| In image processing of mobile laser-aiming system, the effective edge detection impacts on the positioning accuracy of the center directly. The edge which is widely used in image processing is one of the most significant radical characteristics of an image. There are lots of analyses on this subject. Moreover, a number of scientists concentrate on this problem and obtain many edge detection filters and algorithms on successful image distinction.Both conventional and original algorithms which are on the basis of differential and their characteristics are analyzed in this paper.In this paper, the definition of conventional image edge is extended to morphology edge. The morphological operators can withstand the noise and have many superior functions including good performance of location and image details keeping. It is proved that these operators are capable of detecting the image edge more precisely.On the base of a systematical analysis about the theory of mathematical morphology and neural network, in the paper we present a method of mathematical morphology image processing based on radial basis function neural network, which makes some improvement on original morphological image processing. We present the neural network model of grayscale morphologieal operation and educe learning arithmetic to gear the numerical value of structuring element. According to the processing result, we draw the conclusion that the new method is better than the original mathematical morphological method.Spatial moment-based edge operator is widely used because of the advantage in the higher measuring precision and the measuring data being not influenced by noises. In this paper, basic on mathematics morphologic arithmetic was combined to neural network to orientate the edge position and eliminate the noise. Then, the edge was quickly located in subpixel precision by spatial moment method. Afterwards, the least squares fitting was utilized to orientate the circular target center. |
PDF Full Text Request