A sub-pixel feature extraction method for CT image based on Zernike moments

This paper will combine Sobel operator with Zernike moments operator to extract the sub-pixel edge feature for liver CT images. Sobel operator can effectively smooth the noise, provide more intact edge feature, then propose a limited non-optimal suppression improved method based on the Zernike moments. By a large number of experiments, improved method increases the accuracy of feature extraction, besides the speed of extraction is very fast.