Support Vector Machine Classification And Face Detection Applications

Support Vector Machines (SVM) are a kind of novel machine learning methods. It can solve small-sample learning problems better by using Experiential Risk Minimization in place of Structural Risk Minimination. Moreover, this theory can change the problem in non-linearity space to that in the linearity space in order to reduce the algorithm complexity by using the kernel function idea. SVM have become the hotspot of machine learning because of their excellent learning performance. They also have successful applications in many fields, such as: face detection, handwriting digit recognition, text auto-categorization, etc. But as a new technique, SVM also have many shortcomings that need to be researched, including: the adaptive kernel and parameter selection, sensitive to noise, have a limitation in the scale of training set, the shortcomings of training methods, incremental learning, and the combination with the prior knowledge, etc. The applications in many fields are limited because of these problems. In this paper, some of above problems are probed into the application in image processing area. This sisertation mainly focuses on some applications of SVM in image processing area after studing the theory of SVM.The work including:(1) Optimization theory and the essence of SVM training methodThe essence of SVM training method is to solve Quadratic Programming(QP), and it belongs to the area of optimazation theory. In this paper, the basic concept of the optimization theory has been introduced and the essence of improved training method is also discussed.(2) Applying SVM in image processingBased on the deeply analysis of SVM learning, this paper discuss applying SVM in image processing, such as simple image classification, remote sensing image classification and so on.(3) Using SVM in face detectionEigenceface of Principal Component Analysis (PC A) has demonstrated its success in face detection, recognition, and tracking. But PCA method has some limitations, it is built on the low order statistics of the image set, and does not address higher order statistical dependencies such as the relationships among several pixels. To solve this problem, one new method is proposed in this paper. The image features can be extracted by Kernel PCA, and then classified by SVM. This method can not onlyexpress the high order feature, but also can void the bad results issued by dimension explosion.