the point cloud data and the 3D structure matches as accurately as possible. Through the registration and mesh generation process, a complete and accurate 3D model of the saw chain is obtained, showing all the structural features of the saw chain, including the geometric details of each part such as the tooth shape and chain. Combine image analysis and acquisition control parameters for 3D modeling to accurately restore the actual shape of the saw chain and reduce errors caused by factors such as shooting angle and distance.

[0052] Furthermore, the present application also includes the following steps:

[0053] The first image is rasterized to obtain a first raster group; a first raster in the first raster group is locally binary-converted to obtain a first binary code, and a first binary code group is formed; the first binary code group is screened with a preset saw chain binary code as a constraint to obtain a first target binary code group; the first main body slice is calibrated based on a first target raster group obtained by reverse matching the first target binary code group.

[0054] Specifically, the first image is rasterized, that is, the image is converted into a grid composed of pixels. Each pixel in the image corresponds to a grid. The rasterization process divides the image into a regular two-dimensional grid, thereby converting the image information into a discrete format. Through the rasterization process, the first grid group is obtained, each image pixel corresponds to a grid, and each grid represents a small area of ​​the image. The first grid in the first grid group is locally binary converted, and each grid in the raster image is converted into binary data, that is, each grid has only two possible states: 1 (indicating the presence of specific features, usually black or an area of ​​interest) and 0 (indicating the absence of specific features, usually background or blank area). The purpose of binarization is to distinguish different areas in the image (such as tooth shapes or chains) from the background, and to convert the content of the image into black and white states by setting a threshold (such as pixel gray value).

[0055] The first binary code is a binary code composed of raster data after local binary conversion. The binary state (0 or 1) of each raster is encoded into a binary number, usually 1 represents a feature and 0 represents no feature. The first binary code group is a set of multiple first binary codes, which contains the binarization results of all rasters in the image and reflects the feature information of each area in the image.

[0056] The preset saw chain binary code is a predefined binary code based on the design or known standard of the saw chain, which is used as a reference and represents the standard form of the saw chain (such as tooth shape, chain configuration, etc.) and the binary state of these features. The predefined standard saw chain binary code is used as a reference, and the content in the first binary code group is screened accordingly. The binary code group of the actual image (the first binary code group) is matched with the preset saw chain binary code (standard binary code) by comparison, and the part matching the preset saw chain binary code is selected to obtain the first target binary code group, which contains binary code data that conforms to the preset saw chain form. The goal of the screening is to remove the parts that do not meet the saw chain design requirements and only retain the parts that meet the standards. The preset saw chain binary code represents the standard form of the saw chain (such as tooth shape, chain structure), and by comparing the first binary code group of the actual image with the standard binary code, the binary code that meets the standard can be selected. For example, assuming that the preset saw chain binary code is 11110000, and a part in the first binary code group is 11110100, it means that the part does not meet the standard and needs to be excluded.

[0057] The first target binary code group obtained after screening is used for reverse matching. By comparing the first target binary code group with the binary code in the actual image, the geometric shape of the image or model is adjusted, and the first main slice in the image is corrected to better match the preset saw chain binary code, ensuring that the saw chain features in the image are consistent with the geometric shape of the actual saw chain. The goal of reverse matching is to correct the morphological deviation caused by errors or defects in the image so that it conforms to the standard model. The first target binary code group is compared with the raster data in the actual image to adjust the target part (i.e., the first main slice) in the image. The matching error is calculated by shape matching or geometric transformation (such as rigid transformation, affine transformation, etc.), and the part that needs to be adjusted in the image is reversely deduced. The image is translated, rotated or scaled to ensure that the tooth shape or other feature parts in the image are consistent with the preset standard saw chain binary code. Reverse matching is not only the adjustment of shape, but also the correction of spatial position. If the tooth shape in the image deviates from the standard tooth shape in position or angle, reverse matching can correct it by adjusting these parameters.

[0058] After reverse matching, the first main slice in the image (such as the toothed part of the saw chain) will be accurately calibrated. After calibration, the shape, size and position of the first main slice will be closer to the design requirements of the standard saw chain. Through screening and reverse matching, errors and noise in the image are eliminated to ensure the accuracy of saw chain features (such as tooth shape and chain structure). Based on the preset saw chain binary code, only the parts that meet the saw chain standards are retained, reducing the impact of erroneous data, automatically identifying and calibrating the details of the saw chain processing process, and automatically identifying processing deviations.

[0059] Furthermore, step six of this application includes:

[0060] Perform multi-dimensional feature collection on the three-dimensional structure to obtain structural feature parameters; compare the structural feature parameters with predetermined structural feature parameters of the predetermined three-dimensional model to obtain a structural comparison result; determine and analyze the structural comparison result to determine whether the structural abnormality index is within the predetermined abnormality index threshold; if so, activate the structural abnormality prediction model to analyze the structural feature parameters to obtain a first predicted abnormality category; match a first control indicator corresponding to the first predicted abnormality category in a predetermined processing control mapping list stored in the predetermined processing control mechanism; and control parameter information corresponding to the first control indicator in the real-time processing technology information.

[0061] Specifically, multi-dimensional structural feature parameters are extracted from the obtained three-dimensional structural model, including the height, width, pitch, tooth angle, thickness, angle, etc. of the tooth shape. Multi-dimensional feature collection refers to the extraction of various types of feature data from the three-dimensional structure, including geometric dimensions, angles, surface smoothness, complex curves or changes in surfaces, etc., to comprehensively describe the shape and structure of the object. The characteristic parameters of the three-dimensional structure obtained by actual processing are compared with the characteristics of the predetermined three-dimensional model in the design stage to find out the differences between the two, obtain the structural comparison results, and evaluate the accuracy of the processing. Analyze the structural comparison results and calculate the structural anomaly index. The structural anomaly index is a quantitative indicator used to indicate the difference between the actual processing results and the design goals. If the difference between the two is large and the anomaly index is high, it indicates that there are large deviations or quality problems in the processing process.

[0062] It is determined whether the structural abnormality index is within the predetermined abnormality index threshold. If the structural abnormality index is within the predetermined abnormality index threshold, that is, the predetermined abnormality index threshold is reached, the structural abnormality prediction model is activated, and the obtained structural characteristic parameters are further analyzed. The predetermined abnormality index threshold has been explained above. If the predetermined abnormality index threshold is reached, it means that there is an abnormality in the processing process, and a warning is automatically triggered. The structural abnormality prediction model will judge possible abnormal problems based on the existing structural characteristic parameters (such as tooth size, angle, etc.).

[0063] The structural anomaly prediction model refers to the process of manufacturing or processing. By analyzing and predicting various structural anomalies (such as tooth shape error, broken teeth, deformation, etc.) that may occur during the processing, it helps to adjust the processing technology and avoid the production of defective products. Structural anomalies refer to the differences between the geometric shape, size or surface quality of the processed part and the predetermined design model, which may cause the product to fail to meet the quality standards or meet the use requirements, including tooth shape error, broken teeth, deformation, etc. Structural anomaly prediction is to analyze the collected processing data (such as three-dimensional models, real-time parameters of the processing process) to infer the types of anomalies that may occur during the processing and their likelihood.

[0064] By calling historical data, including the processing problems that occurred in the past period of time and the corresponding structural feature parameters, the historical data is preprocessed, including denoising, normalization, missing value filling, etc. Features related to abnormalities are extracted from the processing data, such as geometric features such as the curvature, tooth width, and tooth angle of the tooth shape. Select appropriate models, including machine learning models (such as support vector machines SVM, decision trees, random forests, neural networks, etc.), regression analysis, etc., and use historical processing data for training to train a model that can predict processing abnormalities based on input data. Use support vector machines (SVM) to learn the boundary between normal and abnormal, or use decision trees to build classification rules for processing abnormalities. Train the model by annotating historical data (for example, marking which data is normal and which data is abnormal), compare the predicted results with the actual results, calculate the mean square error, adjust the model parameters, and continuously optimize the model until the accuracy reaches the preset requirements or the number of iterative optimizations reaches the maximum number, stop adjusting the parameters, and use the model at this time as a structural abnormality prediction model.

[0065] The collected structural feature parameters of the three-dimensional structure are input into the structural abnormality prediction model to obtain the first predicted abnormality category, and possible abnormal problems such as tooth shape error, broken teeth, deformation, etc. are judged. The predetermined processing control mechanism is based on historical data, empirical rules or algorithm models, and defines how to adjust processing parameters for different abnormal types. The pre-set processing technology control scheme contains control measures corresponding to various abnormal categories. The first control index corresponding to the first predicted abnormal category is matched in the predetermined processing control mapping list stored in the memory of the predetermined processing control mechanism. The mapping relationship between each abnormal category and the control index is stored in the predetermined processing control mapping list, and the control measures corresponding to different abnormal categories are defined. The first control index is an index related to the first predicted abnormal category, which will directly affect the first predicted abnormal category. For example, the structural feature parameters are compared with the predetermined structural feature parameters, and the length deviation of the tooth is 0.2 mm, the width deviation is 0.1 mm, and the angle deviation is 5 degrees. The calculated structural anomaly index is 0.3. Since the preset anomaly index threshold is 0.2, the structural anomaly prediction model is activated to predict tooth shape anomalies. The cutting speed is adjusted by querying the preset processing control mapping list. The adjusted cutting speed is 100 meters per minute.

[0066] According to the control indicators found in the control mapping list, the real-time processing technology is automatically adjusted, and the processing parameters are adjusted in real time to correct or optimize the processing results during the processing process to ensure product quality and processing efficiency. By automatically identifying and correcting abnormal problems that occur during the processing process, especially when the abnormality reaches a certain level, it will not only detect the abnormality, but also take proactive measures to adjust the processing technology, thereby improving the stability of the processing process, reducing human intervention, improving production efficiency and product quality, and reducing scrap and rework rates.

[0067] Furthermore, the present application also includes the following steps:

[0068] If not, perform multi-dimensional feature collection on the three-dimensional model to obtain model feature parameters; compare the model feature parameters with the predetermined model feature parameters of the predetermined three-dimensional model to obtain a model comparison result; determine and analyze the model comparison result to obtain whether the model abnormality index is within the predetermined abnormality index threshold; if so, activate the model abnormality prediction model to analyze the model feature parameters to obtain a second predicted abnormality category; match the second control indicator corresponding to the second predicted abnormality category in the predetermined processing control mapping list; and regulate the parameter information corresponding to the second control indicator in the real-time processing technology information.

[0069] Specifically, if the structural anomaly index is not within the predetermined anomaly index threshold, that is, it has not reached the predetermined anomaly index threshold, it means that it is still within the controllable range and other inspections or operations are required. Multi-dimensional feature collection is performed on the three-dimensional model to obtain model feature parameters, including surface roughness, dimensional error, surface cracks, etc. The feature parameters of the actual processed three-dimensional model are compared with the predetermined model feature parameters of the predetermined three-dimensional model, and the differences in each feature dimension (such as tooth width, tooth shape, tooth angle) are compared to generate comparison results to detect whether there is a deviation in the processing process. The model comparison results are analyzed to obtain the model anomaly index, which is a quantitative indicator used to evaluate the degree of difference between the actual processed model and the model in the ideal state.

[0070] It is determined whether the model abnormality index is within the predetermined abnormality index threshold. If the predetermined abnormality index threshold is reached, the model abnormality prediction model is activated. The model abnormality prediction model is similar to the structural abnormality prediction model. The difference is that the structural abnormality prediction model is used to analyze a type of processing problem (such as tooth shape, broken teeth, deformation, etc.), and the model abnormality prediction model is used to analyze a type of processing problem (such as fracture, surface roughness, etc.). The construction process is similar and has been introduced in detail in the structural abnormality prediction model, so I will not repeat it here.

[0071] Input the model characteristic parameters into the model abnormality prediction model to obtain the second predicted abnormality category, and predict the second type of abnormality, such as unqualified surface roughness or fracture. Match the second control index corresponding to the second predicted abnormality category in the predetermined processing control mapping list, that is, the control parameter associated with the second predicted abnormality category (such as fracture abnormality, surface roughness abnormality). According to the second control index, directly adjust the relevant parameters in the real-time processing technology to ensure that the processing accuracy is guaranteed. For example, by comparing the model characteristic parameters with the predetermined model characteristic parameters, the surface roughness deviation of the saw chain is 0.5μm, the tooth surface depth deviation is 0.1mm, the model abnormality index is calculated to be 0.3, the predetermined abnormality index threshold is 0.2, and the surface roughness abnormality is discovered through the abnormality prediction model. The process information collected in real time is cutting speed of 25m / min, feed speed of 0.4mm / rev, and tool angle of 25°. The cutting speed is automatically adjusted from 250m / min to 300m / min, feed speed is reduced from 0.4mm / rev to 0.3mm / rev, and tool angle is from 25° to 20° to ensure that the surface smoothness meets the requirements. By timely monitoring the difference between the model and the predetermined model, and accurately adjusting according to the abnormal prediction model, the process parameters are automatically adjusted according to the abnormal situation in the processing process, avoiding problems such as abnormal surface roughness and fracture, improving the processing quality, and reducing waste.

[0072] Furthermore, the present application also includes the following steps:

[0073] Establish an abnormal category set, and extract any abnormal category from the abnormal category set; establish a process indicator set, and perform correlation analysis between the arbitrary abnormal category and the process indicator set to obtain any correlation result; analyze the arbitrary correlation result and screen significantly correlated process indicators to form an arbitrary process indicator set of the arbitrary abnormal category; establish the predetermined processing control mapping list according to the mapping relationship between the arbitrary abnormal category and the arbitrary process indicator set.

[0074] The abnormality category set includes a structural abnormality category set and a model abnormality category set, and the structural abnormality category set includes at least tooth shape abnormality, broken tooth abnormality and deformation abnormality, and the model abnormality category set includes at least fracture abnormality and surface roughness abnormality.

[0075] Specifically, the abnormal category set is a set of possible processing abnormality types, including various problems or defects that may occur during the saw chain processing. The abnormal category set includes the structural abnormal category set and the model abnormal category set. The structural abnormal category set refers to the abnormality types related to the saw chain structure, including tooth shape abnormality (tooth shape does not meet the design requirements due to improper cutting parameters or tool wear), broken tooth abnormality (tooth breakage or falling due to excessive wear or improper operation) and deformation abnormality (whole or partial deformation of the saw chain due to improper processing pressure or temperature, such as twisting or bending), etc.; the model abnormal category set refers to the abnormality types related to the saw chain appearance and surface quality, including fracture abnormality (fracture or crack of the saw chain due to excessive cutting force or material fatigue) and surface roughness abnormality (poor processing surface quality, excessive burrs or rough surface due to improper cutting speed, feed speed or tool state). Extract any abnormal category from the abnormal category set for analysis.

[0076] The process index set is a set of parameters that describe the key factors of the processing process, which directly affects the processing quality of the saw chain, including cutting speed, tool selection, processing temperature, cutting force, feed speed, lubrication conditions, etc. Through statistical methods (such as regression analysis or correlation coefficient), the correlation analysis is performed on any abnormal category and the process index set, and the relationship between each process index and the abnormal category in the processing process is analyzed to find out which process index has a significant relationship with the abnormal category. By calculating the correlation coefficient, the relationship between each process index and a specific abnormal category is found, and any correlation results are obtained, including the correlation coefficient between each process index and any abnormal category.

[0077] Analyze any correlation results and screen the significantly correlated process indicators, that is, select the corresponding process indicators with high correlation to obtain any process indicator set, which includes process indicators that can significantly affect the abnormal category and will directly cause or solve the abnormal problems in processing. For example, if the cutting speed and tool wear have a significant positive correlation with the broken tooth abnormality, they will be included in the process indicator set of the broken tooth abnormality.

[0078] According to the correlation between any abnormality category and the process indicator set, a control mapping list is constructed. The processing control mapping list is a preset rule list that records the mapping relationship between different abnormality categories and the corresponding process parameters. According to these mapping relationships, adjustment measures can be automatically matched when abnormalities are found. Each abnormality category will have a set of process parameters that need to be adjusted. For example, for tooth shape abnormalities, according to the previously screened process indicator set (cutting speed, feed speed, tool, etc.), a set of control rules are defined: if the cutting speed exceeds a certain threshold, the feed speed needs to be reduced; if the tool wear exceeds the preset standard, the tool needs to be replaced. By constructing an abnormality category set and a process indicator set, and performing correlation analysis, the key factors affecting the processing quality can be accurately identified.

[0079] In summary, the saw chain processing method combined with visual inspection provided by the present application has the following technical effects:

[0080] The saw chain processing process is dynamically monitored by intelligent sensing equipment to obtain real-time processing information; the real-time saw chain image set is dynamically acquired by an intelligent image acquisition device arranged on the saw chain, and the real-time saw chain image set includes a first image, wherein the first image refers to the saw chain image acquired by the first acquisition device in the intelligent image acquisition device; a predetermined modeling strategy is introduced to analyze the first image, and a three-dimensional model of the saw chain is constructed; the comparison result between the three-dimensional model and the predetermined three-dimensional model of the saw chain is analyzed to obtain the real-time processing abnormality index of the saw chain; when the real-time processing abnormality index reaches the predetermined abnormality index threshold, a processing abnormality warning signal is issued; based on the processing abnormality warning signal, a predetermined processing control mechanism is called, and the real-time processing information is dynamically processed and controlled. That is to say, through the dynamic monitoring of intelligent sensing equipment and image acquisition equipment, various process data in the saw chain processing process are obtained in real time, a three-dimensional model of the saw chain is constructed, the actual processing results are compared with the predetermined three-dimensional model, any morphological or dimensional deviations in the processing process are accurately identified, the processing parameters are dynamically adjusted, the processing technology is optimized, and the processing accuracy and product quality of the saw chain are improved.

[0081] Embodiment 2, based on the same inventive concept as the saw chain processing method combined with visual detection in the aforementioned embodiment 1, the present application also provides a saw chain processing system combined with visual detection, please refer to the attached Figure 2 , the saw chain processing system combined with visual inspection comprises:

[0082] A real-time information acquisition module 11, the real-time information acquisition module 11 is used to dynamically monitor the saw chain processing process through an intelligent sensor device to obtain real-time processing technology information; an image acquisition module 12, the image acquisition module 12 is used to dynamically acquire a real-time saw chain image set through an intelligent image acquisition device arranged on the saw chain, the real-time saw chain image set includes a first image, wherein the first image refers to a saw chain image acquired by a first collector in the intelligent image acquisition device; a model construction module 13, the model construction module 13 is used to introduce a predetermined modeling strategy to analyze the first image, and construct a three-dimensional model of the saw chain; an abnormality comparison module 14, the abnormality comparison module 14 is used to analyze the comparison result of the three-dimensional model with the predetermined three-dimensional model of the saw chain to obtain the real-time processing abnormality index of the saw chain; an abnormality warning module 15, the abnormality warning module 15 is used to issue a processing abnormality warning signal when the real-time processing abnormality index reaches a predetermined abnormality index threshold; a processing control module 16, the processing control module 16 is used to call a predetermined processing control mechanism based on the processing abnormality warning signal, and dynamically process and control the real-time processing technology information.

[0083] Furthermore, the model building module 13 in the saw chain processing system combined with visual detection is also used for:

[0084] Extracting the subject-background slicing strategy in the predetermined modeling strategy; slicing the first image according to the subject-background slicing strategy to obtain a first slicing result, and extracting a first subject slice in the first slicing result; acquiring a first acquisition control parameter of the first collector; performing a structural analysis on the first subject slice in combination with the first acquisition control parameter to obtain the three-dimensional structure of the saw chain; aligning the first point cloud data of the first image to the three-dimensional structure to obtain the three-dimensional model.

[0085] Furthermore, the model building module 13 in the saw chain processing system combined with visual detection is also used for:

[0086] The first image is rasterized to obtain a first raster group; a first raster in the first raster group is locally binary-converted to obtain a first binary code, and a first binary code group is formed; the first binary code group is screened with a preset saw chain binary code as a constraint to obtain a first target binary code group; the first main body slice is calibrated based on a first target raster group obtained by reverse matching the first target binary code group.

[0087] Furthermore, the processing control module 16 in the saw chain processing system combined with visual detection is also used for:

[0088] Perform multi-dimensional feature collection on the three-dimensional structure to obtain structural feature parameters; compare the structural feature parameters with predetermined structural feature parameters of the predetermined three-dimensional model to obtain a structural comparison result; determine and analyze the structural comparison result to determine whether the structural abnormality index is within the predetermined abnormality index threshold; if so, activate the structural abnormality prediction model to analyze the structural feature parameters to obtain a first predicted abnormality category; match a first control indicator corresponding to the first predicted abnormality category in a predetermined processing control mapping list stored in the predetermined processing control mechanism; and control parameter information corresponding to the first control indicator in the real-time processing technology information.

[0089] Furthermore, the processing control module 16 in the saw chain processing system combined with visual detection is also used for:

[0090] If not, perform multi-dimensional feature collection on the three-dimensional model to obtain model feature parameters; compare the model feature parameters with the predetermined model feature parameters of the predetermined three-dimensional model to obtain a model comparison result; determine and analyze the model comparison result to obtain whether the model abnormality index is within the predetermined abnormality index threshold; if so, activate the model abnormality prediction model to analyze the model feature parameters to obtain a second predicted abnormality category; match the second control indicator corresponding to the second predicted abnormality category in the predetermined processing control mapping list; and regulate the parameter information corresponding to the second control indicator in the real-time processing technology information.

[0091] Furthermore, the processing control module 16 in the saw chain processing system combined with visual detection is also used for:

[0092] Establish an abnormal category set, and extract any abnormal category from the abnormal category set; establish a process indicator set, and perform correlation analysis between the arbitrary abnormal category and the process indicator set to obtain any correlation result; analyze the arbitrary correlation result and screen significantly correlated process indicators to form an arbitrary process indicator set of the arbitrary abnormal category; establish the predetermined processing control mapping list according to the mapping relationship between the arbitrary abnormal category and the arbitrary process indicator set.

[0093] Furthermore, the processing control module 16 in the saw chain processing system combined with visual detection is also used for:

[0094] The abnormality category set includes a structural abnormality category set and a model abnormality category set, and the structural abnormality category set includes at least tooth shape abnormality, broken tooth abnormality and deformation abnormality, and the model abnormality category set includes at least fracture abnormality and surface roughness abnormality.

[0095] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1 The saw chain processing method combined with visual detection and the specific examples in the first embodiment are also applicable to the saw chain processing system combined with visual detection in this embodiment. Through the above detailed description of the saw chain processing method combined with visual detection, those skilled in the art can clearly understand the saw chain processing system combined with visual detection in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0096] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be 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 application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0097] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A saw chain processing method combined with visual inspection, characterized in that: include: The saw chain processing is dynamically monitored through intelligent sensing equipment to obtain real-time processing information; A real-time saw chain image set is obtained by dynamically acquiring through an intelligent image acquisition device arranged on the saw chain, wherein the real-time saw chain image set includes a first image, wherein the first image refers to a saw chain image acquired by a first acquisition device in the intelligent image acquisition device; Introducing a predetermined modeling strategy to analyze the first image, and constructing a three-dimensional model of the saw chain; analyzing a comparison result between the three-dimensional model and a predetermined three-dimensional model of the saw chain to obtain a real-time processing abnormality index of the saw chain; When the real-time processing abnormality index reaches a predetermined abnormality index threshold, a processing abnormality warning signal is issued; Based on the abnormal processing warning signal, a predetermined processing control mechanism is retrieved, and dynamic processing control is performed on the real-time processing technology information; Introducing a predetermined modeling strategy to analyze the first image and constructing a three-dimensional model of the saw chain, including: Extracting a subject-background slicing strategy from the predetermined modeling strategy; Slicing the first image according to the subject-background slicing strategy to obtain a first slicing result, and extracting a first subject slice in the first slicing result; Acquire a first acquisition control parameter of the first collector; performing structural analysis on the first main body slice in combination with the first acquisition control parameter to obtain a three-dimensional structure of the saw chain; The first point cloud data of the first image is registered to the three-dimensional structure to obtain the three-dimensional model.

2. The saw chain processing method combined with visual inspection according to claim 1, characterized in that: After slicing the first image according to the subject-background slicing strategy to obtain a first slicing result and extracting a first subject slice from the first slicing result, the method further includes: Performing rasterization processing on the first image to obtain a first raster group; Performing local binary conversion on the first grid in the first grid group to obtain a first binary code and forming a first binary code group; The first binary code group is screened based on a preset saw chain binary code as a constraint to obtain a first target binary code group; The first body slice is calibrated based on a first target grid group obtained by reverse matching the first target binary code group.

3. The saw chain processing method combined with visual inspection according to claim 1, characterized in that: Based on the abnormal processing warning signal, a predetermined processing control mechanism is retrieved, and dynamic processing control is performed on the real-time processing technology information, including: Collecting multi-dimensional features of the three-dimensional structure to obtain structural feature parameters; Comparing the structural feature parameters with predetermined structural feature parameters of the predetermined three-dimensional model to obtain a structural comparison result; Determine and analyze whether the structural abnormality index obtained by the structural comparison result is within the predetermined abnormality index threshold; If yes, activate the structural abnormality prediction model to analyze the structural characteristic parameters to obtain a first predicted abnormality category; Matching a first control indicator corresponding to the first predicted abnormality category in a predetermined processing control mapping list stored in the predetermined processing control mechanism; The parameter information corresponding to the first control index in the real-time processing technology information is controlled.

4. The saw chain processing method combined with visual inspection according to claim 3, characterized in that: After judging and analyzing the structural comparison result to obtain a structural abnormality index within the predetermined abnormality index threshold, the method further includes: If not, performing multi-dimensional feature collection on the three-dimensional model to obtain model feature parameters; Comparing the model characteristic parameters with predetermined model characteristic parameters of the predetermined three-dimensional model to obtain a model comparison result; Determine and analyze the model comparison result to determine whether the model abnormality index is within the predetermined abnormality index threshold; If yes, activate the model abnormality prediction model to analyze the model feature parameters to obtain a second predicted abnormality category; matching a second control indicator corresponding to the second predicted abnormality category in the predetermined processing control mapping list; The parameter information corresponding to the second control index in the real-time processing technology information is controlled.

5. The saw chain processing method combined with visual inspection according to claim 3, characterized in that: Matching the first control indicator corresponding to the first predicted abnormality category in the predetermined processing control mapping list stored in the predetermined processing control mechanism includes: Forming an abnormal category set, and extracting any abnormal category in the abnormal category set; Establishing a process indicator set, and performing a correlation analysis between the arbitrary abnormality category and the process indicator set to obtain an arbitrary correlation result; Analyze the arbitrary correlation results and select significantly correlated process indicators to form an arbitrary process indicator set of the arbitrary abnormal category; The predetermined processing control mapping list is established according to the mapping relationship between the arbitrary abnormality category and the arbitrary process indicator set.

6. The saw chain processing method combined with visual inspection according to claim 5, characterized in that: The abnormality category set includes a structural abnormality category set and a model abnormality category set, and the structural abnormality category set includes at least tooth shape abnormality, broken tooth abnormality and deformation abnormality, and the model abnormality category set includes at least fracture abnormality and surface roughness abnormality.

7. A saw chain processing system combined with visual inspection, characterized in that: The steps for implementing the saw chain processing method combined with visual inspection according to any one of claims 1 to 6, wherein the saw chain processing system combined with visual inspection comprises: A real-time information collection module, which is used to dynamically monitor the saw chain processing process through intelligent sensing equipment to obtain real-time processing technology information; An image acquisition module, the image acquisition module is used to dynamically acquire a real-time saw chain image set through an intelligent image acquisition device arranged on the saw chain, the real-time saw chain image set includes a first image, wherein the first image refers to a saw chain image acquired by a first collector in the intelligent image acquisition device; A model building module, the model building module is used to introduce a predetermined modeling strategy to analyze the first image and construct a three-dimensional model of the saw chain; an abnormality comparison module, the abnormality comparison module is used to analyze the comparison result between the three-dimensional model and the predetermined three-dimensional model of the saw chain to obtain a real-time processing abnormality index of the saw chain; An abnormality warning module, wherein the abnormality warning module is used to issue a processing abnormality warning signal when the real-time processing abnormality index reaches a predetermined abnormality index threshold; A processing control module is used to call a predetermined processing control mechanism based on the processing abnormality warning signal, and dynamically control the real-time processing technology information.

Images