The Application Of Catastrophe Theory In Digital Image Processing

With the development of science and technology, image processing has become an indispensible part of the life. In the field of image processing, image feature extraction is more important, it can be characterized by many of the image, using the feature point of the image, we can perform a variety of operation, such as image feature points match, image Mosaic, etc.There are a lot of algorithms for image feature point, but each kind of algorithm has its advantages and disadvantages, not only use an algorithm to describe the image, we requires more algorithms of image feature points. Catastrophe theory is based on the stability of the system structure, to deal with the discontinuous phenomenon in the real world which is contradiction, it's a branch of mathematics which in the study of the non-linear science research discontinuous phenomenon. In this paper, we use the cusp model of the catastrophe theory, to get the catastrophe point of the image, and then use the catastrophe point to match and splice, get the images seamless. This program is achieved with Visual C++6.0.This paper mainly includes the following aspects:First, a brief introduction of the development of catastrophe theory and development situation. Second, introduces the digital image processing technology, the image mosaic, and analyze it. Third, introduces image scale space, and how to get the catastrophe points of the image using the cusp model of the catastrophe theory. Fourth, introduces the general flow of image mosaic technique, Harris corner detection algorithm and Sift algorithm, and combined with the Mosaic of cylindrical coordinate system, and then, the experimental results of three kinds of algorithm are analyzed, the experimental results show that this algorithm in detecting and matching rate feature points are taking a lead. Finally, using the program we generated a panorama.