B Spline Wavelet and SVM Threshold Based Medical Image Edge Extraction

In order to extract the significant edges of the medical image without the interference of the internal noise and the external noise, a new medical image edge extraction algorithm is proposed. We use the information measure to remove the internal noise. According to the characteristics of wavelet transform and our requirements, we design a new quadratic B spline wavelet to discriminate the external noise and extract the edge feature information at the same time. Then we choose the Support Vector Machines (SVM) to select the threshold of the wavelet modulus to reconstruct the edge characteristics. After the computer simulation, the experiment results show that the proposed edge extraction algorithm provides the best trade off between noise rejection and accurate edge localization.