An image matching algorithm based on Rough-SIFT descriptor

For the rotation, translation, scale invariant properties of SIFT (Scale Invariant Feature Transform) feature, it has been widely applied in image matching. However, it is represented by a 128 element feature vector, and when it is used for image matching, especially for the case that there are many keypoints in the image, the matching speed will be slow and storage requirement will be huge, and matching precision is low. In order to overcome these disadvantages, a new image matching algorithm based on Rough-SIFT descriptor is proposed in this paper. Firstly, we select the robust and salient SIFT features according to ranking the local invariant features. Then the rough set theory is introduced into image matching by putting forward a new approximation reduction algorithm which is used to reduce the dimension of SIFT feature vector. At last, the reduced feature points are used to image matching. Some experimental results have been provided to show the proposed method not only effectively realizes image matching, but also has higher matching speed and lower storage requirement.