论文部分内容阅读
为解决大范围偏差的控制问题 ,将期望图象按给定的角度间隔旋转 ,离线生成一系列子期望图象。比较实时采集图象与期望子图象间的差异程度可获取目标绕重心的旋转运动参数。结合图象求重心方法给出的平动参数 ,实现了在大范围偏差时迅速将摄象机调整到期望位姿。在期望位姿附近结合直接图象反馈方式 ,实现了基于图象差的平面大范围视觉伺服控制。
In order to solve the control problem of a wide range of deviation, the desired image is rotated at a given angular interval to generate a series of sub-desired images offline. Comparing the degree of difference between the real-time acquired image and the desired sub-image can obtain the rotational motion parameters of the target around the center of gravity. Combining the translational parameters given by the image center of gravity method, the camera can be quickly adjusted to the desired pose in a wide range of deviations. In the vicinity of the expected pose, combined with the direct image feedback, a large-area planar servo control based on image difference is realized.