改进骨刺去除算法的结构光条纹中心检测

Detection of structured light strip center line based on improved skeletal removal algorithm

在基于线结构光的三维视觉检测系统中,激光条纹中心线的提取精度直接影响了整个系统的精度。针对形态学骨架法提取激光条纹中心线时会存在大量骨刺导致条纹中心线粗糙且不连通,传统去骨刺算法无法快速精确的去除骨刺等问题,提出一种改进的基于骨架分枝起点定位去骨刺算法。该方法可以利用分枝起点和交叉点的定位标记准确分离骨架主干和骨架分枝,设定合理阈值去除骨刺,并通过标记点快速恢复骨架主干,从而减少了骨刺去除及骨架主干恢复的时间。实验结果表明：该算法能够有效去除骨刺,保留完整平滑、具有单相连通性的激光条纹中心线,提高了检测精度和检测效率,满足三维表面的快速精确检测。

In the three-dimensional visual detection system based on the line structured light,the accuracy of the entire system is affected by the extraction accuracy of the laser stripe centerline directly. There would be a large number of bone spurs,leading to rough and non-connected centerlines when the morphological skeleton method is adopted to extract the laser centerline. Bone spurs can＇t be removed quickly and accurately by the traditional deburring algorithm and am improved spur-removed algorithm based on skeleton branch starting point. The improved method could accurately separate skeleton skeletons and skeleton branches with the positioning marks of branching starting points and intersections. A reasonable threshold is set to remove bone spurs and quickly recover skeletons through marker points,thereby sprout removal and skeleton backbone recovery time are reduced. The experimental results show that it could effectively remove bone spurs and preserve the integrity and smoothness of the laser stripe center line with single-phase connectivity,improve the detection accuracy and detection efficiency,and satisfy the fast three-dimensional surface accurate detection.