Sonar image enhancement based on particle swarm optimization

Sonar images have some characteristics such as low resolution, strong echo disturbance, small object regions and obscure object edges, so the satisfied results can´t be obtained by the global enhancement algorithm. An adaptive local enhancement algorithm is proposed in this paper, and the evaluation function is defined by edge numbers, edge intensity and the entropy. Due to the high complexity of the algorithm proposed, the particle swarm optimization algorithm is used to search the optimal parameters for the best enhancement. The presented algorithm was compared with other automatic contrast enhancement techniques-histogram equalization and iterative enhancement, the results obtained show the superiority of our method in the sonar images. It can not only improve the vision effect obviously, but also enhance the image contrast and hold the details simultaneously. The qualities of images treated are more suitable for the detection algorithm.