Laser surface treatment method, device and equipment based on visual control

Abstract

The invention provides a laser surface treatment method, device and equipment based on visual control, and the method comprises the steps: carrying out the energy waveform capture processing of pulse laser outputted by laser marking equipment, and carrying out the multi-channel visual detection of a surface reflection region of a to-be-marked part based on a pulse energy reference, and obtaining the surface feature information; planning an initial scanning path according to the surface feature information, and synchronously correcting a trigger time sequence of the pulse laser to obtain a target marking path; and in combination with the target marking path and the pulse energy reference, dynamic observation and local heat dispersion treatment are conducted on the marking area, and the treated part surface is obtained. The method is suitable for performing fine marking on a high-reflectivity or complex geometric surface, and can perform timely adjustment according to local temperature rise or optical characteristic change in the marking process to reduce depth unevenness and thermal damage.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment control, and in particular to a laser surface processing method, device and equipment based on vision control. Background Art

[0002] In the field of laser processing, laser marking technology is widely used in the marking and traceability management of high-precision components or products. Existing laser marking devices are usually operated by fixed power, preset scanning paths, etc., and it is difficult to make real-time adjustments to the reflective and thermal properties of different material surfaces. Especially when the parts to be marked have high reflectivity or complex curved geometry, traditional processes often lack flexibility in laser energy input and scanning trajectory control, which can easily cause uneven marking depth or excessive heat accumulation. For industries with high requirements for surface quality and marking durability, such as medical devices, aviation parts, etc., such problems are more prominent.

[0003] At the same time, as the requirements for processing quality and efficiency continue to increase, various vision-based monitoring methods have begun to be introduced into the laser marking process to identify the surface morphology and temperature distribution of the parts to be marked. However, passive visual observation alone is not enough to provide accurate feedback on the energy fluctuations of the pulsed laser; in addition, the transient scattering and mirror reflection phenomena generated by the highly reflective surface can easily interfere with the recognition results of conventional sensors, resulting in a lack of targeted scheduling of the laser output energy and the scanning galvanometer trajectory. In addition, many of the publicly available vision-based laser marking or processing solutions mostly focus only on the initial monitoring of the marking process, and lack closed-loop control of subsequent dynamic observation and local heat dispersion in conjunction with the pulse energy benchmark. Once local heat accumulation or local melting occurs at the end of the marking process, it is often difficult to take effective strategies to compensate in a timely manner. Summary of the invention

[0004] The main purpose of the present invention is to solve the technical problems that the existing laser marking process is difficult to adjust the energy and trajectory in real time for highly reflective or complex curved surfaces, and cannot effectively deal with local heat accumulation or ablation at the end of marking, resulting in uneven marking depth and thermal damage; A first aspect of the present invention provides a laser surface treatment method based on vision control, the laser surface treatment method based on vision control comprising: Performing energy waveform capture processing on the pulse laser output by the laser marking device to obtain a pulse energy reference, and performing multi-channel visual inspection processing on the surface reflection area of ​​the to-be-marked component according to the pulse energy reference to obtain surface feature information of the to-be-marked component; Planning and processing the trajectory of the laser scanning galvanometer of the laser marking device according to the pulse energy reference and the surface feature information of the component to be marked to obtain an initial scanning path; Performing synchronous correction processing on the triggering timing of the pulsed laser output by the laser marking device according to the initial scanning path to obtain a target marking path; The marking area of ​​the component to be marked is dynamically observed and locally heat dispersed according to the target marking path and the pulse energy reference to obtain a component surface after the processing of the component to be marked.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the pulsed laser output by the laser marking device is subjected to energy waveform capture processing to obtain a pulse energy reference, and a surface reflection area of ​​the component to be marked is subjected to multi-channel visual inspection processing according to the pulse energy reference to obtain surface feature information of the component to be marked, including: Performing high-speed photoelectric sampling and acquisition processing on the pulse signal output by the laser to obtain a pulse sampling sequence; Performing feature extraction processing on the peak segment and the attenuation segment of the pulse signal according to the pulse sampling sequence to obtain a pulse energy reference; Performing multi-channel visual recognition on the surface reflection area of ​​the component to be marked to obtain a multi-channel image set; Optical feature comparison and pixel partition clustering processing are performed on the multi-channel image set according to the pulse energy reference to obtain surface feature information of the component to be marked.

[0006] Optionally, in a second implementation of the first aspect of the present invention, performing optical feature comparison and pixel partition clustering processing on the multi-channel image set according to the pulse energy reference to obtain surface feature information of the component to be marked includes: Performing optical spectrum extraction processing on each pixel point in the multi-channel image set to obtain an optical feature matrix; Performing threshold partitioning processing on corresponding pixel values ​​in the optical feature matrix according to the pulse energy reference to obtain a preliminary partitioning result; Performing grouping and clustering processing on the high-reflection pixel group and the general-reflection pixel group of the preliminary partitioning result to obtain a pixel clustering index table; The concave-convex boundary area is processed for contour positioning according to the pixel clustering index table to obtain surface feature information of the component to be marked.

[0007] Optionally, in a third implementation of the first aspect of the present invention, the trajectory of the laser scanning galvanometer of the laser marking device is planned and processed according to the pulse energy reference and the surface feature information of the component to be marked to obtain an initial scanning path, comprising: Performing pulse adaptability discrimination processing on the pulse sampling sequence and the surface feature information to obtain a pulse adaptability index between the pulse sampling sequence and the surface feature information; According to the pulse adaptability index, pulse energy priorities are divided for different regions of the component to be marked to obtain a region priority table; Arrange the path nodes of the trajectory of the laser scanning galvanometer of the laser marking device according to the area priority table to obtain a node arrangement sequence; The pulse energy amplitude and pulse width allocation parameters of the pulse laser are scheduled according to the node arrangement sequence to obtain the initial scanning path.

[0008] Optionally, in a fourth implementation manner of the first aspect of the present invention, the pulse sampling sequence and the surface feature information are subjected to pulse adaptability discrimination processing to obtain a pulse adaptability index between the pulse sampling sequence and the surface feature information, comprising: Performing energy mapping processing on the pulse energy peak value in the pulse sampling sequence and the multi-channel reflection data in the surface feature information to obtain an energy-feature mapping table; According to the energy-feature mapping table, each pulse node in each pulse sampling sequence is subjected to a difference comparison process with different reflection regions to obtain a pulse-region difference statistical result; Performing segmented correction processing on the pulse nodes with excessive deviations in the pulse-region difference statistical results to obtain a corrected coordination matrix; The pulse energy distribution and surface feature information of the pulse laser are comprehensively evaluated and processed according to the modified matching degree matrix to obtain the pulse matching degree index.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of performing synchronous correction processing on the trigger timing of the pulsed laser output by the laser marking device according to the initial scanning path to obtain the target marking path includes: Comparing and evaluating each coordinate node in the initial scanning path with the ideal pulse triggering time in the pulse adaptability index to obtain a trigger timing difference table; Performing phase calibration processing on the trigger signal of the pulse laser and the coordinate signal of the laser scanning galvanometer according to the trigger timing difference table to obtain a synchronous trigger parameter group; According to the synchronous trigger parameter group, the pulse energy amplitude of each pulse laser and the galvanometer movement speed of the laser scanning galvanometer are processed in a segmented matching operation to obtain a trigger linkage instruction set; The trigger nodes of each scanning section of the initial scanning path are time-marked according to the trigger linkage instruction set to obtain a target marking path.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, dynamically observing and locally dispersing the marking area of ​​the component to be marked according to the target marking path and the pulse energy reference to obtain the component surface after the processing of the component to be marked includes: Performing multi-channel continuous observation processing on the marking results completed by the target marking path to obtain a thermal radiation sequence after marking; Comparing the pulse energy distribution of the pulsed laser with the local temperature curve according to the thermal radiation sequence after marking, and obtaining a high-temperature concentration section; According to the pulse energy standard, the high temperature concentration section is subjected to short-time pulse or airflow heat dissipation treatment to obtain a heat dispersion effect set; According to the heat dispersion effect set, low-energy supplementary engraving is performed on the incomplete positions of the marking lines in the marking result to obtain the processed component surface of the component to be marked.

[0011] A second aspect of the present invention provides a laser surface treatment device based on vision control, the laser surface treatment device based on vision control comprising: A capture detection module is used to perform energy waveform capture processing on the pulse laser output by the laser marking device to obtain a pulse energy reference, and perform multi-channel visual detection processing on the surface reflection area of ​​the to-be-marked component according to the pulse energy reference to obtain surface feature information of the to-be-marked component; A trajectory planning module, used to plan the trajectory of the laser scanning galvanometer of the laser marking device according to the pulse energy reference and the surface feature information of the part to be marked, so as to obtain an initial scanning path; A synchronous correction module, used for performing synchronous correction processing on the triggering timing of the pulse laser output by the laser marking device according to the initial scanning path to obtain a target marking path; The observation and heat dissipation module is used to dynamically observe and perform local heat dispersion processing on the marking area of ​​the component to be marked according to the target marking path and the pulse energy reference, so as to obtain the component surface after the processing of the component to be marked.

[0012] The third aspect of the present invention provides a laser surface processing device based on vision control, comprising: a memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected through lines; the at least one processor calls the instructions in the memory so that the laser surface processing equipment based on vision control performs the steps of the above-mentioned laser surface processing method based on vision control.

[0013] The above-mentioned laser surface treatment method, device and equipment based on visual control, through the energy waveform capture processing of the pulsed laser output by the laser marking equipment, and based on the pulse energy reference, multi-channel visual detection of the surface reflection area of ​​the part to be marked is performed to obtain surface feature information; the initial scanning path is planned according to the surface feature information, and the trigger timing of the pulsed laser is synchronously corrected to obtain the target marking path; in combination with the target marking path and the pulse energy reference, the marking area is dynamically observed and locally heat dispersed to obtain the processed part surface. This method is suitable for fine marking on high reflectivity or complex geometric surfaces, and can make timely adjustments to local temperature rise or optical feature changes during the marking process to reduce depth unevenness and thermal damage.

[0014] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0015] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Schematic diagram of a first embodiment of a laser surface treatment method based on vision control in an embodiment of the present invention; Figure 2 Schematic diagram of an embodiment of a laser surface treatment device based on vision control in an embodiment of the present invention; Figure 3 It is a schematic diagram of an embodiment of a laser surface processing device based on vision control in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0019] To facilitate understanding of this embodiment, a laser surface treatment method based on visual control disclosed in an embodiment of the present invention is first described in detail. Figure 1 As shown, the method comprises the following steps: 101. Perform energy waveform capture processing on the pulse laser output by the laser marking device to obtain a pulse energy reference, and perform multi-channel visual inspection processing on the surface reflection area of ​​the to-be-marked component according to the pulse energy reference to obtain surface feature information of the to-be-marked component; In one embodiment of the present invention, the pulsed laser output by the laser marking device is subjected to energy waveform capture processing to obtain a pulse energy reference, and a surface reflection area of ​​the component to be marked is subjected to multi-channel visual inspection processing according to the pulse energy reference to obtain surface feature information of the component to be marked, including: high-speed photoelectric sampling and acquisition processing of the pulse signal output by the laser to obtain a pulse sampling sequence; feature extraction processing of the peak segment and the attenuation segment of the pulse signal according to the pulse sampling sequence to obtain a pulse energy reference; multi-channel visual recognition of the surface reflection area of ​​the component to be marked to obtain a multi-channel image set; optical feature comparison and pixel partitioning clustering processing of the multi-channel image set according to the pulse energy reference to obtain surface feature information of the component to be marked.

[0020] Specifically, in the process of high-speed photoelectric sampling and acquisition processing of the pulse signal output by the laser to obtain the pulse sampling sequence, it is necessary to set a broadband response detection device between the laser output end and the photoelectric acquisition circuit to capture the energy changes of the pulse signal in a very short time. The detection device usually includes a high-sensitivity photodiode or an avalanche photodiode, and cooperates with a high-speed amplification circuit to convert light energy into an electrical signal, and then perform analog-to-digital conversion at a high sampling frequency and record it in a high-precision data cache. In order to accurately characterize the energy release form of different pulses, the sampling frequency should be at least one order of magnitude higher than the main frequency of the laser pulse. If the main frequency of the laser is , sampling frequency satisfy At the same time, in order to eliminate the interference of ambient stray light on the pulse signal, a narrow-band filter or aperture can be configured at the front end of the detection device to concentrate the collected signal near the specific emission wavelength of the laser. This sampling link obtains multiple pulse waveforms in a time series, so that the subsequent data analysis stage can track the energy changes of each pulse, and provide them to the feature extraction algorithm in the form of a sequence. According to the pulse sampling sequence, the peak segment and the attenuation segment of the pulse signal are subjected to feature extraction processing. In the process of obtaining the pulse energy benchmark, the signal processing module will first filter and denoise the sampling sequence, and lock the boundaries of each pulse in the time domain through a short-time window function. Next, by extracting the amplitude value and time width at the rising edge, platform area and attenuation segment of each pulse, a pulse feature vector set is formed and recorded as In order to quantitatively evaluate the energy of each pulse, the integral of the pulse in one cycle can be recorded as d ,in Indicates the pulse waveform at time The amplitude distribution on the surface. After the feature extraction is completed, the feature vectors of all pulses are aggregated to form a pulse energy benchmark for subsequent algorithms to match different surface reflection features. This process takes into account the pulse peak intensity and the time domain information of the entire pulse release process, and can reflect the transient energy difference of the laser output with a higher resolution; In the process of multi-channel visual recognition of the surface reflection area of ​​the component to be marked and obtaining a multi-channel image set, the industrial camera and the optical sensor will synchronously collect the surface image of the component to be marked from different spectral bands. Usually, a visible light channel and an infrared channel can be set to identify reflection features and thermal radiation features. In order to improve the clarity of the image, a polarizing filter or a tunable attenuator can be used for the specular reflection phenomenon on the surface of different materials to avoid image distortion caused by oversaturation or bright spots. The core of multi-channel visual recognition is to synchronize the images in each band and mark each frame of data with corresponding shooting parameters, such as exposure time and exposure time, so that the subsequent processing algorithm can accurately compare the brightness and texture information of the same surface area in different spectral channels. In order to obtain a complete reflection distribution on a complex surface, light sources can be arranged around the parts to be marked or a ring lighting design can be used to ensure that the high curvature area and the occluded area can also obtain effective imaging brightness. Finally, through multi-channel synchronous acquisition, a multi-channel image set is formed, in which each pixel point can be used with multi-dimensional attributes such as visible light intensity and infrared intensity. According to the pulse energy benchmark, the multi-channel image set is optically compared and pixel partitioned clustered. In the process of obtaining the surface feature information of the parts to be marked, the algorithm will first lock the area associated with the pulse energy benchmark in the multi-channel image. For example, when the pulse energy is high, it may cause the local area to increase thermal radiation, and the infrared channel will show brighter pixel values. By associating the pixel value of the area with the energy release peak or attenuation information in the pulse energy benchmark, it can be determined what kind of reflection or temperature distribution appears on the surface of the component under what kind of energy input. Next, based on the results of the optical feature comparison, the algorithm performs pixel partitioning and clustering on the areas with large brightness changes and prominent textures in the image, separates the high-reflection area, low-reflection area, and intermediate transition area, and marks their thermal response characteristics and corresponding pulse energy ranges at the cluster center. Clustering can use algorithms such as K-means or Gaussian mixture models, and according to the distance metric formula After iterative convergence, pixels with similar optical response patterns are classified into the same category. The partition results contain information such as surface normal direction, mirror reflection intensity, and heat accumulation, which can help the subsequent trajectory planning and synchronization correction links to accurately allocate laser power.

[0021] Furthermore, the optical feature comparison and pixel partitioning and clustering processing are performed on the multi-channel image set according to the pulse energy reference to obtain the surface feature information of the component to be marked, including: performing optical spectrum extraction processing on each pixel point in the multi-channel image set to obtain an optical feature matrix; performing threshold division processing on the corresponding pixel values ​​in the optical feature matrix according to the pulse energy reference to obtain a preliminary partitioning result; grouping and clustering the high-reflection pixel group and the general-reflection pixel group of the preliminary partitioning result to obtain a pixel clustering index table; performing contour positioning processing on the concave and convex boundary area according to the pixel clustering index table to obtain the surface feature information of the component to be marked.

[0022] Specifically, when performing optical spectrum extraction processing on each pixel point in the multi-channel image set to obtain an optical feature matrix, it is necessary to first align the multi-channel images at the same spatial coordinates, and establish a set of independent pixel value mappings for each channel according to the response characteristics of different spectral bands. In order to more comprehensively describe the brightness and texture differences of highly reflective surfaces in the visible light and infrared regions, a matrixed pixel fusion algorithm can be used at the camera output end to merge the grayscale or color data of each channel. If tunable filters or polarization processing are used in certain channels, the spectral response function of each pixel point must be normalized when reading out the image to make the pixel values ​​between the visible light channel and the infrared channel comparable. In this way, the pixel coordinates can be calculated. Multidimensional attributes are obtained , and then all Merge them into a two-dimensional plane with the same resolution to form a multi-channel stack data. Then, a two-dimensional or three-dimensional discrete Fourier transform can be performed on the pixel vector in the stack, which is denoted as or , in order to identify certain bands with more significant periodic or texture features in the frequency domain. Finally, the amplitude, phase and other information obtained after the inverse transformation can be fused with the original pixel value and output into the optical feature matrix, recorded as ,in Contains the intensity, texture, and amplitude and phase components of each channel that may be obtained through frequency domain analysis, which are used for subsequent comparative analysis and clustering. According to the pulse energy benchmark, the corresponding pixel values ​​in the optical feature matrix are thresholded and processed. When the preliminary partitioning result is obtained, the value of each pixel in the optical feature matrix needs to be correlated with the energy range or pulse peak recorded in the pulse energy benchmark. The peak amplitude and rising edge duration contained in the pulse energy benchmark provide the conditions for distinguishing different pulse energy levels. If the brightness of a pixel in the visible light channel is too high but corresponds to a low-energy pulse, it indicates that the surface reflectivity is high. On the contrary, if it shows an extremely high response in the infrared channel but corresponds to a high-energy pulse, it may indicate that the area has absorbed more energy at the thermal coupling level. Specifically, the intensity vector of each pixel in the optical feature matrix can be With pulse energy Perform a distance or similarity metric such as dist , where W is the weight matrix used to map the pulse energy components to the optical feature space, and is the adjustment coefficient. If the distance is less than a certain threshold, it is determined that the partition to which the pixel belongs meets the corresponding pulse energy segment. By traversing all pixels and matching them with multiple energy segments, the entire multi-channel image can be preliminarily divided into regions and a label matrix can be output, which is recorded as , where the element value indicates that a pixel belongs to a high-energy zone, a low-energy zone or other predefined reflection level zone. This is the preliminary partitioning result. The high-reflection pixel group and the general-reflection pixel group of the preliminary partitioning result are grouped and clustered. When obtaining the pixel clustering index table, it is necessary to further finely distinguish different reflection intensities and different thermal response modes on the basis of the preliminary division. Since high-reflection pixel groups are usually concentrated on concave and convex surfaces or polished areas, and general-reflection pixel groups are distributed in relatively flat areas or areas with higher material absorption rates, clustering algorithms such as K-means or DBSCAN can be used to perform secondary subdivision on the labeled pixels. If K-means is used, different reflection levels are first set as the initial clustering centers in the preliminary partitioning results, and the Euclidean distance between each cluster element and each center is iteratively calculated. , and then merge the pixel into the nearest cluster center. After the iteration converges, each cluster will be assigned a cluster ID or index value to form a pixel cluster index table. The clustering category to which the pixel belongs is recorded in the form of, and it can also indirectly reflect the reflection and absorption differences of the surface material under different energy inputs. By distinguishing between high-reflection pixel groups and general-reflection pixel groups, the system can distinguish the areas on the surface that are most prone to specular reflection or heat accumulation from other areas, so as to adopt differentiated scanning paths or pulse energy configurations in subsequent steps. When the concave-convex boundary area is contour-positioned according to the pixel clustering index table and the surface feature information of the part to be marked is obtained, it is necessary to identify significant deformation edges or concave-convex edges in the clustering results and mark them accurately. In order to better capture the contours of curved surface transitions or significant height changes, the depth estimation method can be combined. If the multi-channel imaging system has the function of light structure projection or binocular parallax measurement, the clustering information and height information are integrated in the spatial domain to extract the concave-convex interface. The specific process can be achieved by scanning the boundary pixels of each clustering category. If there is a significant depth difference or optical response difference in the adjacent area, the boundary point is defined as a concave-convex edge point and added to the contour marking list. The pixel spatial gradient can also be combined in the contour marking stage A series of contour segments are generated together with the cluster label discontinuity area, and a multi-segment polyline is used for approximate fitting for subsequent analysis or laser path planning. These precisely located concave and convex boundaries are stored in the surface feature information as a data structure, which may include the coordinate sequence of each boundary, pixel clustering category, and the correlation value with the pulse energy mapping, so that differentiated energy management and dynamic heat dispersion can be performed in the subsequent processing link.

[0023] 102. Planning and processing the trajectory of the laser scanning galvanometer of the laser marking device according to the pulse energy reference and the surface feature information of the component to be marked, and obtaining an initial scanning path; In one embodiment of the present invention, the trajectory of the laser scanning galvanometer of the laser marking device is planned and processed according to the pulse energy reference and the surface feature information of the component to be marked to obtain the initial scanning path, including: performing pulse adaptability discrimination processing on the pulse sampling sequence and the surface feature information to obtain a pulse adaptability index between the pulse sampling sequence and the surface feature information; performing pulse energy priority division processing on different areas of the component to be marked according to the pulse adaptability index to obtain a regional priority table; arranging the path nodes of the trajectory of the laser scanning galvanometer of the laser marking device according to the regional priority table to obtain a node arrangement sequence; and scheduling the pulse energy amplitude and pulse width allocation parameters of the pulse laser according to the node arrangement sequence to obtain the initial scanning path.

[0024] Specifically, in the process of performing pulse adaptability discrimination processing on the pulse sampling sequence and the surface feature information to obtain the pulse adaptability index between the pulse sampling sequence and the surface feature information, it is necessary to uniformly manage the various parameters of the laser pulse and the surface feature in the data structure dimension, so as to establish a mapping relationship between the pulse and the material property. First, the key indicators such as the peak value, pulse width, power integral, etc. of each pulse in the pulse sampling sequence can be standardized to eliminate the amplitude deviation caused by the interference such as the sensitivity of the test equipment or the background light, and mark the triggering time of each pulse on the timing coordinate. Then, the optical properties or geometric properties of each region in the surface feature information are discretized, and the different absorption characteristics of the high reflection area, the concave area and the general plane area are distinguished by numerical labels, and correspond to several energy levels in the pulse sequence. On this basis, the coupling degree of each pulse and each regional label is scored by a multidimensional matching discriminant function. If the power peak of a pulse and the reflection coefficient of the corresponding area are too different in value, it is judged that its adaptability is low. If the two are close, the adaptability increases. The entire discrimination process performs a cyclic operation between all pulses and all regions, and summarizes the results in a matrix form, with the numerical value as a measure of the fit between the pulse and the region in the energy dimension. When some pulses show a low fit in both high-reflection areas and low-reflection areas, the system will make special marks or adjustments to these pulses in subsequent steps. After the operation is completed, the maximum matching values ​​between each pulse and each region are integrated into a vector to form a pulse fitness index. This index will be frequently called during subsequent energy priority division to assist in determining which area different pulses are suitable for placement in, and whether each partition lacks sufficient pulse energy coverage.

[0025] According to the pulse fitness index, the pulse energy priorities of different areas of the component to be marked are divided and processed. In the process of obtaining the regional priority table, the importance of each partition needs to be sorted in combination with the aforementioned index value and surface feature information. If a certain area continues to obtain a high score in the pulse matching scoring stage, it means that the coupling between the area and the pulse energy is stable. It can be marked as a partition with a higher energy input requirement in the priority table, or it can be reversely marked as the partition that is most susceptible to ablation. In order to enable the marking process to achieve more detailed energy distribution in highly reflective or highly heat-sensitive parts, these partitions can be assigned a higher level in the priority table, so that the trajectory planning link can include them in the early marking queue. If the matching index of a partition has a large discreteness or low stability, the system will classify it as a special concern partition to make corresponding redundant settings when arranging subsequent path nodes. By comparing the pulse fitness index with the regional geometric features, it can also be determined which areas have higher material absorption rates or more difficult to reach curved corners, so as to add necessary pulse power limits or pulse number limits to the entries in the priority table. The final generated regional priority table is not just a simple sequential enumeration, but also includes pulse energy levels, partition coordinate ranges, and general scanning strategy prompts, providing a more fine-grained planning reference for the next step of node arrangement.

[0026] According to the regional priority table, the path nodes of the trajectory of the laser scanning galvanometer of the laser marking equipment are arranged and processed. In the process of obtaining the node arrangement sequence, it is necessary to plan the movement direction of the galvanometer in both spatial coordinates and time coordinates. First, the three-dimensional model or partitioned information of the surface of the component to be marked should be gridded so that each grid node corresponds to a galvanometer target position. The system traverses the regional priority table and lists the nodes of the high-priority partition at the front of the path queue to solve the most sensitive or critical marking parts as soon as possible. If a certain area is over-concentrated and may cause heat accumulation, a jumping sorting scheme is inserted when arranging the nodes, so that the galvanometer scans the nodes far apart in a short time to relieve the local temperature. When calculating the node movement sequence, the path optimization algorithm can be used to constrain the acceleration limit and positioning accuracy of the galvanometer, so that the actual walking trajectory covers the partitions with higher priority and does not spend too long in the low-priority area. When the node arrangement process is completed, a set of node arrangement sequences will be output, and each node in the sequence is associated with its partition number and energy priority label, thereby providing accurate coordinates and timing instructions for the next step of pulse energy scheduling. In this way, the entire galvanometer scanning will be able to balance the detailed processing of high-priority partitions and the efficient coverage of low-priority partitions, and achieve differentiated energy distribution between partitions over a large range.

[0027] According to the node arrangement sequence, the pulse energy amplitude and pulse width distribution parameters of the pulse laser are scheduled and processed. In the process of obtaining the initial scanning path, it is necessary to allocate appropriate pulse output forms to each node in the scanning timing. The system reads the node coordinates and the corresponding regional priorities from the node arrangement sequence in turn, and then combines the pulse adaptability index to determine whether the current node is more suitable for high-energy short pulses or low-energy long pulses. If a node falls in a high-reflection partition and the matching index is low, it is advisable to select an output with a higher energy peak but a narrower pulse width to ensure that the marking depth can offset the high reflection loss of the surface. Conversely, in a planar area with high optical absorption, the pulse width can be relaxed to cover a larger area in one scan. When generating a pulse sequence, the scheduling algorithm combines the light emission limit of the laser and the acceleration characteristics of the galvanometer to calculate the time domain alignment of the pulse and the node one by one. If the physical distance between the nodes is far, a longer pulse interval is arranged to keep the galvanometer in a reasonable thermal balance during movement. The final output of the initial scanning path not only includes the movement trajectory of the galvanometer in two-dimensional or three-dimensional space, but also marks the pulse power and pulse width triggered by each node in the time dimension, providing clear operating instructions for the actual marking process and ensuring that the effective energy input in the high-reflection area and the processing speed requirements in the general area are taken into account during the scanning process.

[0028] Furthermore, the pulse adaptability discrimination processing of the pulse sampling sequence and the surface feature information to obtain the pulse adaptability index between the pulse sampling sequence and the surface feature information includes: performing energy mapping processing on the pulse energy peak in the pulse sampling sequence and the multi-channel reflection data in the surface feature information to obtain an energy-feature mapping table; performing difference comparison processing on each pulse node in each pulse sampling sequence and different reflection areas according to the energy-feature mapping table to obtain pulse-area difference statistics; performing segmented correction processing on the pulse nodes with excessive deviations in the pulse-area difference statistics to obtain a corrected matching matrix; performing comprehensive evaluation processing on the pulse energy distribution and surface feature information of the pulse laser according to the corrected matching matrix to obtain the pulse adaptability index.

[0029] Specifically, in the process of performing energy mapping processing on the pulse energy peak in the pulse sampling sequence and the multi-channel reflection data in the surface feature information to obtain the energy-feature mapping table, it is necessary to first calibrate the pulse sampling sequence and the multi-channel reflection data in the time domain and the space domain respectively. The pulse sampling sequence usually records the amplitude and energy integration information of the laser pulse in a high-frequency sampling manner, wherein the peak value and its corresponding pulse width and pulse integration value are the key parameters that need to be acquired. The multi-channel reflection data can be collected synchronously through the visible light channel and the infrared channel, so as to record the optical response of the material surface to the laser at different wavelengths. In order to associate the two to the same scale, it is necessary to first normalize the peak value in the pulse sampling sequence, and map the brightness or thermal radiation value of each pixel in the multi-channel reflection data accordingly, so that the two can be compared in the same or similar numerical range. An interpolation algorithm or a nearest neighbor matching algorithm can be used to perform one-to-one or one-to-many corresponding processing on the pulse peak sequence and the reflection data, so as to establish a coupling relationship between the pulse energy and the optical response of the material numerically. If the pulse peak is too high but the reflection brightness of the corresponding area is at a medium level, it will be recorded as a feature combination of "high energy-medium reflection" in the energy-feature mapping table. If both are high, it will be recorded as "high energy-high reflection". In this mapping table, each table cell may contain the pulse energy range, pulse peak duration, and corresponding multi-channel optical information. Through this refined mapping structure, the energy matching degree of each pulse in different surface areas can be measured, laying a numerical foundation for the difference comparison in the next stage.

[0030] According to the energy-feature mapping table, each pulse node in each pulse sampling sequence is subjected to a difference comparison process with different reflection areas. In the process of obtaining the pulse-area difference statistical results, it is necessary to evaluate the degree of deviation between the pulse node and the surface reflection area at the statistical level. In order to complete this operation, the energy-feature mapping table can be regarded as a multidimensional relationship matrix, in which each pulse node contains parameters such as peak energy, pulse width and pulse duration, and different reflection areas contain brightness values, surface roughness indicators and possible thermal radiation intensity. The difference comparison will first pair each pulse node with one or more candidate areas, and calculate the absolute difference or square difference between the pulse index and the reflection index to measure whether the energy input of the pulse in the area is excessive or insufficient. If the peak value of a node is much higher than the average absorption capacity of the area, it is recorded as a high difference. If the match is close, it can be regarded as a low difference. After completing the matching calculation for all nodes, a pulse-area difference statistical result will be generated, and each record indicates the difference level of a pulse in a certain area. If some pulses cannot form a reasonable energy correspondence with any area in the multi-channel reflection data, the difference will be expressed as a continuous extreme value. The statistical results are stored in the form of numerical values ​​or graphs, indicating the deviation of each pulse in the global range, providing quantifiable guidance for subsequent segment correction.

[0031] In the process of obtaining the corrected matching degree matrix by performing segmented correction processing on the pulse nodes with excessive deviation in the pulse-region difference statistical results, it is necessary to first screen out those pulse nodes whose difference exceeds the preset threshold, and confirm that the deviation is caused by reasons such as too high or too low laser energy or the pulse duration not matching the thermal characteristics of the material. For nodes with too high deviation, the energy coefficient corresponding to the pulse peak value can be reduced or the pulse width can be shortened during segmented correction, so that the pulse energy tends to be consistent with the actual absorption capacity of the surface. For nodes with too low deviation, it is possible to try to increase the power calibration value of the pulse during the correction stage to compensate for the higher reflection or rapid heat dissipation on the surface of the material. Segmented correction can divide the sequence into several intervals according to the pulse triggering time, and each interval only performs adjustments on the pulse nodes with larger deviations to avoid affecting the part where the matching degree has met the requirements. If some pulse nodes show excessive deviations in different regions, they can be handed over to a higher level of correction strategy for fusion processing, such as comprehensively considering factors such as reflection data and thermal diffusion rate and making sliding corrections to the pulse energy curve as a whole. When all the out-of-limit nodes are corrected, a corrected matching matrix can be reconstructed by updating the corresponding energy-feature mapping relationship. This matrix records the degree of coupling between the pulse and the material surface with a new energy ratio and time mark, thus providing a reliable basis for the final evaluation.

[0032] According to the corrected matching matrix, the pulse energy distribution and surface feature information of the pulse laser are comprehensively evaluated and processed. In the process of obtaining the pulse fitness index, the corrected pulse energy curve and the multi-channel reflection data need to be scored for overall matching. First, by traversing the corrected matching matrix, the energy attributes of each pulse node are matched again with the optical characteristics of the corresponding area, and the difference between the two is evaluated to see whether it has dropped to an acceptable level. If there are still a few pulse nodes that maintain a high deviation value after correction, a punitive weight can be given when calculating the pulse fitness index to reduce the overall fitness score, so that the system recognizes that the batch of pulses still needs to be vigilant in subsequent use. If the matching degree of most nodes and regions reaches the ideal range after correction, the cumulative score or forward weighting can be set in the evaluation algorithm to push up the final index value. The comprehensive evaluation process can adopt a weighted average or a weighted geometric average method, with pulse peak energy, pulse width, surface reflectivity and material thermal conductivity as the main evaluation dimensions. The obtained pulse fitness index numerically reflects the excellent matching degree of the pulse sequence to the surface characteristics of the marked component.

[0033] 103. Perform synchronous correction processing on the trigger timing of the pulse laser output by the laser marking device according to the initial scanning path to obtain a target marking path; In one embodiment of the present invention, the trigger timing of the pulsed laser output by the laser marking device is synchronously corrected according to the initial scanning path to obtain the target marking path, including: comparing and evaluating each coordinate node in the initial scanning path with the ideal pulse trigger moment in the pulse adaptability index to obtain a trigger timing difference table; performing phase calibration processing on the trigger signal of the pulsed laser and the coordinate signal of the laser scanning galvanometer according to the trigger timing difference table to obtain a synchronous trigger parameter group; performing segmented matching operation processing on the pulse energy amplitude of each pulsed laser and the galvanometer movement speed of the laser scanning galvanometer according to the synchronous trigger parameter group to obtain a trigger linkage instruction set; performing time marking processing on the trigger nodes of each scanning segment of the initial scanning path according to the trigger linkage instruction set to obtain the target marking path.

[0034] Specifically, each coordinate node in the initial scanning path is compared and evaluated with the ideal pulse triggering time in the pulse adaptability index. In the process of obtaining the trigger timing difference table, it is necessary to first establish a clear pulse triggering expected value for each node in the coordinate domain and time domain, and combine the mapping relationship between pulse energy and regional characteristics in the pulse adaptability index to compare the expected value with the trigger time actually assigned to the node in the initial scanning path. In order to uniformly manage these two types of information in the data structure, it is necessary to record the galvanometer coordinates and the initial trigger time label for each node in the initial scanning path, and to clarify when each pulse should be released under ideal conditions in the pulse adaptability index. When a node is read, the algorithm extracts the coordinate position of the node and the originally configured pulse triggering time, and then retrieves the pulse adaptability index to confirm the optimal pulse release time corresponding to the node. If the two do not coincide, the time difference between them is recorded in the difference table, and the positive and negative difference are marked to distinguish between advance or delay. The numerical calculation stage will traverse all nodes, thereby accumulating a multi-dimensional timing difference table. The table rows usually correspond to the node numbers, and the columns represent the trigger time and its error. If the absolute value of the difference is small, it means that the pulse trigger is relatively close to the ideal state. If the deviation is significant, a subsequent phase calibration step is required to correct it. The difference table not only saves the error information at the time level, but also records the pulse energy level or pulse width difference retrieved from the pulse adaptability index for cross-dimensional synchronization scheduling during subsequent corrections. In this way, it can be ensured that when the scanning galvanometer moves to the target coordinate, each node can be consistent with the pulse output time with the highest matching degree, so that the pulse energy and the material reflection area are better matched.

[0035] According to the trigger timing difference table, the trigger signal of the pulse laser and the coordinate signal of the laser scanning galvanometer are phase-calibrated to obtain the synchronous trigger parameter group. In the process, it is necessary to set the relative delay or relative advance amount for the pulse trigger instruction and the galvanometer movement instruction respectively within the framework of the combination of hardware and software. At the hardware level, a programmable delay module can be added between the laser trigger port and the galvanometer controller. At the software level, a series of delay values ​​or phase offsets are generated by calculating and analyzing the data of each record in the timing difference table. If the pulse corresponding to a node should be released in advance before the coordinates arrive, a negative offset is assigned so that the laser emits a pulse before the galvanometer completes the forward movement, otherwise a positive offset is configured. In practice, the acceleration and deceleration of the dynamic response of the galvanometer are also considered to ensure that the pulse after phase calibration will not cause the energy concentration position to deviate due to the galvanometer being too fast or too slow. When the pulse trigger and coordinate arrival time of all nodes are calibrated, the system will output a synchronous trigger parameter group, which is usually stored in the form of an array or dictionary, in which each element contains the node number, the delay (or advance) amount, and the necessary fine-tuning of the pulse energy. If some nodes are marked as extremely high energy or extremely low energy in the pulse adaptability index, an additional energy correction field may be added to the parameter group to be submitted to the segment matching process for execution during subsequent operations.

[0036] According to the synchronous trigger parameter group, the pulse energy amplitude of each pulse laser and the galvanometer movement speed of the laser scanning galvanometer are processed by segmented matching operation. In the process of obtaining the trigger linkage instruction set, it is necessary to comprehensively calculate the time offset of each node, its energy correction amount and the galvanometer movement state. The segmented matching operation refers to dividing the scanning process into several sections on the timeline, each section corresponds to a group of nodes, and batch adjustments are made for the unified or similar energy correction requirements of this group of nodes. For example, if multiple nodes in a section belong to the high reflection area, and the synchronous trigger parameter group prompts that these node pulses need a higher energy peak, the system will also increase the amplitude upper limit of the pulse in the section to coordinate the marking depth with the surface properties. If another section is located in the buffer zone or the ordinary material area, the parameter group will allocate a lower energy amplitude to the section and appropriately speed up the galvanometer movement speed. The operation process accumulates data at the end of each section, forms an instruction set with the allocated pulse amplitude and phase calibration value, and inserts the necessary transition time at the section transition to prevent the laser power from changing suddenly or the galvanometer speed from changing suddenly and affecting the overall accuracy. The final trigger linkage instruction set will list the pulse parameters and galvanometer parameters of each section, and attach timestamps or sequence codes to facilitate subsequent unified scanning execution.

[0037] According to the trigger linkage instruction set, the trigger nodes of each scanning section of the initial scanning path are time-stamped. In the process of obtaining the target marking path, it is necessary to clarify a final trigger moment for each trigger node in each section based on the completed phase calibration and segment matching operation results. At this time, the timing distribution of the section where each node is located will be replaced on the basis of the initial scanning path. If the pulse energy amplitude or galvanometer speed in the section has been corrected as a whole, the new amplitude and motion parameters need to be written when marking these nodes. In this way, a complete target marking path can be generated, including precise position marking for node coordinates, final moment definition for pulse triggering, and dynamic speed configuration for galvanometer motion. If there are special heat-sensitive areas in some sections, the temperature balance waiting time in the trigger linkage instruction set can also be included in the time marking link, so that the node reserves enough time before arrival to avoid overheating or excessive reflection. When all nodes are time-stamped, the target marking path will be presented in a list or graphic form. In addition to the coordinate value, each node also has a corresponding pulse time, energy parameter and galvanometer speed. In this way, the system can release pulses with higher synchronization accuracy when performing laser marking, avoiding misalignment between the galvanometer and the laser signal, thereby maintaining stable marking accuracy and uniform energy distribution under high reflection and multi-curved surface conditions.

[0038] 104. Perform dynamic observation and local heat dispersion processing on the marking area of ​​the component to be marked according to the target marking path and the pulse energy reference to obtain the component surface after processing of the component to be marked.

[0039] In one embodiment of the present invention, the marking area of ​​the component to be marked is dynamically observed and locally thermally dispersed according to the target marking path and the pulse energy reference to obtain the component surface after the processing of the component to be marked, including: performing multi-channel continuous observation processing on the marking result completed by the target marking path to obtain a thermal radiation sequence after marking; comparing the pulse energy distribution of the pulse laser with the local temperature curve according to the thermal radiation sequence after marking to obtain a high-temperature concentration section; performing short-time pulse or airflow heat dissipation processing on the high-temperature concentration section according to the pulse energy reference to obtain a thermal dispersion effect set; performing low-energy supplementary engraving processing on the incomplete positions of the marking lines in the marking result according to the thermal dispersion effect set to obtain the component surface after the processing of the component to be marked.

[0040] Specifically, in the process of performing multi-channel continuous observation and processing on the marking results completed by the target marking path and obtaining the thermal radiation sequence after marking, it is necessary to track the optical and thermal changes of the material surface after marking by coordinating the visible light channel and the infrared channel for simultaneous acquisition. In order to ensure that the values ​​of reflectivity and thermal radiation can be recorded under the same time reference, multiple synchronous cameras or thermal imagers can be arranged near the laser marking equipment to capture images of the target area immediately after each marking is completed. The visible light channel mainly identifies the integrity and grayscale contrast of the marking lines, while the infrared channel monitors the temperature rise of the material surface and generates a thermal distribution map. In order to obtain dynamic information, it is necessary to continuously collect multiple frames of data for a period of time near the end or after the end of the marking process to form a time series data stack. Each frame of data carries an accurate timestamp so that the temperature distribution of each frame image can be associated with the evolution law of time during subsequent calculations. If there are grooves or blind spots in the marking area, the camera can be equipped with multiple angles or an optical reflector group to ensure that the light can effectively reach each surface detail. After the acquisition of multi-channel continuous observation images is completed, they will be uniformly organized into a storage unit and stacked into a three-dimensional or four-dimensional data set in chronological order. The system then performs illumination correction and image distortion correction on the data set to eliminate the influence of ambient light changes or viewing angle changes, thereby retaining the true thermal dynamic information of the material in a short period of time after marking. Finally, after denoising and interpolation processing, a post-marking thermal radiation sequence is output, which includes radiation values ​​or pixel grayscales at different times and different spectral channels, which is used for subsequent identification of potential local overheating or incomplete heat dissipation areas; According to the thermal radiation sequence after marking, the pulse energy distribution of the pulsed laser is compared with the local temperature curve. In the process of obtaining the high-temperature aggregation section, it is necessary to first perform time-series slicing on the thermal radiation sequence, align the visible light and infrared image of the same time slice, and perform edge detection on the infrared pixels in space to identify areas with high temperature or uneven heat dissipation. Then match these temperature distributions with the pulse energy distribution, which can be derived from the previously recorded pulse energy benchmark or the energy parameter assigned to each scanning node during marking. If the surface temperature value is detected to be continuously high at a certain moment, and the pulse energy delivery is relatively concentrated, it can be determined that this is a high-temperature aggregation phenomenon. In order to quantify this phenomenon, partition statistics can be performed on the infrared image. If the average temperature of a certain area exceeds a certain threshold of the surrounding benchmark temperature, the area is marked as a candidate for the high-temperature aggregation section. Next, it is necessary to analyze the thermal decay rate in the time dimension, fit the temperature decay rate with the pulse power, and if the value shows a slow decay or a continuous increase at multiple moments, it can be determined that the section has significant thermal concentration characteristics. At this time, these high-temperature accumulation sections will be individually marked, their coordinate range and pulse energy input time will be recorded, and the corresponding thermal curve or temperature gradient information will be extracted to lay an accurate coordinate and energy basis for the next step of local heat dissipation treatment; According to the pulse energy benchmark, the high-temperature concentrated section is subjected to short-time pulse or airflow heat dissipation treatment. In the process of obtaining the heat dispersion effect set, it is necessary to locate the specific coordinates and thermal values ​​of the area according to the marking information of the high-temperature concentrated section, and quickly apply targeted heat dissipation strategies on the marking head or auxiliary cooling module. If the short-time pulse method is selected, the system will dispatch a group of low-energy or extremely short-width pulses to disturb the surface of the area, so that the area will produce local convection or evaporation effects due to the trace energy input, thereby driving the surface cooling. On the other hand, if the high-temperature section is large and the temperature exceeds the safety threshold, it is necessary to call the airflow heat dissipation module to use gas jets to cool down at the coordinates, and stop or slow down the repeated irradiation of the laser in this area to prevent further heating. In order to accurately evaluate the effect of the heat dissipation process, it is necessary to continue to monitor the temperature change rate of the infrared image during the execution of the heat dissipation process. If the temperature has fallen back to the safe range after heat dissipation, the result of the treatment is recorded in the heat dispersion effect set, and parameters such as the heat dissipation duration, the pulse energy level used, or the airflow speed are indicated. If the heat dissipation is not enough to reverse the temperature rise trend in the area, the heat dispersion strategy can increase the airflow intensity or increase the frequency of low-energy pulse disturbances until a significant drop in the temperature curve is detected. Ultimately, the heat dispersion effect includes not only the temperature reduction after heat dissipation, but also the image comparison before and after heat dissipation, so as to determine whether the part needs further repair later; According to the heat dispersion effect set, low-energy supplementary engraving is performed on the incomplete position of the engraved line in the engraving result. In the process of obtaining the surface of the component after the processing of the component to be marked, it is necessary to analyze the multi-channel continuous observation image again after the heat dissipation is completed to confirm whether there is excessive melting, interruption, insufficient depth and other phenomena of the engraved line in the previous high temperature area or the area with large temperature contrast. If the engraved line is detected to be discontinuous or the edge is unclear in some positions, the low-energy pulse scheme can be called for secondary processing to avoid causing a large-scale temperature increase again after the previous heat dissipation operation. The low-energy pulse can be sprayed for a short time at a local position, and the line deformation or nicking can be corrected with fine galvanometer scanning. After the line edge is identified in the optical detection, the laser light position is aligned with the gap for filling-in marking. If the surface of the component contains high-precision patterns such as scales or QR codes, the pattern information collected in real time can be compared with the vector shape of the original design through a matching algorithm. If the error exceeds the set tolerance, a low-energy impact is performed again at the corresponding coordinate, and the integrity of the corrected line is immediately evaluated. In this way, the marking quality can be finely trimmed while maintaining local temperature balance to avoid repeated impacts under high energy. After completing the low-energy engraving of all incomplete positions, the surface will be tested again in the visible light and infrared channels. Once it is confirmed that the line contour is complete and the surface temperature is balanced, the system can determine that the part to be marked has met the requirements and output the final processed part surface.

[0041] In this embodiment, the pulsed laser output by the laser marking device is processed for energy waveform capture, and multi-channel visual inspection is performed on the surface reflection area of ​​the component to be marked based on the pulse energy reference to obtain surface feature information; the initial scanning path is planned according to the surface feature information, and the trigger timing of the pulsed laser is synchronously corrected to obtain the target marking path; the marking area is dynamically observed and locally heat dispersed in combination with the target marking path and the pulse energy reference to obtain the processed component surface. This method is suitable for fine marking on high reflectivity or complex geometric surfaces, and can make timely adjustments to local temperature rise or optical feature changes during the marking process to reduce depth unevenness and thermal damage.

[0042] The above describes the laser surface treatment method based on vision control in the embodiment of the present invention. The following describes the laser surface treatment device based on vision control in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a laser surface treatment device based on visual control includes: The capture detection module 201 is used to perform energy waveform capture processing on the pulse laser output by the laser marking device to obtain a pulse energy reference, and perform multi-channel visual detection processing on the surface reflection area of ​​the to-be-marked component according to the pulse energy reference to obtain surface feature information of the to-be-marked component; A trajectory planning module 202 is used to plan the trajectory of the laser scanning galvanometer of the laser marking device according to the pulse energy reference and the surface feature information of the part to be marked, so as to obtain an initial scanning path; A synchronization correction module 203 is used to perform synchronization correction processing on the trigger timing of the pulse laser output by the laser marking device according to the initial scanning path to obtain a target marking path; The observation and heat dissipation module 204 is used to dynamically observe and perform local heat dispersion processing on the marking area of ​​the component to be marked according to the target marking path and the pulse energy reference, so as to obtain the component surface after the processing of the component to be marked.

[0043] In an embodiment of the present invention, the laser surface treatment device based on vision control runs the above-mentioned laser surface treatment method based on vision control, and the laser surface treatment device based on vision control performs energy waveform capture processing on the pulsed laser output by the laser marking device, and performs multi-channel visual detection on the surface reflection area of ​​the component to be marked based on the pulse energy reference to obtain surface feature information; plans the initial scanning path according to the surface feature information, and synchronously corrects the trigger timing of the pulsed laser to obtain the target marking path; combines the target marking path with the pulse energy reference, performs dynamic observation and local heat dispersion processing on the marking area, and obtains the processed component surface. This method is suitable for fine marking on high reflectivity or complex geometric surfaces, and can make timely adjustments to local temperature rise or optical feature changes during the marking process to reduce depth unevenness and thermal damage.

[0044] above Figure 2 The laser surface processing device based on vision control in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The laser surface processing device based on vision control in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0045] Figure 3: is a structural schematic diagram of a laser surface processing device based on vision control provided by an embodiment of the present invention. The laser surface processing device 300 based on vision control may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage device terminals) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 can be short-term storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the laser surface processing device 300 based on vision control. Furthermore, the processor 310 can be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the laser surface processing device 300 based on vision control to implement the steps of the above-mentioned laser surface processing method based on vision control.

[0046] The laser surface processing device 300 based on vision control may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It can be understood by those skilled in the art that Figure 3 The structure of the laser surface processing equipment based on vision control shown does not constitute a limitation on the laser surface processing equipment based on vision control provided by the present invention, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0047] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0048] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A laser surface treatment method based on visual control, characterized in that: The laser surface treatment method based on vision control comprises: Performing energy waveform capture processing on the pulsed laser output by the laser marking device to obtain a pulse energy reference, and performing multi-channel visual inspection processing on the surface reflection area of ​​the to-be-marked component according to the pulse energy reference to obtain surface feature information of the to-be-marked component; Planning and processing the trajectory of the laser scanning galvanometer of the laser marking device according to the pulse energy reference and the surface feature information of the component to be marked to obtain an initial scanning path; Performing synchronous correction processing on the triggering timing of the pulsed laser output by the laser marking device according to the initial scanning path to obtain a target marking path; The marking area of ​​the component to be marked is dynamically observed and locally heat dispersed according to the target marking path and the pulse energy reference to obtain a component surface after the processing of the component to be marked.

2. The laser surface treatment method based on vision control according to claim 1, characterized in that: The energy waveform capture process is performed on the pulse laser output by the laser marking device to obtain a pulse energy reference, and a multi-channel visual inspection process is performed on the surface reflection area of ​​the to-be-marked component according to the pulse energy reference to obtain the surface feature information of the to-be-marked component, including: Performing high-speed photoelectric sampling and acquisition processing on the pulse signal output by the laser to obtain a pulse sampling sequence; Performing feature extraction processing on the peak segment and the attenuation segment of the pulse signal according to the pulse sampling sequence to obtain a pulse energy reference; Performing multi-channel visual recognition on the surface reflection area of ​​the component to be marked to obtain a multi-channel image set; Optical feature comparison and pixel partition clustering processing are performed on the multi-channel image set according to the pulse energy reference to obtain surface feature information of the component to be marked.

3. The laser surface treatment method based on vision control according to claim 2 is characterized in that: The optical feature comparison and pixel partition clustering processing are performed on the multi-channel image set according to the pulse energy reference to obtain the surface feature information of the part to be marked, including: Performing optical spectrum extraction processing on each pixel point in the multi-channel image set to obtain an optical feature matrix; Performing threshold partitioning processing on corresponding pixel values ​​in the optical feature matrix according to the pulse energy reference to obtain a preliminary partitioning result; Performing grouping and clustering processing on the high-reflection pixel group and the general-reflection pixel group of the preliminary partitioning result to obtain a pixel clustering index table; The concave-convex boundary area is processed for contour positioning according to the pixel clustering index table to obtain surface feature information of the component to be marked.

4. The laser surface treatment method based on vision control according to claim 2 is characterized in that: The planning and processing of the trajectory of the laser scanning galvanometer of the laser marking device according to the pulse energy reference and the surface feature information of the part to be marked to obtain the initial scanning path includes: Performing pulse adaptability discrimination processing on the pulse sampling sequence and the surface feature information to obtain a pulse adaptability index between the pulse sampling sequence and the surface feature information; According to the pulse adaptability index, pulse energy priorities are divided for different regions of the component to be marked to obtain a region priority table; Arrange the path nodes of the trajectory of the laser scanning galvanometer of the laser marking device according to the area priority table to obtain a node arrangement sequence; The pulse energy amplitude and pulse width allocation parameters of the pulse laser are scheduled according to the node arrangement sequence to obtain the initial scanning path.

5. The laser surface treatment method based on vision control according to claim 4 is characterized in that: The performing pulse adaptability discrimination processing on the pulse sampling sequence and the surface feature information to obtain the pulse adaptability index between the pulse sampling sequence and the surface feature information comprises: Performing energy mapping processing on the pulse energy peak value in the pulse sampling sequence and the multi-channel reflection data in the surface feature information to obtain an energy-feature mapping table; According to the energy-feature mapping table, each pulse node in each pulse sampling sequence is subjected to a difference comparison process with different reflection regions to obtain a pulse-region difference statistical result; Performing segmented correction processing on the pulse nodes with excessive deviations in the pulse-region difference statistical results to obtain a corrected coordination matrix; The pulse energy distribution and surface feature information of the pulse laser are comprehensively evaluated and processed according to the modified matching degree matrix to obtain the pulse matching degree index.

6. The laser surface treatment method based on vision control according to claim 4, characterized in that: The step of synchronously correcting the trigger timing of the pulsed laser output by the laser marking device according to the initial scanning path to obtain a target marking path includes: Comparing and evaluating each coordinate node in the initial scanning path with the ideal pulse triggering time in the pulse adaptability index to obtain a trigger timing difference table; Performing phase calibration processing on the trigger signal of the pulse laser and the coordinate signal of the laser scanning galvanometer according to the trigger timing difference table to obtain a synchronous trigger parameter group; According to the synchronous trigger parameter group, the pulse energy amplitude of each pulse laser and the galvanometer movement speed of the laser scanning galvanometer are processed in a segmented matching operation to obtain a trigger linkage instruction set; The trigger nodes of each scanning section of the initial scanning path are time-marked according to the trigger linkage instruction set to obtain a target marking path.

7. The laser surface treatment method based on vision control according to claim 1, characterized in that: The dynamically observing and locally dispersing the marking area of ​​the component to be marked according to the target marking path and the pulse energy reference to obtain the component surface after the processing of the component to be marked comprises: Performing multi-channel continuous observation processing on the marking results completed by the target marking path to obtain a thermal radiation sequence after marking; Comparing the pulse energy distribution of the pulsed laser with the local temperature curve according to the thermal radiation sequence after marking, and obtaining a high-temperature concentration section; According to the pulse energy standard, the high temperature concentration section is subjected to short-time pulse or airflow heat dissipation treatment to obtain a heat dispersion effect set; According to the heat dispersion effect set, low-energy supplementary engraving is performed on the incomplete positions of the marking lines in the marking result to obtain the processed component surface of the component to be marked.

8. A laser surface treatment device based on visual control, characterized in that: The laser surface treatment device based on vision control comprises: A capture detection module is used to perform energy waveform capture processing on the pulse laser output by the laser marking device to obtain a pulse energy reference, and perform multi-channel visual detection processing on the surface reflection area of ​​the to-be-marked component according to the pulse energy reference to obtain surface feature information of the to-be-marked component; A trajectory planning module, used to plan the trajectory of the laser scanning galvanometer of the laser marking device according to the pulse energy reference and the surface feature information of the part to be marked, so as to obtain an initial scanning path; A synchronous correction module, used for performing synchronous correction processing on the triggering timing of the pulse laser output by the laser marking device according to the initial scanning path to obtain a target marking path; The observation and heat dissipation module is used to dynamically observe and perform local heat dispersion processing on the marking area of ​​the component to be marked according to the target marking path and the pulse energy reference, so as to obtain the component surface after the processing of the component to be marked.

9. A laser surface treatment device based on visual control, characterized in that: The laser surface processing device based on vision control includes: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the vision-controlled laser surface processing device to perform the steps of the vision-controlled laser surface processing method according to any one of claims 1 to 7.

Images