技术领域Technical field
本发明涉及图形图像处理技术领域，尤其是一种极柱焊接环形区域的提取方法。The invention relates to the technical field of graphics and image processing, in particular to a method for extracting the annular area of pole welding.
背景技术Background technique
对于电池极柱焊接区域的检测识别，以前国内主要是使用人工测量特征参数的方法进行缺陷的鉴定，该方法检测效率低，鉴定结果的主观性大，不能适应机器视觉行业发展的需要。随着计算机技术和数字图像处理技术的发展，国内不少研究者将机器视觉技术和数字图像处理技术应用于机器视觉特征参数的测量中，实现了焊接缺陷某一项特征参数的测量但是这些研究与工程实际应用尚有一段距离。For the detection and identification of battery pole welding areas, in the past, manual measurement of characteristic parameters was mainly used to identify defects. This method has low detection efficiency, high subjectivity in identification results, and cannot meet the needs of the development of the machine vision industry. With the development of computer technology and digital image processing technology, many domestic researchers have applied machine vision technology and digital image processing technology to the measurement of machine vision characteristic parameters, achieving the measurement of a certain characteristic parameter of welding defects. However, these studies There is still some distance between it and practical engineering applications.
因此，极柱焊接区域的征参数的快速低成本测量技术的研究仍是该行业目前迫切需要解决的问题。Therefore, the research on fast and low-cost measurement technology of characteristic parameters in the pole welding area is still an urgent problem that the industry needs to solve.
发明内容Contents of the invention
本发明提出一种极柱焊接环形区域的提取方法，对焊接部位典型的对接和角接接头都有效，且能推广至其他接头轮廓焊缝的提取，利于焊接信息采集与处理系统的集成。The present invention proposes a method for extracting the annular area of pole welding, which is effective for typical butt and corner joints of the welding part, and can be extended to the extraction of other joint contour welds, which is beneficial to the integration of welding information collection and processing systems.
本发明采用以下技术方案。The present invention adopts the following technical solutions.
一种极柱焊接环形区域的提取方法，可用于在图像处理过程中提取电池极柱的焊接环形区域，所述提取方法包括以下步骤；A method for extracting the welding annular area of the pole pole, which can be used to extract the welding annular area of the battery pole during image processing. The extraction method includes the following steps;
步骤一、采集极柱焊接部位的俯视向图形，以图形中心处的内圆为第一特征模板；Step 1: Collect the top view pattern of the pole welding part, and use the inner circle at the center of the pattern as the first feature template;
步骤二、用Canny算子对内圆的边缘进行边缘检测和膨胀腐蚀处理；Step 2: Use the Canny operator to perform edge detection and expansion corrosion processing on the edge of the inner circle;
步骤三、以俯视向图形的外圆为第二特征模板；Step 3: Use the outer circle of the top-view figure as the second feature template;
步骤四、用Canny算子对外圆的边缘进行边缘检测和膨胀腐蚀处理；Step 4: Use the Canny operator to perform edge detection and expansion corrosion processing on the edges of the outer circle;
步骤五、获取俯视向图形处的环形边界XLD轮廓，进行灰度处理，获得对应的二值图；Step 5: Obtain the XLD outline of the annular boundary at the top view graph, perform grayscale processing, and obtain the corresponding binary image;
步骤六、识别得出焊接环形区域，识别得到焊缝的边界；Step 6: Identify the welding annular area and identify the boundary of the weld;
步骤七、提取焊缝区域。Step 7: Extract the weld area.
在步骤五中，其获取及处理XLD轮廓的步骤为，In step five, the steps to obtain and process the XLD outline are:
步骤S1、对采集的图像进行图像分割，选择感兴趣区域；Step S1: Perform image segmentation on the collected image and select the area of interest;
步骤S2、对感兴趣区域进行亚像素边缘提取以获取XLD轮廓，并且确定轮廓的属性；Step S2: Perform sub-pixel edge extraction on the area of interest to obtain the XLD outline, and determine the attributes of the outline;
步骤S3、对XLD轮廓进行处理以实现轮廓拟合；Step S3: Process the XLD contour to achieve contour fitting;
步骤S4、对拟合后的XLD轮廓进行计算，需得到的计算结果包括XLD轮廓的各端点坐标和拟合圆直径；Step S4: Calculate the fitted XLD profile. The calculation results to be obtained include the coordinates of each endpoint of the XLD profile and the diameter of the fitting circle;
步骤S5、根据步骤S4中的计算结果，对拟合后的XLD轮廓进行特征参数计算，以识别其是否为焊接环形区域。Step S5: According to the calculation results in step S4, calculate the characteristic parameters of the fitted XLD profile to identify whether it is a welding annular area.
所述步骤S1中，采集的图像为RGB三通道的彩色图像，以decompose3算子将RGB图像分解成R、G和B三幅单通道图像，分别对应红色、绿色和蓝色三个颜色通道；然后再用自动阈值算子bin_threshold对R通道图像或G通道图像进行图像分割处理。In step S1, the collected image is an RGB three-channel color image, and the decompose3 operator is used to decompose the RGB image into three single-channel images of R, G and B, corresponding to the three color channels of red, green and blue respectively; Then use the automatic threshold operator bin_threshold to perform image segmentation processing on the R channel image or G channel image.
若步骤S1中，图像分割处理的结果是得到多个不连通的区域，则在所述步骤S2中，把这些不连通的区域合并为一个单一的连通区域，并去除干扰区域，采用以下方法进行处理，首先用算子select_shape选择感兴趣的区域ROI，然后用预设边长的矩形结构元素对区域ROI进行膨胀操作，经膨胀操作后ROI可合并成一个单一连通区域，最后用算子reduce_domain将该单一连通区域与原始图像进行裁剪处理，裁剪出ROI图像，最后可以用算子threshold_sub_pix从该ROI图像中提取图像轮廓边缘，所提取的图像轮廓边缘具有亚像素精度。If in step S1, the result of the image segmentation processing is to obtain multiple disconnected areas, then in step S2, these disconnected areas are merged into a single connected area, and the interference area is removed, using the following method. Processing, first use the operator select_shape to select the area of interest ROI, then use the rectangular structural element with a preset side length to expand the area ROI. After the expansion operation, the ROI can be merged into a single connected area, and finally use the operator reduce_domain to This single connected area is cropped with the original image to crop out the ROI image. Finally, the operator threshold_sub_pix can be used to extract the image contour edge from the ROI image. The extracted image contour edge has sub-pixel accuracy.
在步骤S3中，首先以算子select_shape_ xld筛选识别XLD轮廓位置，然后以算子shape_trans_xld将XLD轮廓图形转换为可供拟合处理的XLD轮廓形状，再对XLD轮廓形状进行拟合处理。In step S3, the operator select_shape_xld is first used to filter and identify the XLD outline position, and then the operator shape_trans_xld is used to convert the XLD outline shape into an XLD outline shape that can be fitted, and then the XLD outline shape is fit.
在以算子select_shape_ xld筛选识别XLD轮廓位置的过程中，以算子select_shape_ xld对XLD轮廓所在区域进行识别，识别出极柱焊接区域的灰白域轮廓，将其与极柱外轮廓进行区分。In the process of filtering and identifying the XLD outline position with the operator select_shape_ xld, the area where the XLD outline is located is identified with the operator select_shape_
在以算子shape_trans_xld将XLD轮廓图形转换为可供拟合处理的XLD轮廓形状的过程中，所述可供拟合处理的XLD轮廓形状为由XLD轮廓转换的XLD凸包形成的多边形。In the process of using the operator shape_trans_xld to convert the XLD outline shape into an XLD outline shape available for fitting processing, the XLD outline shape available for fitting processing is a polygon formed by the XLD convex hull converted from the XLD outline.
在步骤S4、S5中，以机器视觉图像处理软件HALCON拟合后的XLD轮廓进行计算。In steps S4 and S5, the XLD profile fitted by the machine vision image processing software HALCON is used for calculation.
对拟合后的XLD轮廓进行计算所得到的特征参数包括极柱焊接环形区域的半径、圆度、椭圆度、净度和形状。The characteristic parameters calculated from the fitted XLD profile include the radius, roundness, ovality, clarity and shape of the pole welding annular area.
在步骤S1中，采集焊接前不起弧的焊缝图像作为参考图像，利用参考图像自动确定Gabor滤波的滤波角度和感兴趣区域，并且采集焊接图像，对焊接图像进行自适应Gabor滤波获得多方向特征图，同时获取参考图像和方向特征图的感兴趣区域部分；In step S1, the image of the weld seam without arcing before welding is collected as a reference image, and the reference image is used to automatically determine the filtering angle and area of interest of the Gabor filter, and the welding image is collected, and adaptive Gabor filtering is performed on the welding image to obtain multi-directional Feature map, obtain the area of interest part of the reference image and the direction feature map at the same time;
在步骤S1、S2、S3、S4和S5中，利用尺度不变特征变换算法处理上述两个感兴趣区域部分，获得两个图像相似点的匹配位置；对多方向特征图感兴趣区域部分进行自适应局部7值分割处理，获得二值图像，根据已获取的匹配位置的坐标和聚类结果提取焊缝轮廓。In steps S1, S2, S3, S4 and S5, the scale-invariant feature transformation algorithm is used to process the above two regions of interest to obtain the matching positions of similar points in the two images; the region of interest in the multi-directional feature map is automatically Adapt to the local 7-value segmentation process to obtain a binary image, and extract the welding seam contour based on the obtained coordinates of the matching position and the clustering results.
本发明为自动焊接焊缝轮廓提取的通用方法，对典型的对接和角接接头都有效，且能推广至其他接头轮廓焊缝的提取，利于焊接信息采集与处理系统的集成。The invention is a general method for automatic welding seam contour extraction, which is effective for typical butt joints and corner joints, and can be extended to the extraction of other joint contour weld seams, which is beneficial to the integration of welding information collection and processing systems.
本发明将图像处理技术与HALCON技术相结合，提出一种极柱焊接环形区域的提取方法。能将焊接区域提取出来，消除背景杂质带来的影响，提高检测缺陷检测的准确率。从而提高产业链的效率；此外，本专利针对于类似环形，圆形焊接区域的检测提取也是具有实际参考意义的。The present invention combines image processing technology with HALCON technology to propose a method for extracting the annular area of pole welding. It can extract the welding area, eliminate the influence of background impurities, and improve the accuracy of defect detection. Thereby improving the efficiency of the industrial chain; in addition, this patent is also of practical reference significance for the detection and extraction of similar ring-shaped and circular welding areas.
附图说明Description of the drawings
下面结合附图和具体实施方式对本发明进一步详细的说明：The present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments:
附图1是本发明所述方法的图像分步处理示意图；Figure 1 is a schematic diagram of the step-by-step image processing of the method of the present invention;
附图2是本发明获取及处理XLD轮廓的流程示意图；Figure 2 is a schematic flow chart of obtaining and processing XLD contours according to the present invention;
图中：1-内圆；2-外圆；3-焊接环形区域。In the picture: 1-inner circle; 2-outer circle; 3-welding annular area.
具体实施方式Detailed ways
如图所示，一种极柱焊接环形区域的提取方法，可用于在图像处理过程中提取电池极柱的焊接环形区域，所述提取方法包括以下步骤；As shown in the figure, a method for extracting the welding annular area of the pole pole can be used to extract the welding annular area of the battery pole during image processing. The extraction method includes the following steps;
步骤一、采集极柱焊接部位的俯视向图形，以图形中心处的内圆1为第一特征模板；Step 1: Collect the top view pattern of the pole welding part, and use the inner circle 1 at the center of the pattern as the first feature template;
步骤二、用Canny算子对内圆的边缘进行边缘检测和膨胀腐蚀处理；Step 2: Use the Canny operator to perform edge detection and expansion corrosion processing on the edge of the inner circle;
步骤三、以俯视向图形的外圆2为第二特征模板；Step 3: Use the outer circle 2 of the top-view figure as the second feature template;
步骤四、用Canny算子对外圆的边缘进行边缘检测和膨胀腐蚀处理；Step 4: Use the Canny operator to perform edge detection and expansion corrosion processing on the edges of the outer circle;
步骤五、获取俯视向图形处的环形边界XLD轮廓，进行灰度处理，获得对应的二值图；Step 5: Obtain the XLD outline of the annular boundary at the top view graph, perform grayscale processing, and obtain the corresponding binary image;
步骤六、识别得出焊接环形区域3，识别得到焊缝的边界；Step 6: Identify the welding annular area 3 and identify the boundary of the weld;
步骤七、提取焊缝区域。Step 7: Extract the weld area.
在步骤五中，其获取及处理XLD轮廓的步骤为，In step five, the steps to obtain and process the XLD outline are:
步骤S1、对采集的图像进行图像分割，选择感兴趣区域；Step S1: Perform image segmentation on the collected image and select the area of interest;
步骤S2、对感兴趣区域进行亚像素边缘提取以获取XLD轮廓，并且确定轮廓的属性；Step S2: Perform sub-pixel edge extraction on the area of interest to obtain the XLD outline, and determine the attributes of the outline;
步骤S3、对XLD轮廓进行处理以实现轮廓拟合；Step S3: Process the XLD contour to achieve contour fitting;
步骤S4、对拟合后的XLD轮廓进行计算，需得到的计算结果包括XLD轮廓的各端点坐标和拟合圆直径；Step S4: Calculate the fitted XLD profile. The calculation results to be obtained include the coordinates of each endpoint of the XLD profile and the diameter of the fitting circle;
步骤S5、根据步骤S4中的计算结果，对拟合后的XLD轮廓进行特征参数计算，以识别其是否为焊接环形区域。Step S5: According to the calculation results in step S4, calculate the characteristic parameters of the fitted XLD profile to identify whether it is a welding annular area.
所述步骤S1中，采集的图像为RGB三通道的彩色图像，以decompose3算子将RGB图像分解成R、G和B三幅单通道图像，分别对应红色、绿色和蓝色三个颜色通道；然后再用自动阈值算子bin_threshold对R通道图像或G通道图像进行图像分割处理。In step S1, the collected image is an RGB three-channel color image, and the decompose3 operator is used to decompose the RGB image into three single-channel images of R, G and B, corresponding to the three color channels of red, green and blue respectively; Then use the automatic threshold operator bin_threshold to perform image segmentation processing on the R channel image or G channel image.
若步骤S1中，图像分割处理的结果是得到多个不连通的区域，则在所述步骤S2中，把这些不连通的区域合并为一个单一的连通区域，并去除干扰区域，采用以下方法进行处理，首先用算子select_shape选择感兴趣的区域ROI，然后用预设边长的矩形结构元素对区域ROI进行膨胀操作，经膨胀操作后ROI可合并成一个单一连通区域，最后用算子reduce_domain将该单一连通区域与原始图像进行裁剪处理，裁剪出ROI图像，最后可以用算子threshold_sub_pix从该ROI图像中提取图像轮廓边缘，所提取的图像轮廓边缘具有亚像素精度。If in step S1, the result of the image segmentation processing is to obtain multiple disconnected areas, then in step S2, these disconnected areas are merged into a single connected area, and the interference area is removed, using the following method. Processing, first use the operator select_shape to select the area of interest ROI, then use the rectangular structural element with a preset side length to expand the area ROI. After the expansion operation, the ROI can be merged into a single connected area, and finally use the operator reduce_domain to This single connected area is cropped with the original image to crop out the ROI image. Finally, the operator threshold_sub_pix can be used to extract the image contour edge from the ROI image. The extracted image contour edge has sub-pixel accuracy.
在步骤S3中，首先以算子select_shape_ xld筛选识别XLD轮廓位置，然后以算子shape_trans_xld将XLD轮廓图形转换为可供拟合处理的XLD轮廓形状，再对XLD轮廓形状进行拟合处理。In step S3, the operator select_shape_xld is first used to filter and identify the XLD outline position, and then the operator shape_trans_xld is used to convert the XLD outline shape into an XLD outline shape that can be fitted, and then the XLD outline shape is fit.
在以算子select_shape_ xld筛选识别XLD轮廓位置的过程中，以算子select_shape_ xld对XLD轮廓所在区域进行识别，识别出极柱焊接区域的灰白域轮廓，将其与极柱外轮廓进行区分。In the process of filtering and identifying the XLD outline position with the operator select_shape_ xld, the area where the XLD outline is located is identified with the operator select_shape_
在以算子shape_trans_xld将XLD轮廓图形转换为可供拟合处理的XLD轮廓形状的过程中，所述可供拟合处理的XLD轮廓形状为由XLD轮廓转换的XLD凸包形成的多边形。In the process of using the operator shape_trans_xld to convert the XLD outline shape into an XLD outline shape available for fitting processing, the XLD outline shape available for fitting processing is a polygon formed by the XLD convex hull converted from the XLD outline.
在步骤S4、S5中，以机器视觉图像处理软件HALCON拟合后的XLD轮廓进行计算。In steps S4 and S5, the XLD profile fitted by the machine vision image processing software HALCON is used for calculation.
对拟合后的XLD轮廓进行计算所得到的特征参数包括极柱焊接环形区域的半径、圆度、椭圆度、净度和形状。The characteristic parameters calculated from the fitted XLD profile include the radius, roundness, ovality, clarity and shape of the pole welding annular area.
在步骤S1中，采集焊接前不起弧的焊缝图像作为参考图像，利用参考图像自动确定Gabor滤波的滤波角度和感兴趣区域，并且采集焊接图像，对焊接图像进行自适应Gabor滤波获得多方向特征图，同时获取参考图像和方向特征图的感兴趣区域部分；In step S1, the image of the weld seam without arcing before welding is collected as a reference image, and the reference image is used to automatically determine the filtering angle and area of interest of the Gabor filter, and the welding image is collected, and adaptive Gabor filtering is performed on the welding image to obtain multi-directional Feature map, obtain the area of interest part of the reference image and the direction feature map at the same time;
在步骤S1、S2、S3、S4和S5中，利用尺度不变特征变换算法处理上述两个感兴趣区域部分，获得两个图像相似点的匹配位置；对多方向特征图感兴趣区域部分进行自适应局部7值分割处理，获得二值图像，根据已获取的匹配位置的坐标和聚类结果提取焊缝轮廓。In steps S1, S2, S3, S4 and S5, the scale-invariant feature transformation algorithm is used to process the above two regions of interest to obtain the matching positions of similar points in the two images; the region of interest in the multi-directional feature map is automatically Adapt to the local 7-value segmentation process to obtain a binary image, and extract the welding seam contour based on the obtained coordinates of the matching position and the clustering results.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011099777.XACN112200786B (en) | 2020-10-15 | 2020-10-15 | A method for extracting the annular area of pole welding |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011099777.XACN112200786B (en) | 2020-10-15 | 2020-10-15 | A method for extracting the annular area of pole welding |
| Publication Number | Publication Date |
|---|---|
| CN112200786A CN112200786A (en) | 2021-01-08 |
| CN112200786Btrue CN112200786B (en) | 2023-09-22 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011099777.XAActiveCN112200786B (en) | 2020-10-15 | 2020-10-15 | A method for extracting the annular area of pole welding |
| Country | Link |
|---|---|
| CN (1) | CN112200786B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115456995A (en)* | 2022-09-15 | 2022-12-09 | 杭州安脉盛智能技术有限公司 | Method, device, equipment, and readable storage medium for detecting weak welds in welds |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109859186A (en)* | 2019-01-31 | 2019-06-07 | 江苏理工学院 | A kind of lithium battery mould group positive and negative anodes detection method based on halcon |
| CN111462110A (en)* | 2020-04-20 | 2020-07-28 | 广东利元亨智能装备股份有限公司 | Welding seam quality detection method, device and system and electronic equipment |
| CN111598831A (en)* | 2020-03-26 | 2020-08-28 | 浙江科技学院 | Detection method and system for cylindrical roller bearing |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104484857B (en)* | 2014-12-26 | 2017-08-18 | 国网重庆市电力公司电力科学研究院 | Method and system for reading instrument data |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109859186A (en)* | 2019-01-31 | 2019-06-07 | 江苏理工学院 | A kind of lithium battery mould group positive and negative anodes detection method based on halcon |
| CN111598831A (en)* | 2020-03-26 | 2020-08-28 | 浙江科技学院 | Detection method and system for cylindrical roller bearing |
| CN111462110A (en)* | 2020-04-20 | 2020-07-28 | 广东利元亨智能装备股份有限公司 | Welding seam quality detection method, device and system and electronic equipment |
| Title |
|---|
| 双列金属零件加工质量机器视觉检测系统研发;李倩;《中国优秀硕士学位论文全文数据库 工程科技II辑》;第1-75页* |
| 基于机器视觉的圆环电阻片缺陷检测算法研究;朱铮涛;李波锋;何修远;;组合机床与自动化加工技术(第05期);第80-83页* |
| Publication number | Publication date |
|---|---|
| CN112200786A (en) | 2021-01-08 |
| Publication | Publication Date | Title |
|---|---|---|
| CN109816652B (en) | Complex casting defect identification method based on gray level significance | |
| CN114821114B (en) | Groove cutting robot image processing method based on vision system | |
| CN110717872B (en) | Method and system for extracting characteristic points of V-shaped welding seam image under laser-assisted positioning | |
| CN109784344A (en) | An image non-target filtering method for ground plane identification recognition | |
| CN113496483A (en) | Weld seam air hole defect detection method based on image processing | |
| CN112435272B (en) | High-voltage transmission line connected domain removing method based on image contour analysis | |
| CN113077437B (en) | Workpiece quality detection method and system | |
| CN110298344A (en) | A kind of positioning of instrument knob and detection method based on machine vision | |
| CN110443791B (en) | Workpiece detection method and device based on deep learning network | |
| CN114782346B (en) | Cloth image defect detection method based on polymorphic data amplification and block identification | |
| CN111932490B (en) | Visual system grabbing information extraction method for industrial robot | |
| CN109697717B (en) | Lining crack identification method based on image automatic search | |
| CN110060239B (en) | A kind of defect detection method for bottle mouth | |
| CN109544513A (en) | A kind of steel pipe end surface defect extraction knowledge method for distinguishing | |
| CN105057899A (en) | Scanned image recognition method applied to intelligent laser cutting | |
| CN107977960A (en) | A kind of car surface scratch detection algorithm based on improved SUSAN operators | |
| CN117911516B (en) | A weld feature extraction method and system for line laser sensor | |
| CN112200786B (en) | A method for extracting the annular area of pole welding | |
| CN110866916A (en) | Machine vision-based photovoltaic cell black-core black-corner detection method, device and equipment | |
| CN113971681A (en) | A method for edge detection of belt conveyor in complex environment | |
| CN118691542A (en) | A method, device and medium for detecting ceramic tile surface based on machine vision | |
| CN116958714B (en) | Automatic identification method for wafer back damage defect | |
| CN113160163A (en) | Engine cylinder block appearance defect detection algorithm based on machine vision | |
| CN114943710B (en) | Method for locating the center of thin-walled circular tube parts using improved least squares fitting | |
| CN116329842A (en) | An Automatic Tracking Method for Narrow Butt Weld |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information | Inventor after:Huang Jing Inventor after:Qi Yanfeng Inventor after:Lin Zhe Inventor after:Wang Zezhou Inventor before:Huang Jing Inventor before:Qi Yanfeng Inventor before:Lin Zhe Inventor before:Wang Zezhou | |
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address | ||
| CP03 | Change of name, title or address | Address after:No.33, Xuefu South Road, University New District, Minhou County, Fuzhou City, Fujian Province Patentee after:Fujian University of Science and Technology Country or region after:China Address before:No.33, Xuefu South Road, University New District, Minhou County, Fuzhou City, Fujian Province Patentee before:FUJIAN University OF TECHNOLOGY Country or region before:China |