SURF Feature Description Method of Color Image Based on Quaternion

To solve the problem of feature description in color image, in this paper, a novel method of SURF (speed up robust features) feature description of color image based on quaternion is proposed according to the influences of color information to image feature attribute. Firstly, the color information of three color channels is expressed by the three imaginary parts of pure quaternion separately. Then, taking advantage of the characteristic of rotation invariance of quaternion norm, feature description vectors are described by pure quaternion norm. Finally, the simulation experiments show that the correct matching ratios of our method is higher than SURF method and fully affine invariant SURF (FAIR-SURF) method, which verifies the effectiveness of our method in this paper.