A fast SAR image segmentation algorithm based on particle swarm optimization and grey entropy

To speed up the segmentation procedure and improve the segmentation quality of SAR image, the paper suggests a PSOGE algorithm, which is based on particle swarm optimization and grey entropy. In the algorithm, after a filtered image and a gradient image are deduced from the origin SAR image respectively, their grey-level co-occurrence matrix is constructed. On the basis of the matrix, a grey entropy based fitness function is designed for Particle Swarm Optimization (PSO). And then, after several groups of thresholds and their moving speeds are acquired by the initialization of the particle swarm, all of the particles change positions iteratively and concurrently, and approach to the best threshold, depending on two types of experiences: personal best and global best experiences. The experimental results indicate that the algorithm not only shortens the segmenting time obviously, but also ignores the disturbance of inherent speckle in SAR image.