Design Of Detecting System For Six Degrees Of Freedom Based On Single-camera

Along with rapid development of image acquiring and processing technology, image measurement of six degrees of freedom(DOF) based on machine vision has been widely applied in some fields such as robotics, ocean engineering and flight simulator. This paper uses monocular vision and image processing technology for detecting six DOF of platform, and completes the building of a test model of the system, finally detects and proves the six DOF moving parameters of platform.In this paper, a method of measuring six DOF parameters with single camera between camera and opposite flat goal is presented. The ADSP-BF533 EZ-KIT Lite evaluation board, which is produced by ADI company, is used for designing of the system. Camera's internal parameters are calibrated in this paper. This system sends the image acquired by camera to SDRAM of the evaluation board and processes the image with Blackfin 533 digital signal processor. The six DOF moving parameters are computed by detecting method. And then the camera moves to next location according to the parameters, finally moves to the desired pose. The six DOF parameters in the whole detecting process are displayed on a monitor.Before computing the six DOF parameters, two circles with different diameter are marked on the target plant, and some characters are filled in great circle in order to distinguish the moving type. Then the collected image is processed by de-noise, image segmentation, region filled and etc. Characteristic elements which are used for calculating the six DOF are extracted, and the characters in the great circle are identified. Last six DOF parameters which are used for feedback movement of the platform are computed with method of this paper.At the end of this paper the system which is used for detecting six DOF is simulated, and the process of platform's movement and errors are analyzed. The results of the analysis prove that the system model is correct and effective. Results of the experiment show that the system can move intelligently without superfluous movement, and finally the camera can move to the desired place. Therefore, this system establishes basis for practical project application.