基于多层次特征向量匹配和支持向量机非刚性图像配准.pdfVIP

Non-Rigid Registration based on Attribute Vector Matching of Different Levels and Support Vector Machine Bo Pan, Zhaowei Gu, Fuming Qu, Gongjian Shen, Anna Wang College of Information Science and Engineering Northeastern University Shenyang, P. R. China 110004 pan_bo_mail@, zhaoweigu123@163.com, qfmqfmqfm@163.com, gj_shen@163.com, wanganna@ Abstract This paper presents a novel deformable registration method based on Gabor Wavelet and SVM. The method introduces Gabor Wavelet and SVM theory to match brain images. Firstly, the coefficients of wavelet are extracted and optimal search radiuses are reached. With the guidance of the SVM, the deformable registration overcomes the shortages of rigid registration, owning the generalization and presenting an excellent robustness. Keywords: Gabor Wavelet, SVM, Deformable Registration, multi-level registration, feature extraction 1 Introduction Image registration is the process of determining a mapping between the coordinates in one image and those in another, to achieve biological, anatomical or functional correspondence. The medical image registration owns great value. When two images are registered, it is possible to analyze differences between the images (e.g. images taken at two different time instants or difference between the template and a tested product in visual inspection) and combine information contained in multiple images into a single image (image fusion) with the goal of easier interpretation by humans