Aquatic Image Segmentation Method Based on HS-PCNN for Automatic Operation Boat in Crab Farming

Abstract

Crab farming mainly adopts the pattern of aquatic feeding, which is a hard work for farmers. To reduce the labor intensity and production costs for farmers, it is great significance to develop an automatic aquatic-cleaning boat with visual navigation. In visual navigation system, image segmentation is a difficult problem. In this paper, we combine the advantages of pulse coupled neural network in image segmentation with the global optimization characteristic of harmony search algorithm, an image segmentation algorithm of optimized pulse coupled neural network based on harmony search (HS-PCNN) is proposed. In order to improve the operating efficiency and segmentation accuracy of PCNN, this new algorithm can optimize the weighted combination of PCNN maximum Shannon entropy and minimum cross entropy by harmony search (HS), and evaluate the optimization effect of parameters by using yield function. Experimental results show that the proposed method can provide a more effective method for the aquatic image segmentation in crab pond.