Depth-Supervised Fusion Network for Seamless-Free Image Stitching

2025-10-24

AI Summary

Image stitching suffers from ghosting and misalignment under significant scene depth variations due to parallax-induced geometric inconsistencies. To address this, we propose a depth-consistent seamless stitching framework. First, a multi-stage alignment network ensures robust geometric registration. Second, a global depth regularization constraint is introduced to suppress pixel-level displacements caused by depth inconsistency. Third, graph-cut optimization is integrated with a soft seam diffusion mechanism to generate visually continuous and structurally preserved seam paths. Finally, structural reparameterization accelerates inference without compromising accuracy. Experiments demonstrate that our method significantly reduces alignment errors across diverse challenging depth scenes, producing wide-field-of-view images with seamless transitions, natural appearance, and high fidelity—while maintaining real-time performance. Quantitative and qualitative evaluations confirm superior overall performance compared to state-of-the-art approaches.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

Image stitching synthesizes images captured from multiple perspectives into a single image with a broader field of view. The significant variations in object depth often lead to large parallax, resulting in ghosting and misalignment in the stitched results. To address this, we propose a depth-consistency-constrained seamless-free image stitching method. First, to tackle the multi-view alignment difficulties caused by parallax, a multi-stage mechanism combined with global depth regularization constraints is developed to enhance the alignment accuracy of the same apparent target across different depth ranges. Second, during the multi-view image fusion process, an optimal stitching seam is determined through graph-based low-cost computation, and a soft-seam region is diffused to precisely locate transition areas, thereby effectively mitigating alignment errors induced by parallax and achieving natural and seamless stitching results. Furthermore, considering the computational overhead in the shift regression process, a reparameterization strategy is incorporated to optimize the structural design, significantly improving algorithm efficiency while maintaining optimal performance. Extensive experiments demonstrate the superior performance of the proposed method against the existing methods. Code is available at https://github.com/DLUT-YRH/DSFN.

Problem

• Addresses parallax-induced ghosting in multi-view image stitching
• Develops depth-constrained alignment for objects across different depths
• Optimizes seam computation and fusion for seamless transition areas

Innovation

• Depth-consistency constraints enhance multi-view alignment accuracy
• Graph-based optimal seam computation with soft-seam diffusion
• Reparameterization strategy optimizes structure for efficiency