Progressive Per-Branch Depth Optimization for DEFOM-Stereo and SAM3 Joint Analysis in UAV Forestry Applications

TL;DR

This paper presents a comprehensive multi-stage depth optimization pipeline that significantly improves per-branch 3D reconstruction accuracy from UAV stereo imagery for forestry applications, enabling precise autonomous pruning.

Contribution

It introduces a novel progressive pipeline combining DEFOM-Stereo, SAM3 segmentation, and multi-stage depth refinement specifically tailored for UAV forestry scenarios.

Findings

Reduces per-branch depth standard deviation by 82%

Produces geometrically coherent 3D point clouds

Enhances robustness of UAV-based forestry analysis

Abstract

Accurate per-branch 3D reconstruction is a prerequisite for autonomous UAV-based tree pruning; however, dense disparity maps from modern stereo matchers often remain too noisy for individual branch analysis in complex forest canopies. This paper introduces a progressive pipeline integrating DEFOM-Stereo foundation-model disparity estimation, SAM3 instance segmentation, and multi-stage depth optimization to deliver robust per-branch point clouds. Starting from a naive baseline, we systematically identify and resolve three error families through successive refinements. Mask boundary contamination is first addressed through morphological erosion and subsequently refined via a skeleton-preserving variant to safeguard thin-branch topology. Segmentation inaccuracy is then mitigated using LAB-space Mahalanobis color validation coupled with cross-branch overlap arbitration. Finally, depth noise - the most persistent error source - is initially reduced by outlier removal and median filtering, before being superseded by a robust five-stage scheme comprising MAD global detection, spatial density consensus, local MAD filtering, RGB-guided filtering, and adaptive bilateral filtering. Evaluated on 1920x1080 stereo imagery of Radiata pine (Pinus radiata) acquired with a ZED Mini camera (63 mm baseline) from a UAV in Canterbury, New Zealand, the proposed pipeline reduces the average per-branch depth standard deviation by 82% while retaining edge fidelity. The result is geometrically coherent 3D point clouds suitable for autonomous pruning tool positioning. All code and processed data are publicly released to facilitate further UAV forestry research.