Particle swarm optimization method for very low bit rate segmented image compression

In this paper a new very low bit rate segmented image compression method based on particle swarm optimization (PSO) is proposed. In the segmented image coding schemes, the given image is firstly segmented and, then, the contours and textures of regions are extracted and encoded. In our method seeded region growing method (SRG) is used to segment the image. The proposed algorithm randomly initializes each particle in the swarm to contain K seed points and then seeded region growing algorithm is applied to each particle. The segmented image of each particle is encoded, the contours are encoded by using chain codes method and the textures are encoded by using the gray means. After that the PSO algorithm is applied to find the best positions for seed points that produce the best segmentation, which gives us the best compression ratio. So, in the proposed method, the objective function of the PSO algorithm is the compression ratio (CR).