基于主成分分析的结构光条纹中心提取方法  

Centerline Extraction of Structured Light Stripe Based on Principal Component Analysis

作　　者：蔡怀宇[1] 冯召东[1] 黄战华[1] 

机构地区：[1]天津大学精密仪器与光电子工程学院光电信息技术教育部重点实验室,天津300072

出　　处：《中国激光》2015年第3期270-275,共6页Chinese Journal of Lasers

基　　金：国家科技支撑计划(2007BAF14B03)

摘　　要：条纹中心提取是线结构光表面形貌测量的关键问题,提出一种基于主成分分析方法实现结构光条纹中心提取的方法。利用大津阈值法提取图像的感兴趣区域(ROI),在经两次高斯卷积得到条纹的梯度分布的基础上确定条纹中心的初始位置,利用主成分分析来确定这些点上法线方向;以初始位置为基准点,对条纹的灰度分布函数在法线方向进行二阶泰勒展开求得条纹中心的精确位置。实验结果表明,该算法具有速度快、精度高等特点,相对Steger算法,均方误差(MSE)小于0.003 pixel,速度提高了近3倍。该方法能够实现结构光条纹中心线的快速高精度提取,为结构光视觉检测的实时应用奠定了基础。Centerline extraction of structured light stripe is the key step of topography measurement. The centerline extraction of structured light stripe based on the principal component analysis is proposed. Otsu′s method is used to extract the region of interest（ROI） area of image. After twice Gaussian convolution, the gradient distribution of the ROI area is got, then the rough position of the stripe center is obtained based on the gradient distribution. Using principal component analysis the normal vector is obtained. Based on the rough position of the stripe center, the center of structured light stripe can be found in normal directions with second order Taylor series expansion of grayscale distribution function. The experimental results show that the algorithm has higher speed and precision characteristics. Compared with Steger algorithm, the mean square error（MSE） is less than 0.003 pixel, and it is nearly three times as fast as Steger algorithm. This method can realize the rapid and high precision extraction of structured light stripe, and lays a foundation of real- time application for structure light vision detecting.

关 键 词：测量 形貌检测 线结构光 中心提取 主成分分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]