全息干涉条纹图的计算机自动处理与识别

本文应用数字图象处理与识别技术在IBM-PC/AT微处理计算机图象处理系统上对复杂的全息干涉条纹图进行了自动分析,文中采用二维灰度值检测的方法获取条纹的准峰或谷值二值图形,然后通过细化得到条纹中心,提出了用记忆跟踪算法自动去除毛刺和噪声线段.在用光标进行人机对话标定零级条纹后,可自动识别出其它整数条纹级数.分数级条纹级数可在整数级条纹数据之间进行线性或非线性插值得到.最后给出了全息条纹图分析的实例.

Automatic Holographic Interferometric Fringe Processing and Recognition with a Computer Aided Image-Processing System

An IBM-PC/AT image processing system, based on digital image processing techniques, is used for, auto-analysing a complicated holographic interferometric fringe pattern. Two-dimensional grey value detecting method is applied to get binary pattern of fringes. The fringe center can be obtained through thinning the pattern. In order to remove noise lines and fringe burrs automatically, it is proposed to used the memory tracing algorithm. Integer fringe orders can be recognized automatically after zero-order fringes are determined manually. Then the fraction fringe order can be obtained by the method of interpolation. Finally, an example of pattern analysing is presented.