论文部分内容阅读
机器人的强耦合和强非线性使得6自由度视觉定位成为机器人视觉伺服领域的一个热点和难点问题。提出了一种基于自适应kalman滤波的机器人无标定6自由度视觉定位方法。首先利用图像的全局特征描述子—图像矩设计了一组图像矩组并以它的变化来表征摄像机与目标之间的相对平动与转动。在不标定摄像机与机器人坐标变换关系的情况下,应用自适应kalman滤波器来在线估计图像雅可比矩阵,并在此基础上设计视觉控制律从而计算出机器人的运动控制量,最后在MATLAB环境下建立了眼固定机器人无标定6自由度视觉定位Simulink模型,实现了机器人6自由度视觉定位。仿真实验结果表明,在噪声的统计特性不完全已知的情况下,所设计的自适应kalman滤波器能使6自由度机器人到达期望的位置,且定位精度高。
The strong coupling and strong non-linearity of robot make 6-DOF visual positioning become a hot and difficult issue in the field of robot vision servo. A 6-DOF visual positioning method based on adaptive kalman filtering for robot is proposed. Firstly, a group of image moments is designed by using the global feature descriptor-image moment of the image and its change is used to characterize the relative translation and rotation between the camera and the target. In the case of not calibrating the relationship between camera and robot coordinate transformation, an adaptive kalman filter is used to estimate the image Jacobian matrix on the basis of which the visual control law is designed to calculate the robot’s motion control. Finally, in the MATLAB environment The 6-DOF Vision-Positioning Simulink model of eye-fixed robot is established, which realizes the 6-DOF visual positioning of the robot. Simulation results show that the adaptive Kalman filter can make the 6-DOF robot reach the desired position and the positioning accuracy is high under the condition that the statistical properties of noise are not completely known.