Spherical Image Stitching Based on ORB and PROSAC Algorithm
Xin Li, Yarong Liu, Donghui Li, Liyan Yang, Xiaofei Yang, Wang Yuchen · 2018
This paper provides a real-time image stitching method. During stitching users often discover that panoramic picture is not ideal because of the spherical camera taking the picture of wide-angle and large-area distortion. We presents the ORB algorithm extracting feature points, and PROSAC algorithm purification. While traditional stitching researchers use the SIFT or SURF method detecting feature points, and the RANSAC algorithm removes outpoints. We discover that SIFT and SURF algorithms require very complex calculations, and the RANSAC algorithm is equally and random for all feature point pairs, so the real-time processing is slow. In this paper, the use of ORB and non-random PROSAC algorithm greatly shortens the running time. The SURF+RANSAC and ORB+PROSAC algorithms are used to compare the stitching speed. Experiments show that the ORB+PROSAC algorithm increase matching speed with better matching accuracy.