This is the authors' abstract. We don't add key points for this paper.
Multi-aperture system-based high-resolution panoramic image stitching technology, is an effective method to overcome the limitations of spatial bandwidth in single imaging systems. However, significant computational demands and inadequate real-time capabilities frequently challenge panoramic image stitching. However, due to the fast-moving objects in the scene, the stitching result will exist misalignment and artifacts, which affecting the overall quality of the panoramic image. To solve these challenges, we proposed a dynamic panoramic image stitching technique based on a 5×5 compound eye array. We use Compute Unified Device Architecture (CUDA) technology to boost the processes of image registration, furthermore, this method optimizes the stitching result of multiple frames by introducing the Visual Background Extraction (ViBE) algorithm and an advanced seam line search criterion. This method not only reduces the computational complexity associated with stitching adjacent frames, but also eradicates misalignments and artifacts resulting from moving objects. This approach achieves high-quality, rapid stitching of dynamic panoramic images, providing a billion-pixel panoramic image with a horizontal field of view of more than 90 degrees.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Wu et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: