The Improvement On Real-time Control Of Microassembly Robot's Visual Servo

Microassembly robot is the combination of robotic assembly and micromanipulation, and microscopic vision servo is the main controlling method for the microassembly robot. This dissertation mainly aims to the application of microscopic system and several topics have been discussed such as stereo microscopic vision framwork, micro object detection and recognition. The desighed microassembly system is sponsored by National 863 Plan. The dissertation gets a deep research on the structure of the microassembly robot.The thesis is focused on semiautomatic video object segmentation and tracking. Based on the in-depth research of active contours models which have been studied and applied as image segmentation methods during the past ten years, a framework for semiautomatic segmentation and tracking using Snake is presented. A new snake model—VSnakes is studied and two improvements of the definition of contour energy are proposed. One of them makes the points which control the detection of the contour more evenly, and the other smooths the candidate contour. And the dissertation compares the different three models.The detection of micromanipulator's movements in the sequence is the basis of the realization of the visual control. A method of contour prediction is presented which is based on the path coherenceand limited contour expand. Applying the novel snake algorithm to actual contour extraction leads to satisfying result, because it can not only simplify snake's initialization to decrease its sensitivity to its initial position, but also converge to the global minimal energy quickly.In the end of the dissertation, we introduce the Open Source Computer Vision Library (OpenCV) which is extensively applied in the image processing and computer vision. And we make a emphasis on the CamShift algorithm which is used for the object tracking. All the algorithm for image processing is based on OpenCV.