Support Vector Machine And In The Application Of Image Enhancement

Statistical Learning Theory is a small-sample statistics by Vapnik, Which concerns mainly the statistic principles when samples are limited, Especially the properties of learning procedure in such cases. Statistical Learning Theory provides us a new framework for the general learning problem, and a novel powerful learning method calls Support Vector Machine or SVM, which can solve small-sample learning problems better. It is believed that the study of Support Vector Machine is becoming a new hot area in the field of machine learning. With the wide range of support vector machines studies, there have been many improvements in the Support Vector Machine algorithm, such as: Least Squares Support Vector Machine algorithm, Incremental Learning methods, they mainly to reduce traditional Support Vector Machines computational complexity and solve traditional Support Vector Machine in the large-scale training samples slower problems.In this dissertation, Support Vector Machine algorithm and the Support Vector Machine in image processing applications are studied form the support vector machine algorithm and the application point of view, the main work of this thesis is as follows:First, Compared with the classical Support Vector Machines, the Least Square Support Vector Machines lose the sparseness, which would influence the efficiency of re-learning. This paper presents an improved incremental Least Squares Support Vector Machine learning methods ,in the solution of the sparsity using adaptive pruning method, according to the initial class- ification be to set performance pruning threshold and increase the size of samples and the simulation results show that the algorithm feasible.Noise pollution was one of the images through sharpening processing, image sharpening will reflect the noise, spatial features can be described as isolated and pixel value smaller points. Through the noise characteristics of this paper using the Least Squares Support Vector Machine sharpening image denoising handling, noise corresponding value is set to zero, in the final strengthen neglected, the entire processing time reduction. In practical application, the system access to the original image is not generally perfect, image quality is not high, so it is necessary to enhance the image processing with a view to increasing the quality requirements. This paper presents a weighted local adaptive contrast enhancement algorithms, it took into account the details in the image, at the same time enhancing the noise suppression, after filtering the target image noise processed by Least Squares Support Vector Machine , algorithm also takes into the image different contrast in the different regions, it is enhanced less for a high contrast region and low contrast region is enhanced more, so that the result of image is softer and improve the image of the visual effects. Through experiments show that the method used in image enhancement is a feasible method.Finally, to summary Least Squares Support Vector Machine algorithm improvement and used in image denoising and image enhancement research, and has carried on the forecast to the research work.