首篇在真实树枝影像上训练十种深度匹配网络，为无人机林业应用提供实时深度估计基准。
Training Deep Stereo Matching Networks on Tree Branch Imagery: A Benchmark Study for Real-Time UAV Forestry Applications
- 用真实树枝图像和DEFOM生成的视差图训练十种深度网络，覆盖3D卷积与轻量设计。
- BANet-3D质量最优（SSIM=0.883），RAFT-Stereo场景理解最强（ViTScore=0.799）。
- AnyNet在1080P下达6.99帧/秒，唯一接近实时；适合对速度要求高的林业无人机。
自主无人机修剪树木需要从立体相机中实现精确、实时的深度估计。深度由视差图通过公式 $Z = f B/d$ 计算，微小视差误差在工作距离下即导致显著深度偏差。基于我们此前发现的DEFOM-Stereo是植被场景最佳视差生成器，本研究首次在真实树枝图像上训练并测试十种深度立体匹配网络。数据集采用Canterbury Tree Branches——5,313对立体图像，来自ZED Mini相机在1080P与720P分辨率下的采集，以DEFOM生成的视差图为训练目标。所涉方法涵盖逐步精炼、3D卷积、边缘感知注意力及轻量化设计。通过感知指标（SSIM、LPIPS、ViTScore）与结构指标（SIFT/ORB特征匹配）评估，发现BANet-3D整体质量最佳（SSIM=0.883，LPIPS=0.157），RAFT-Stereo在场景理解上表现最优（ViTScore=0.799）。在搭载于无人机的NVIDIA Jetson Orin Super（16 GB，独立供电）上测试显示，AnyNet在1080P下达到6.99 FPS，是唯一接近实时的方案；而BANet-2D在质量与速度间取得最佳平衡，达1.21 FPS。同时对比720P与1080P处理时间，为林业无人机系统分辨率选择提供依据。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。