Research And Design Of Bionic Imaging Platform Based On Two-axis Linear Drive

This paper intends to present a stable imaging platform for the visual system of autonomous robot, aiming at improving the non-optimal quality of image which is caused by tremor, vibration, slope alteration as well as rapid movement of the target during the moving process in intricate circumstances. According to the mechanism of human optical system, a bionic imaging platform driven by multi-axis linear motors is given, in which the function of swift tracing is integrated.This paper gives both the general design and the construction of the visual platform at the beginning, and then it expands to describe three important processing stages in details including data input stage(pattern recognition and target tracking),intermediate stage(attitude adjustment strategy) as well as motion output stage(motion control of linear motor).During the phase of pattern recognition and target tracking, some preprocessing methods are used to achieve high-quality images, and then the characters of target can be extracted from those images by image sharpening, image segmentation and edge tracing. After the procedures mentioned above, we utilize kalman filter to predict and trace target and furthermore verify the validity of these processes in matlab.According to the result of recognition and prediction, the target positions can be transferred as inputs to the attitude adjustment strategy, generating the relations between target position and motor displacement by kinematics models of the platform. And finally, this paper analyzes the motion control process of linear motors and completes the simulation of velocity loop and position loop.Moreover, we set up some interference to verify the performance of this visual platform. From the simulation, it is obvious to see that a good tracking effect and capacities of anti-interference can be available due to the high response of linear motor.This research also contributes fundamental approaches to the framework designs of the visual system of autonomous robot.