节点文献

像素法裂纹参数测量技术研究

【作者】 曹梅丽

【导师】 王振林;

【作者基本信息】 浙江大学 , 固体力学, 2006, 硕士

【摘要】 本文针对工程结构开裂评估的需求,运用光测力学原理与近年间迅速发展起来的图像处理新技术,研究开发了一套“像素法”裂纹长度和宽度参数测定的新系统。论文前期,参与完成了由光学镜头—CCD—图像卡组成的非接触裂纹检测系统的硬件制作。之后作者独立完成了裂纹长度与宽度计算计量专用软件的开发。其中图像处理模块包含了滤波去噪、二值化、图像分割、轮廓提取与细化等子模块。参数计算块中,运用数字散斑相关原理进行像素标定与裂纹长度的像素搜索计算;进而用亚像素原理解决了裂纹宽度测量时整像素分辨率不够的技术难点。在论文的后阶段,设计了实验室中进行的两个层次的考核和标定试验,并对应用中必然遇到的物面光照度、图像对比度、测量距离(物距)和焦距的改变,对试验样本测量精度的规律作出了初步分析,为新方法的进一步完善提供了有价值的工作基础。最后对进一步拓展研究提出了自己的分析与建议。 论文提交的新方法具有非接触和快速自动测量的特点,精度满足工程要求,特别在现场和危险环境下的裂纹参数测量时具有特色和优势。可以预测,经过以产品化为目的的深度开发(软件界面、辅助显示、报告打印功能和提升硬件指标等)后,并辅以更多的实验考核,完全有可能制作成一种新型的仪器,在众多工程检测领域推广应用。

【Abstract】 According to the need in evaluation of the crack in structure, the author of this paper developed a new system based of photoelastic mechanics and digital image proccessing, which is specialy used to measure the length and width of crack. In the prophase the author paticpated the mading of the hardware of the system made of optial lens 、 CCD、 image card and computer. Then the software which composed the filtering of noise、 the product of binary image、 image segmentation、 edge detector、 and thinning is finished. Also the image element is calibrated accoring to the digital speckle correlation method (DSCM)oAnd then the length divided by the interal pixel is hunted and caculated, the width divided by subpixel is also done. Finaly many experiments are completed in the lab in diffrient conditions , which contains illuminance of crack 、 contrast of image 、 distance of shotting and focal length, to review the accuracy of the new system and to analys the rule of th effection of the accuracy .This system can non-contacting and fast measure , and the measuring accuracy satisfied the demand of engineering.It shows its dominance and characteristic when it is use to measure the crack which is in dangerous or high temperature field . It can be forecasted that this system can be made to a new equipment applied in the field of detection widespreadly, when the system is improved .

  • 【网络出版投稿人】 浙江大学
  • 【网络出版年期】2006年 12期
  • 【分类号】O348
  • 【被引频次】12
  • 【下载频次】216
节点文献中: 

本文链接的文献网络图示:

本文的引文网络