Optimal multilevel thresholding based on Tsallis entropy and half-life constant PSO for improved image segmentation

This paper proposes a multilevel thresholding technique for image segmentation, using an image enhancing technique and an improvised algorithm called Half-life Constant Particle Swarm Optimization (HCPSO). The multilevel thresholding is based on Tsallis entropy. This is done so that the thresholds are optimized and thus a better segmentation can be obtained. This algorithm has been tested on various standard test images. This method provides better results in terms of increased objective values and PSNR values when compared with the basic PSO and other optimization algorithms such as Genetic Algorithm (GA) and Bacterial Foraging (BF) algorithm.