Image Denoising And Image Segmentation Based On Improved Particle Swarm Optimization Algorithm

With the development of science and technology, digital image processing technology has been used more and more widely. Due to the diversity and complexity of image information, there are some problems like imperfection, uncertainty and modeling difficulties in image processing field. Therefore, depend on the characteristics like easiness of understanding and convenience of achievement, intelligent optimization algorithms have been widely used in image processing field. Although some intelligent optimization algorithms like particle swarm optimization (PSO) have achieved good performance in image processing field, there are still some issues like image denoising and image segmentation.On the research of fundamental theory of PSO algorithm, an improved particle swarm optimization algorithm is proposed in this dissertation. This algorithm is based on inertia weight adjustment, and is used in the issues of image denoising and image segmentation. Main contributions of the dissertation are shown as following:1. For improving the shortcoming about the balance of global search and local search of basic PSO, an improved PSO algorithm with binary dynamic adaptive inertia weight adjustment is proposed in this dissertation. Use the pre-adjustment of finding correct direction and delicate adjustment of references about evolution speed and concentration, to improve the performance about the balance of global search and local search.2. An image denoising method based on the improved PSO is proposed. Because of the noise interference on the image, finding threshold of wavelet coefficient becomes very difficult, and the wavelet transform cannot achieve good performance on image denoising. This dissertation uses the improved PSO algorithm to find the threshold value between noise coefficient and image coefficient to achieve image denoising.3. An image segmentation method based on the improved PSO is proposed. Because of the anti-noise performance of two-dimensional gray histogram, choose particle swarm on the area between object and background and use the improved PSO to find the best threshold value which makes two-dimensional reach to the maximum, thus achieve image segmentation.