Research On Underwater Acoustic Image Thresholding Segmentation And Intelligent Optimization Algorithm

This paper mainly studies the underwater acoustic image thresholding segmentation and the application of the intelligent optimization algorithm in underwater acoustic image segmentation. There are the underwater acoustic image thresholding segmentation based on the Otsu and based on maximum entropy, the application of the intelligent optimization algorithm in underwater acoustic image segmentation.First, this paper studies the underwater acoustic image thresholding segmentation based on the Otsu,and researches on both the classical method and improved the method of the Otsu; combines the characteristics of underwater acoustic images, extends this method to the double-thresholding segmentation, and a relaxing background of the within-class distance method of Otsu is proposed, which is more suitable for the underwater acoustic image segmentation.Second, this paper studies the underwater acoustic image thresholding segmentation based on the maximum entropy method. One-dimensional entropy, two-dimensional entropy and the improved two-dimensional entropy method are studied. For the characteristics of underwater acoustic images, the image segmentation based on the maximum entropy of the amended gray-gradient 2-D histogram method is proposed. This method can not only be free to choose the dark areas or bright areas, but also to get better segmentation results in noisy circumstances, as part of the interference pixels is wiped off.Finally, this paper works for the intelligent optimization algorithm method. It is used to enhigh the speed of the complexity and slower speed thresholding segmentation method of the underwater acoustic image segmentation.