Efficient feature point detection in CT images using Discrete Curvelet Transform

摘要

In this work, a multi-scale feature point detection algorithm in CT slices based on discrete curvelet transform is presented. Discrete curvelet transformation is applied to input CT slices and the behavior of curvelet coefficients in all the scales are examined. The information in the fine and detail levels which contains the edge and singularity details are processed to extract the feature points. Performance comparison is made against wavelet and canny edge detectors based on SSIM index.