Image-based manipulator visual servoing using the Kalman Filter

Abstract

This paper presents a manipulator visual servoing algorithm with the Kalman filter. The objective is to control the relative position and orientation between a robot's gripper and an object using a single camera mounted on the manipulator. The feature-based visual servoing control algorithm is used, which continually updates the pose of the robot relative to the object by using the differential changes in image plane. But the acquired image from a camera often has several noise elements such as image signal spatial quantization error, lens distortion, pixel error and etc. In order to reduce the measurement errors, a Kalman Filter was adopted to estimate the motion states of a moving object. Its real implementation was performed for its effectiveness ofthe presented algorithm.