Rotational Invariant LBP-SURF for Fast and Robust Image Matching

Speeded Up Robust Features (SURF) is one of the most robust and widely used image matching algorithms based on local features. However, the performance for rotation image is poor when one image is a rotated version of the other. To improve the matching accuracy of rotation image, we present an modified image matching algorithm combining Haar wavelet and the rotation invariant Local Binary Patterns (LBP). Firstly, keypoints are extracted from the images for matching by applying the Hessian matrix and integral images. Secondly, each keypoint is described by the Rotation Invariant LBP patterns and Haar wavelet, which are computed from the image patch centered at the keypoint. Finally, the matching pairs between the two sets of keypoints are determined by using the nearest neighbor distance based on matching strategy. The experimented results show that in comparison with prior works the proposed algorithm is efficient when tested on the images of scaling, rotation, blurring and brightening.