Ortho-Fuse: Orthomosaic Generation for Sparse High-Resolution Crop Health Maps Through Intermediate Optical Flow Estimation
Rugved Katole, Christopher Stewart
TL;DR
This work tackles the bottleneck of generating reliable orthomosaics from sparse aerial imagery in AI-driven crop health mapping. It introduces Ortho-Fuse, an intermediate optical-flow-based pipeline that uses Real-Time Flow Estimation (RIFE) to synthesize intermediate frames between consecutive images, linearly interpolates GPS metadata, and integrates with OpenDroneMap to produce high-quality orthomosaics with reduced image overlap. Experiments show a 20% reduction in the minimum required overlap while preserving NDVI-based crop health analysis accuracy, demonstrating practical data-collection cost savings. The study also discusses adoption barriers in precision agriculture and proposes diffusion-based and hybrid generative approaches as future directions to broaden applicability and integration with existing photogrammetric workflows.
Abstract
AI-driven crop health mapping systems offer substantial advantages over conventional monitoring approaches through accelerated data acquisition and cost reduction. However, widespread farmer adoption remains constrained by technical limitations in orthomosaic generation from sparse aerial imagery datasets. Traditional photogrammetric reconstruction requires 70-80\% inter-image overlap to establish sufficient feature correspondences for accurate geometric registration. AI-driven systems operating under resource-constrained conditions cannot consistently achieve these overlap thresholds, resulting in degraded reconstruction quality that undermines user confidence in autonomous monitoring technologies. In this paper, we present Ortho-Fuse, an optical flow-based framework that enables the generation of a reliable orthomosaic with reduced overlap requirements. Our approach employs intermediate flow estimation to synthesize transitional imagery between consecutive aerial frames, artificially augmenting feature correspondences for improved geometric reconstruction. Experimental validation demonstrates a 20\% reduction in minimum overlap requirements. We further analyze adoption barriers in precision agriculture to identify pathways for enhanced integration of AI-driven monitoring systems.