Technology Research On Computer Vision Of Robot

The main study objective of computer vision is that how to realize human being's vision function using computers, namely to recognize and understand the three-dimensional (3-D) world from two-dimensional projective images. It is very significant to study computer vision, because it is not only beneficial to satisfy the application requirements of artificial intelligence, but also to be helpful to deeply understand and study the mechanism of human being's vision system.Camera calibration is the precondition and foundation of computer vision and image matching is the difficult and important part of computer vision. After read a large number of materials, a new accurate calibrating technique for camera is described. Mathematic model and mid-matrix could be founded by perspective transformation theory. So initial value iterate could be get by vector product. In condition of radial distortion and tangential distortion, restraint optimization equation can be created by penalty function. So nonlinear model parameter can be resolved by iteration. This camera calibration technique is accurate with high speed proved theoretically and practically.Here, two kind of matching approaches are studied, based on mixed image matching Algorithm (the multiresolution pagoda structure algorithm, the sufficiency function and projection algorithm were analyzed) and improved chaos algorithms. Both methods have good matching results in matching experiment. In order to decrease computation the multiresolution pagoda structure algorithm would be used to laminated image in mixed image matching Algorithm. When use sufficiency function rough matching position could be get in each layer. At last projection algorithm should be used to find exact position. A contrast experiment is conducted and the result demonstrates that the algorithm can reduce the matching time sharply. So the algorithm has been applied to the image matching system. For the algorithm of Chaos Optimization basic chaos optimization algorithm is used to search for approximate matching points firstly. Simultaneity, sufficiency function should be utilized to filter needless points. After the approximate matching points getting, the right matching points can be get with the second carrier wave by the approximate matching points. Finally, effectiveness is shown by a contrast experiment.