| Image segmentation is one of the most important and difficult tasks in image processing, therefore it has received the people to take seriously. With the development of modern computer technology, a lot of efficient theories and methods of calculating on image segmentation bring up. Rough set theory is one of them, it is an effective tool dealing with imprecise, vague description of the targets of mathematics. At present, rough set theory has been widely applied in image processing, but the application of variable precision rough set theory in image segmentation is rare.This paper presents two new gray-degree image segmentation algorithms, and the threshold of image segmentation method has been studied. One is image segmentation algorithm based on particle swarm optimization (PSO) and entropy of variable precision rough set based on boundary region; the other one is image segmentation algorithm using genetic algorithm and wavelet transform by variable precision rough set standards for the evaluation of gray-degree.The first algorithm uses variable precision rough set based on boundary areas as evaluation function, through PSO searching to find the biggest approximate classification quality and the corresponding interval ofβ. The gray-degree value corresponds to the upper boundary of β is the optimal segmentation threshold. This approach reduces sensitivity of the algorithm on the sub-piece size, and to some extent reduces the running time of the algorithm. The second one concentrates on cutting down the effect of noise and reducing the running time of the algorithm. Firstly, we use wavelet transformation to decompose the image. Secondly, this approach uses the genetic algorithm and the evaluation function with variable precision rough set to find the best threshold for the image. Finally, the obtained image is reorganized by wavelet transformation. Two algorithms are test by MATLAB simulation, and the effectiveness and feasibility of algorithms are shown in these tests. |
PDF Full Text Request