Font Size: a A A

Study On Intelligent Image Segmentation

Posted on:2006-06-22Degree:MasterType:Thesis
Country:ChinaCandidate:C ZhangFull Text:PDF
GTID:2168360155954890Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Image is an important information source of human intelligence. It is the primary media of communication. Along with the concept of information highway and digital earth being proposed, the need of image processing technique increase steadily. At the same time, the developments of computer provide a platform for image processing technique.Image segmentation is a vital image processing technique and the main problem of computer vision. It is also a key step from image processing to image analysis and holds significant place in image engineering. On one side, image segmentation is the basic of target expression and affects feature measure greatly. On the other side, original image can be translated into more abstract and more compact format by image segmentation and some factor based on segmentation, such as target expression, feature extraction, parameter measure, and so on. In image applications, the target extraction and measure can't be done without image segmentation. As the complexity of image information, some problems about fragmentariness, non-accuracy and non-structure will be in the image processing. Intelligent information processing methods, which used in image segmentation will promote the quality of whole image processing system, including real time, automatization and intelligentize.Based on the research of classic image segmentation methods, intelligent information processing methods are used in image segmentation. In this paper, it mainly include analysis of classic image segment methods, research of image segmentation based on chaos genetic algorithm and the segmentation based on some of fuzzy theory. The main research outcomes are as follows:1. Based on the research of evolutionary algorithms, a quantum-inspired evolutionary programming is proposed and simulations show that QEP is better than conventional EP, because of its rapid convergence and global search capability. It can be effectively used in the optimal design of FIR filter.2. Based on the analysis of chaos genetic algorithm, alterable scale...
Keywords/Search Tags:Intelligent image processing, Image segmentation, Evolutionary algorithms, Local fuzzy entropy, Transition region
PDF Full Text Request
Related items