Reducing perspective distortion for stereoscopic image stitching
Weiqing Yan, Chunping Hou · 2016
This paper presents a stereoscopic image stitching method, which reduces perspective distortion of stitched stereoscopic image. We propose a hybrid warping model for stereoscopic image stitching by combining derivative half-projective and content-preserving warping. Firstly, a derivative warp model is proposed to pre-warp left and right images, and thus reduce perspective distortions, meantime, ensure the consistent changes. Secondly, a content-preserving warping is introduced to locally refine the alignment and reduce the vertical disparity. At last, we blend the warped input stereoscopic images to a whole stereoscopic image. Experimental results show that the proposed method can effectively stitch stereoscopic image, which not only reduces the perspective distortion, but also reduces the vertical disparity reasonably.