SIA OpenIR  > 智能产线与系统研究室
基于多特征的激光拼焊板焊缝质量检测技术研究
Alternative TitleResearch on Laser Tailored Blank Inspection Based on Multiple Features
寇淼1,2
Department装备制造技术研究室
Thesis Advisor房灵申
ClassificationTG665
Keyword激光拼焊 视觉检测 焊缝检测 多特征
Call NumberTG665/K55/2016
Pages71页
Degree Discipline机械电子工程
Degree Name硕士
2016-05-25
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract本文以激光拼焊为背景,针对基于结构光的视觉测量系统中相关的技术问题和应用需求,参考国内外相关技术和研究,结合相关图像处理算法对焊缝质量检测相关技术进行研究。首先本文介绍了基于结构光的焊缝视觉检测系统,之后对运用于焊缝检测的相关图像处理算法进行研究,最后,搭建了检测平台,并对算法进行验证。主要研究内容如下:1. 对基于结构光的焊缝视觉检测传感器进行建模,确定图像中像素点与结构光平面上点的对应关系;对传感器标定方法进行研究,建立了工业相机的成像模型,采用了基于平面靶标的张正友标定方法,利用Matlab标定工具箱对相机进行标定,对结构光标定中常用的方法进行了介绍和研究。2. 针对采集的焊缝检测图像特征,设计了一套光纹中心线提取方法。首先,为减少图像处理时间,提出了一种感兴趣区域自动提取算法;讨论了图像的滤波去噪方法,并选取了中值滤波方法对图像进行平滑滤波;对几种常用的阈值分割方法进行分析对比,选用大律法确定阈值,并对结构光进行分割提取;针对结构光二值图像存在噪声问题,选用形态学方法进行处理;对几种中心条纹提取方法进行对比分析,综合考虑处理时间和精度选取灰度重心法提取条纹中心线;设计了中心条纹直线拟合方法。3. 对背面焊缝分割方法进行研究。首先,研究了焊缝纹理特征提取方法;使用基于支持向量机的图像分割方法利灰度共生矩阵对焊缝区域进行粗提取;针对焊缝区域纹理结构特点,提出一种焊缝区域精确提取方法;对焊缝边缘进行提取,确定焊缝具体位置。4. 针对基于结构光的视觉测量系统难以有效地检测拼焊板背面焊缝几何尺寸的问题,设计了一种背面焊缝检测方法,以焊缝区域与结构光中心条纹的交点代替不易检测的畸变特征点对焊缝背面几何尺寸进行测量。并搭建了视觉检测平台,对算法进行验证。最后,对研究工作作出总结,并对今后工作进行了展望。
Other AbstractThis paper plan to research some problems in visual inspection of tailored blanks and try to solve some questions by referring to the related technology and research at home and abroad. First this paper introduced the visual inspection method based on structured light, and then researched related image processing algorithms. At last, set up a test platform to verify the algorithm. The main research contents are as follows: 1. Build a model of the visual inspection system based on structured light, and give some formulas about the corresponding relation of the the pixels in the image with the points on structured light plane. Then, researched the sensor calibration methods. The camera imaging model was established, and then chose Zhang’s method to implement camera calibration. At last, the problem of how to calibrate the structured light was researched. 2. The characteristics of the gray image of weld seam in the back of the blanks was analyzed, and a image processing flow of how to extract the center line of structured light was put forward . First of all, to reduce the processing time, a automatic extraction algorithm of the region of interest was proposed. The filtering and noise reduction methods of image were discussed; Analyzed and compared several commonly used threshold segmentation methods, choose a suitable one to extract the structured light; To solve the noise problems of binary image, put forward a morphological processing methods; Considering precision and processing time, selected a quick accurate algorithm to get the center stripe; A fitting method of center stripe line was designed. 3. Then researched the extraction methods of seam region. First, researched the texture feature extraction methods; Used image segmentation method based on support vector machine (SVM) to get the approximate region of the weld seam area. According to the characteristics of seam, this paper puts forward a kind of method to get the accurate position of seam. 4. The visual system only based on structured is difficult to detect the size of the defects on the back of the blanks. This paper provided a method to solve this, that is using the intersection points of the structured line and the weld seam instead of the distortion points on the structured line which are difficult to detect. At last, a visual inspection platform was set up to verify this algorithm. Finally, the research work is summarized,and the future work is prospected.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/19665
Collection智能产线与系统研究室
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院大学
Recommended Citation
GB/T 7714
寇淼. 基于多特征的激光拼焊板焊缝质量检测技术研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2016.
Files in This Item:
File Name/Size DocType Version Access License
基于多特征的激光拼焊板焊缝质量检测技术研(3386KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[寇淼]'s Articles
Baidu academic
Similar articles in Baidu academic
[寇淼]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[寇淼]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.