Training Deep Stereo Matching Networks on Tree Branch Imagery: A Benchmark Study for Real-Time UAV Forestry Applications

• URL: http://arxiv.org/abs/2602.19763v1
• Date: Mon, 23 Feb 2026 12:12:43 GMT
• Title: Training Deep Stereo Matching Networks on Tree Branch Imagery: A Benchmark Study for Real-Time UAV Forestry Applications
• Abstract summary: We present the first study to train and test ten deep stereo matching networks on real tree branch images.<n>We use the Canterbury Tree Branches dataset -- 5,313 stereo pairs from a ZED Mini camera at 1080P and 720P.<n>Ten methods cover step-by-step refinement, 3D convolution, edge-aware attention, and lightweight designs.
• Score: 5.266753902938501
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Autonomous drone-based tree pruning needs accurate, real-time depth estimation from stereo cameras. Depth is computed from disparity maps using $Z = f B/d$, so even small disparity errors cause noticeable depth mistakes at working distances. Building on our earlier work that identified DEFOM-Stereo as the best reference disparity generator for vegetation scenes, we present the first study to train and test ten deep stereo matching networks on real tree branch images. We use the Canterbury Tree Branches dataset -- 5,313 stereo pairs from a ZED Mini camera at 1080P and 720P -- with DEFOM-generated disparity maps as training targets. The ten methods cover step-by-step refinement, 3D convolution, edge-aware attention, and lightweight designs. Using perceptual metrics (SSIM, LPIPS, ViTScore) and structural metrics (SIFT/ORB feature matching), we find that BANet-3D produces the best overall quality (SSIM = 0.883, LPIPS = 0.157), while RAFT-Stereo scores highest on scene-level understanding (ViTScore = 0.799). Testing on an NVIDIA Jetson Orin Super (16 GB, independently powered) mounted on our drone shows that AnyNet reaches 6.99 FPS at 1080P -- the only near-real-time option -- while BANet-2D gives the best quality-speed balance at 1.21 FPS. We also compare 720P and 1080P processing times to guide resolution choices for forestry drone systems.

Related papers

• Robostral Navigate [138.74472687550374]
  Robostral Navigate is a vision-language model for navigation systems at scale.<n>It consumes only a stream of monocular RGB images and predicts waypoints by pointing to the next target location in the current camera view.<n>We generate 2.4 million trajectories across 350k simulated scenes to reduce the reliance on real-world data collection.
  arXiv  Detail & Related papers  (2026-07-22T23:13:05Z)
• Low-Cost Stereo Vision for Robust 3D Positioning of Thin Radiata Pine Branches in Autonomous Drone Pruning [5.266753902938501]
  Existing autonomous pruning platforms typically rely on expensive sensors such as LiDAR.<n>This paper investigates whether a single low-cost stereo camera mounted on a drone can provide sufficiently accurate branch detection.
  arXiv  Detail & Related papers  (2026-05-06T10:01:25Z)
• Real-Time Branch-to-Tool Distance Estimation for Autonomous UAV Pruning: Benchmarking Five DEFOM-Stereo Variants from Simulation to Jetson Deployment [5.266753902938501]
  We train five variants of DEFOM-Stereo on a task-specific synthetic dataset and deploy the checkpoints on an NVIDIA Jetson Orin Super 16 GB.<n>The training corpus is built in Unreal Engine 5 with a simulated ZED Mini stereo camera 5,520 stereo pairs across 115 tree instances from three viewpoints at 2m distance.<n>We conclude that DEFOM-PrunePlus provides the most practical accuracy-latency balance for onboard distance estimation.
  arXiv  Detail & Related papers  (2026-03-27T10:16:39Z)
• UE5-Forest: A Photorealistic Synthetic Stereo Dataset for UAV Forestry Depth Estimation [5.266753902938501]
  We present a photorealistic synthetic stereo dataset built entirely in Unreal Engine 5 (UE5)<n>One hundred and fifteen photogrammetry-scanned trees from the Quixel Megascans library are placed in virtual scenes.<n> Orbiting each tree at up to 2 m across three elevation bands (horizontal, +45 degrees, -45 degrees) yields 5,520 rectified 1920 x 1080 stereo pairs with pixel-perfect disparity labels.
  arXiv  Detail & Related papers  (2026-03-13T04:27:24Z)
• DashCam Video: A complementary low-cost data stream for on-demand forest-infrastructure system monitoring [1.6064410860203764]
  This study introduces a novel, low-cost, and reproducible framework for real-time, object-level structural assessment and geolocation of roadside vegetation and infrastructure.<n>We developed an end-to-end pipeline that combines monocular depth estimation, depth error correction, and geometric triangulation to generate accurate spatial and structural data from vehicle-mounted dashcams.<n>Our approach complements conventional RS methods, such as LiDAR and image by offering a fast, real-time, and cost-effective solution for object-level monitoring of vegetation risks and infrastructure exposure.
  arXiv  Detail & Related papers  (2025-08-15T16:55:12Z)
• Rapid Salient Object Detection with Difference Convolutional Neural Networks [49.838283141381716]
  This paper addresses the challenge of deploying salient object detection (SOD) on resource-constrained devices with real-time performance.<n>We propose an efficient network design that combines traditional wisdom on SOD and the representation power of modern CNNs.
  arXiv  Detail & Related papers  (2025-07-01T20:41:05Z)
• DepthVanish: Optimizing Adversarial Interval Structures for Stereo-Depth-Invisible Patches [52.797945735414665]
  Adversarial attacks against stereo depth estimation can help reveal vulnerabilities before deployment.<n>In this work, we discover that introducing regular intervals among the repeated textures, creating a grid structure, significantly enhances the patch's attack performance.<n>Our generated adversarial patches can be inserted into any scenes and successfully attack advanced stereo depth estimation methods.
  arXiv  Detail & Related papers  (2025-06-20T02:22:21Z)
• FlashDepth: Real-time Streaming Video Depth Estimation at 2K Resolution [50.55876151973996]
  A versatile video depth estimation model should (1) be accurate across frames, (2) produce high-resolution depth maps, and (3) support real-time streaming.<n>We propose FlashDepth, a method that satisfies all three requirements, performing depth estimation on a 2044x1148 streaming video at 24 FPS.<n>We evaluate our approach across multiple unseen datasets against state-of-the-art depth models, and find that ours outperforms them in terms of boundary sharpness and speed by a significant margin, while maintaining competitive accuracy.
  arXiv  Detail & Related papers  (2025-04-09T17:59:31Z)
• RayMVSNet++: Learning Ray-based 1D Implicit Fields for Accurate Multi-View Stereo [21.209964556493368]
  RayMVSNet learns sequential prediction of a 1D implicit field along each camera ray with the zero-crossing point indicating scene depth. RayMVSNet++ achieves state-of-the-art performance on the ScanNet dataset.
  arXiv  Detail & Related papers  (2023-07-16T02:10:47Z)
• CRN: Camera Radar Net for Accurate, Robust, Efficient 3D Perception [20.824179713013734]
  We propose Camera Radar Net (CRN), a novel camera-radar fusion framework. CRN generates semantically rich and spatially accurate bird's-eye-view (BEV) feature map for various tasks. CRN with real-time setting operates at 20 FPS while achieving comparable performance to LiDAR detectors on nuScenes.
  arXiv  Detail & Related papers  (2023-04-03T00:47:37Z)
• TransVisDrone: Spatio-Temporal Transformer for Vision-based Drone-to-Drone Detection in Aerial Videos [57.92385818430939]
  Drone-to-drone detection using visual feed has crucial applications, such as detecting drone collisions, detecting drone attacks, or coordinating flight with other drones. Existing methods are computationally costly, follow non-end-to-end optimization, and have complex multi-stage pipelines, making them less suitable for real-time deployment on edge devices. We propose a simple yet effective framework, itTransVisDrone, that provides an end-to-end solution with higher computational efficiency.
  arXiv  Detail & Related papers  (2022-10-16T03:05:13Z)
• Deep Learning on Home Drone: Searching for the Optimal Architecture [54.535788447839884]
  We suggest the first system that runs real-time semantic segmentation via deep learning on a weak micro-computer such as the Raspberry Pi Zero v2 attached to a toy-drone. In particular, since the Raspberry Pi weighs less than $16$ grams, and its size is half of a credit card, we could easily attach it to the common commercial DJI Tello toy-drone. The result is an autonomous drone that can detect and classify objects in real-time from a video stream of an on-board monocular RGB camera.
  arXiv  Detail & Related papers  (2022-09-21T11:41:45Z)
• Real-time dense 3D Reconstruction from monocular video data captured by low-cost UAVs [0.3867363075280543]
  Real-time 3D reconstruction enables fast dense mapping of the environment which benefits numerous applications, such as navigation or live evaluation of an emergency. In contrast to most real-time capable approaches, our approach does not need an explicit depth sensor. By exploiting the self-motion of the unmanned aerial vehicle (UAV) flying with oblique view around buildings, we estimate both camera trajectory and depth for selected images with enough novel content.
  arXiv  Detail & Related papers  (2021-04-21T13:12:17Z)
• CoordiNet: uncertainty-aware pose regressor for reliable vehicle localization [3.4386226615580107]
  We investigate visual-based camera localization with neural networks for robotics and autonomous vehicles applications. Our solution is a CNN-based algorithm which predicts camera pose directly from a single image. We show that our proposal is a reliable alternative, achieving 29cm median error in a 1.9km loop in a busy urban area.
  arXiv  Detail & Related papers  (2021-03-19T13:32:40Z)
• ZoomNet: Part-Aware Adaptive Zooming Neural Network for 3D Object Detection [69.68263074432224]
  We present a novel framework named ZoomNet for stereo imagery-based 3D detection. The pipeline of ZoomNet begins with an ordinary 2D object detection model which is used to obtain pairs of left-right bounding boxes. To further exploit the abundant texture cues in RGB images for more accurate disparity estimation, we introduce a conceptually straight-forward module -- adaptive zooming.
  arXiv  Detail & Related papers  (2020-03-01T17:18:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.