自适应复杂光条纹中心提取方法研究  

Research on Self-adptive Extraction Method of Central Points in Complex Light Stripe

作　　者：岳龙[1] 郭喜庆[1] 余锦[1] 

机构地区：[1]中国科学院光电研究院,北京100094

出　　处：《科学技术与工程》2016年第14期236-239,244,共5页Science Technology and Engineering

基　　金：中国科学院创新交叉与合作团队项目资助

摘　　要：激光条纹中心提取作为线结构光测量中最关键的环节,其提取方法的准确性及适应性直接影响着测量系统的整体性能。针对环境光干扰、待测物表面反射率差异较大引起的宽度变化的复杂激光条纹,在Steger方法的基础上,提出了一种自适应复杂光条纹中心提取算法。该方法首先根据线激光宽度选择固定的尺度因子,对图像进行高斯核卷积,得到各点Hessian矩阵;然后求得对应的法向宽度,并以此来选择对应的二阶法向偏导数的阈值,再结合一阶法向偏导为零计算出条纹在不同宽度处的中心点。实验结果表明该算法可以实现对于宽度变化的复杂光条纹中心的亚像素提取。The extraction of the laser strip center is the most critical part of the line-structured measurement, and its aceuration and adaptability directly affect the performance of the measurement system. For the variable-width complex laser strip caused by the ambient light interference and the different surface reflectivity of the object, a self- adptive extraction methods is investigated based on Stegcr＇ s algorism. In this method, firstly a fixed scale according was choosed to the width of laser strip to convolve the image with gaussian kernels and got the Hessian matrix of ev- ery point, then calculated the normal width to choose the corresponding threshold of second derivative along normal direction, combining the character that the first directional directive along normal direction takes zero to get the center points of laser strip at different width. The results show that the algorithm can achieve the subpixel center extraction of laser strip with variable width.

关 键 词：图像处理 激光条纹 Steger llcssian矩阵 三维视觉测量 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]