Research of image segmentation algorithm based on wavelet transform

In this paper, image segmentation algorithm based on wavelet transform is studied, and AISWT model is created. The algorithm this paper presented obtains the initial parameters through histogram-based algorithm, then the final parameters by EM algorithm, and finally completes image segmentation by ML. Experimental results show that this algorithm can effectively prevent over-segmentation, greatly reduce the segmentation regions, obtaining ideal segmentation results; and comparing to the traditional FGM and FGGD algorithms, effectively reduce the computation time, greatly improving the segmentation efficiency.