Regular image fast matching based on adaptive genetic algorithm

In order to improve the speed and precision of image match, for regular image (cross curve), a fast image matching algorithm using image project feature based on adaptive genetic algorithm is presented. Design matching rule with taking the image feature into account, so this algorithm suits for some noise, scaling, revolution and deformed images. According to the property that low resolution image can still hold the majority of features of high resolution image, the image is transformed into low resolution by 2X2 sampling by which the computation capacity can be reduced to 1/4. Image matching is a random search problem in 2-D image space. Design an adaptive genetic algorithm, which vary the probabilities of crossover and mutation in response to the relationship of fitness values in evolution course. This makes it to obtain global optimum point quickly and accurately. The experiments demonstrate that this algorithm increases the matching speed and precision efficiently, and has high anti-noise ability and robust.