Ortho-Fuse: Orthomosaic Generation for Sparse High-Resolution Crop Health Datasets

Overview

Ortho-Fuse presents a novel approach for generating high-quality orthomosaics from sparse high-resolution imagery collected by UAVs for precision agriculture applications. The framework leverages intermediate optical flow estimation to improve alignment and stitching quality even when image overlap is limited.

Key Features

• Intermediate optical flow estimation for improved image alignment
• Handles sparse image datasets with limited overlap
• Optimized for high-resolution crop health imagery
• Efficient processing pipeline for large-scale agricultural monitoring

Technical Approach

The system uses a multi-stage pipeline that first estimates optical flow between sequential images, then uses this information to guide the orthomosaic generation process. This approach is particularly effective for UAV-collected imagery where traditional feature-matching approaches may struggle due to repetitive crop patterns and varying lighting conditions.

Applications

This work enables more efficient crop health monitoring by reducing the number of images needed for complete coverage while maintaining high-resolution detail. The generated orthomosaics can be used for:

• Precision agriculture and crop health assessment
• Disease and pest detection
• Yield prediction and optimization
• Environmental monitoring