Design Of Texture Feature Extraction And Classification Algorithm For Rib Cortex

The extraction and classification of the texture features of rib X-ray plain film is important clinical significance for the diagnosis of non dominant rib fractures. To reduce the missed diagnosis and improve the diagnostic accuracy rate has a significant role.There is no existing algorithm in the field of medical image processing. The purpose of this paper is that using mature image processing technology, according to the characteristics of medical image of rib, presents a including preprocessing, texture feature extraction and classification algorithm, computer display, the algorithm can do improve rib texture feature extraction classification accuracy.Firstly, this paper studies the extraction method of texture of the basic concepts and properties and characteristics, the currently widely used texture feature extraction and classification method are introduced in this paper, summary introduces several with the aid of statistical texture characteristic extraction method, and points out their respective advantages and disadvantages. For texture feature extraction and classification principles, models, methods, and the use of the technology were summarized, in rib texture algorithm in feature extraction techniques were based on the analysis of, procedure of a rib image texture analysis is put forward. Canny edge detection operator was used to extract the contour of the image, and then Gabor wavelet transform was used to extract texture features. Finally, the image texture features were classified by the superiority of KPCA dimension reduction. And according to this algorithm, is traditional Gabor wavelet transform features extraction and adding rotation normalization circular operator of Gabor wavelet transform feature extraction; traditional KPCA feature classification and the combination of information to measure KPCA feature vectors of the feature classification, pair wise comparison experiment. Results show, Canny edge detection operator pre processed rib texture image, by adding standard rotating of circular operator of Gabor wavelet transform feature extraction. Finally, the binding information to measure feature vector of KPCA feature classification obtained by the combination of a highest accuracy of classification. And compared to other types of algorithms in the accuracy of the algorithm has been greatly improved.