Analyzing Classical and LDI Depth-Aware Image Stitching for Enhanced Virtual View Representation

This article examines two image stitching methods for creating seamless images in stereo vision applications: a classical OpenCV-based approach and a depth-aware technique. Conventional methods, while effective for 2D image stitching, face limitations when applied to 3D scene reconstruction, where accurate depth information is crucial. We introduce a depth-aware image stitching method that utilizes depth maps to enhance alignment and blending in 3D scenes, demonstrating its superiority through comparative experiments. Depth-aware stitching offers notable advantages over classical methods in 3D scene reconstruction, reducing artifacts, minimizing distortions, and providing an accurate 3D representation, emphasizing the importance of depth estimation in image stitching and encouraging further research in this area.