Automatic image mosaic based on SIFT using bidirectional matching

In the paper, an image mosaic algorithm based on SIFT feature matching is proposed. For an image mosaic method based on feature matching, feature detection is needed to perform in each image. Thus a rapid detection operator is essential to the efficiency of the whole algorithm. In this paper, we use SIFT to extract features. The extracted features are matched by k-d tree and bidirectional matching strategy to enhance the accuracy of matching. Then, a RANSAC algorithm is applied to eliminate outliers to ensure effectiveness of the matching. Finally images are stitched by weighted average blending algorithm. The presented algorithm overcomes the disadvantages of the traditional image mosaic methods which are susceptible to different scale and moving objects, and can achieve sub-pixel accuracy of matching and algorithm is still available to the images at different scale. Experimental results show that the method with strong robustness performs effectively.