摘要
本文提出了一种应用小波变换对固体表面位移场进行测量的新方法 .在图像分析时 ,用各点属于不同尺度的小波变换系数来表征该点周围的子区 .将试件表面位移前后的两幅数字图像进行小波变换 ,通过变形前后两幅图像小波变换系数之间的相互匹配 ,使位移前后两幅图中的子区对应起来 ,从而确定图像的位移场 .本文应用Haar小波变换进行了计算机模拟实验和实物位移实验 .引入亚像素技术 ,获得了 0 .0
This paper presents a new method applied to measure displacement field using wavelets transform(WT). Using digital image processing system, we acquired the digital images of the specimen before and after deformation. The WT of the acquired images was analyzed by two-dimensional Haar wavelets, so we could find the local space-spectrum information of the images. Then, we compared the WT of the images before deformation with that after deformation, and could get the relative displacement of the point related to two images. In the experiment, the displacement of specimen was carried out by a fine micron system. The experimental results show the advantages of WT in the displacement measurement. By using the technique, the sensitivity of this method is 0.02 pixel.
出处
《实验力学》
CSCD
北大核心
2002年第2期165-169,共5页
Journal of Experimental Mechanics
基金
国家自然科学基金资助项目 ( 1 9972 0 4 7)
关键词
位移场
小波变换
数字图像
亚像素
固体实验力学
displacement field
wavelet transform
strain
digital image processing
sub-pixel