An improved method for eliminating ghosting in image stitching

In the process of image stitching, Image stitching algorithms based on SIFT (Scale Invariant Feature Transform) feature have the robustness to the complex transformation between images. However, image stitching appears discontinuous and ghosting in the image registration. We proposed a fast image stitching algorithm to eliminate the seam line and ghosting. The algorithm through two cameras to capture images, and then combined with the OpenCV library to process image splicing using the improved fusion algorithm for the overlapping areas of two images. Firstly, the algorithm extracts the overlapping range of SIFT features of image mosaics. Secondly, the algorithm used the RANSAC algorithm to compute the homograph matrix H. Thirdly, an integral image without seam line was received by the combination of the improved image fusion algorithm which realizes seamless transition between the weight coefficients of images and changes in distance. The experimental results show that the algorithm not only eliminates the seam line and ghosting effectively, but also improves the speed of image stitching.