Content-Preserving Image Stitching with Regular Boundary Constraints

Abstract: This paper proposes an approach to content-preserving stitching of images with regular boundary constraints, which aims to stitch multiple images to generate a panoramic image with regular boundary. Existing methods treat image stitching and rectangling as two separate steps, which may result in suboptimal results as the stitching process is not aware of the further warping needs for rectangling. We address these limitations by formulating image stitching with regular boundaries in a unified optimization. Starting from the initial stitching results produced by traditional warping-based optimization, we obtain the irregular boundary from the warped meshes by polygon Boolean operations which robustly handle arbitrary mesh compositions, and by analyzing the irregular boundary construct a piecewise rectangular boundary. Based on this, we further incorporate straight line preserving and regular boundary constraints into the image stitching framework, and conduct iterative optimization to obtain an optimal piecewise rectangular boundary, thus can make the panoramic boundary as close as possible to a rectangle, while reducing unwanted distortions. We further extend our method to panoramic videos and selfie photography, by integrating the temporal coherence and portrait preservation into the optimization. Experiments show that our method efficiently produces visually pleasing panoramas with regular boundaries and unnoticeable distortions.