Image of Plant Disease Segmentation Model Based on Pulse Coupled Neural Network with Shuffle Frog Leap Algorithm

Abstract

Image segmentation is a key step in feature extraction and disease recognition of plant diseases images. To avoid the subjectivity of using traditional PCNN (pulse-coupled neural network) to segment plant disease image, a new image segmentation model (SFLA-PCNN) is proposed in this paper to get the parameters configuration of PCNN. The weighted sum of cross entropy and compactness degree of image segmentation is chosen as fitness function of shuffled frog leap algorithm to optimize the parameters PCNN, which could improve the performance of PCNN. After 100 times local iteration and 1500 times global iteration, we get the best parameter configure. The extensive tests prove that SFLA-PCNN model could be used to extract the lesion from the background effectively, which could provide a foundation for following disease diagnose.