基于深度相机的船舶小构件打磨装置视觉识别算法设计

基于深度相机，针对船舶小构件智能打磨装置的视觉识别系统进行算法模块设计，确定视觉识别系统的模块框架与模块功能，并基于OpenCV库与PCL库，利用C++程序语言对模块功能进行算法实现。同时搭建算法试验平台，对模块算法进行验证，完成包括相机位姿估计与坐标变换、工件点云分割、工件质心提取、工件边缘提取在内的算法功能实现，为船舶小构件智能打磨装备提供准确全面的视觉数据，保证智能打磨装备的上料及打磨作业顺利执行。

Visual Identification Algorithm Design of Grinding Device for Ship Small Component Based on Depth Camera

Based on the depth camera, the algorithm module design is made in view of the visual identification system of intelligent grinding device for ship small component, the module framework and module functions of visual identification system are determined, and based on OpenCV library and PCL library, the algorithm to module functions is realized with C++ program language. At the same time, the algorithm test platform is established to verify the module algorithm, to complete the algorithm function realization including camera pose estimation and coordinate transformation, workpiece point cloud segmentation, workpiece centroid extraction, workpiece edge extraction, and to provide the accurate and comprehensive visual data for the intelligent grinding device for ship small component, which guarantees that the loading and grinding operation of intelligent grinding device can be performed smoothly.