AUTHOR=Tang Wei , Jia Fangxiu , Wang Xiaoming 

TITLE=An improved adaptive triangular mesh-based image warping method

JOURNAL=Frontiers in Neurorobotics

VOLUME=Volume 16 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/neurorobotics/articles/10.3389/fnbot.2022.1042429

DOI=10.3389/fnbot.2022.1042429

ISSN=1662-5218

ABSTRACT=This paper presents an improved mesh-based joint optimization image stitching model, which could realize accurate alignment and less distortion in the overlapping and non-overlapping regions. Since the uniform vertices are usually used in mesh-based warps, we consider the matched feature points and uniform points as grid vertices to strengthen constraints on deformed vertices. Simultaneously, we define an improved energy function and add a color similarity term to perform alignment beyond color consistency. Therefore, the final panorama via the global optimization will present similar colors to the reference image. In addition to good alignment and minimal local distortion, a regularization parameter strategy of combining our method with APAP warp is introduced. And controlling the proportion of each part by calculating the distance between the vertex and the nearest matched feature point to the vertex. This ensures a more natural stitching effect in non-overlapping areas. A comprehensive evaluation shows that the proposed method achieves more accurate image stitching, with significantly reduced ghosting effects in the overlapping regions and more natural results in the other areas. The comparative experiments outperform the state-of-the-art image stitching warps.