Single-Perspective Warps in Natural Image Stitching
This work addresses perspective consistency issues in image stitching for applications like photography and computer vision, representing an incremental improvement over existing methods.
The paper tackles the problem of projective distortion in single-perspective image stitching by proposing two new warps, a parametric and a mesh-based warp, and demonstrates that these outperform state-of-the-art methods like homography and APAP in evaluation.
Results of image stitching can be perceptually divided into single-perspective and multiple-perspective. Compared to the multiple-perspective result, the single-perspective result excels in perspective consistency but suffers from projective distortion. In this paper, we propose two single-perspective warps for natural image stitching. The first one is a parametric warp, which is a combination of the as-projective-as-possible warp and the quasi-homography warp via dual-feature. The second one is a mesh-based warp, which is determined by optimizing a total energy function that simultaneously emphasizes different characteristics of the single-perspective warp, including alignment, naturalness, distortion and saliency. A comprehensive evaluation demonstrates that the proposed warp outperforms some state-of-the-art warps, including homography, APAP, AutoStitch, SPHP and GSP.