空间绳系机器人单目视觉伺服控制系统

为解决空间绳系机器人近距离交会对接中的目标测量问题，设计了一种单目视觉伺服控制系统。该系统针对小模板在大场景中匹配时存在鲁棒性差、耗时长等缺点，采用了一种新的模板匹配算法。匹配模板借鉴FAST算法的空心圆环结构，使得算法具有旋转不变性，也减小了因背景变化引起的累积误差。同时匹配准则采用灰度值+梯度值的复合算子，具有较强的抗噪声和抑制光照变化的能力。为避免因错误匹配和遮挡等引起的跟踪失败,还设计了动态模板更新策略；最后结合最小二乘综合位置预测方法和PD控制器，实现了对复杂场景中的非合作目标的实时跟踪逼近。

Monocular Visual Servo Control System Scheme for Tethered Space Robot

In order to solve the problem of non-cooperative target recognition during short-distance rendezvous for the TSR, a prototype framework of servo control system using monocular-CMOS is proposed in this paper. Aiming at the problems resulted from small template matching with large scene, such as low robust and long time-consumption, a novel template matching algorithm is used in the system. First, the novel matching algorithm uses a hollow annulus structure according to FAST algorithm to make the method be rotation-invariant and help to decrease the accumulative error caused by the changes of background. The matching criteria is composed of gray and gradient differences between template and object image, which restrain the impact of noise and illumination. Then, a dynamic template update strategy is designed to avoid tracking failures brought about by wrong matching or occlusion. Finally, the algorithm combines the least square integrated predictor with the PD controller, thus realizing real-time tracking and approaching non-cooperative target under complex circumstances.