基于深度图像和点云边缘特征的典型零部件识别  

Recognition of Typical Components Based on Edge Features of Depth Image and Point Cloud

作　　者：张志佳[1,2] 魏信 周自强[2] 吴天舒 贾梦思 

机构地区：[1]沈阳工业大学软件学院,辽宁沈阳110020 [2]江苏省机电产品循环利用技术重点建设实验室,江苏常熟215500

出　　处：《信息与控制》2017年第3期358-364,共7页Information and Control

基　　金：江苏省机电产品循环利用技术重点建设实验室基金资助项目(KF1508);国家自然科学基金资助项目(61540069)

摘　　要：为解决自动化拆卸中零部件识别问题,提出了一种基于Kinect深度图像和点云边缘特征的典型零部件识别方法.首先利用非线性滤波优化算法对获取的深度图像进行处理获得优化后的目标点云并提出八邻域深度差算法提取点云边缘;然后利用随机抽样一致性(RANSAC)算法对分割后的点云边缘进行检测,并提取所定义的边缘特征以识别零部件.该方法能够实现对典型零部件的识别,实验结果验证了方法的有效性.To solve the problem of component identification in automatic disassembly,a typical components identification method is proposed based on the depth image of Kinect and point cloud features. Firstly,a nonlinear filtering algorithm is used to process the acquired depth image and obtain the optimized point cloud target,and an eight neighborhood depth difference algorithm is proposed to extract the point cloud edge. The segmented point cloud edge is then detected using the RANSAC algorithm,and edge features are extracted to identify components. This method can identify typical components,and experimental results verify the effectiveness of the proposed method.

关 键 词：Kinect传感器点云数据边缘提取随机抽样一致性算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]