VEPL

Vegetation encroachment in power line corridors has multiple problems for modern energy-dependence societies, failures due to contact of lines and vegetation, can result in power outages and millions of dollars in losses. To address this problem, UAVs have emerged as a promising solution due to their ability to quickly and affordably monitor power line corridors for vegetation encroachment. However, the extensive manual task that requires analyzed each photo acquired by UAVs searching for the existence of vegetation encroachment has led many authors to propose using Deep Learning to automate the detection process. Despite the advantages of using UAVs and Deep Learning there is currently a lack of datasets that help to train any Deep Learning models. In this paper, we present a dataset for the segmentation of vegetation encroachment in power line corridors. We use RGB orthomosaics acquired in a rural road using a commercial UAV. RGB sliced images and multi-label mask for vegetation segmentation are provided. We provide detailed description of the image acquisition, the labeling task, the data augmentation techniques among other relevant details to produce the dataset. Researchers would benefit from using the proposed dataset to develop and improved strategies of vegetation encroachment monitoring using UAVs and Deep Learning.

输电走廊植被入侵对现代能源依赖型社会存在多重隐患：线路与植被接触引发的故障，可能导致停电及数百万美元的经济损失。为解决该问题，无人机（Unmanned Aerial Vehicle, UAV）凭借快速且低成本监测输电走廊植被入侵的能力，成为极具前景的解决方案。但人工逐一分析无人机拍摄的每张照片以排查植被入侵的工作体量庞大，因此诸多学者提出利用深度学习（Deep Learning）自动化检测流程。尽管无人机与深度学习结合具备诸多优势，但当前缺乏可用于训练深度学习模型的相关数据集。本文提出一款面向输电走廊植被入侵分割任务的数据集。我们使用商用无人机在乡村道路场景采集的RGB正射影像（RGB orthomosaics）作为数据源。本次公开的数据集包含切片后的RGB图像以及用于植被分割的多标签掩码（multi-label mask）。我们详细说明了图像采集流程、标注任务、数据增强技术及其他与该数据集构建相关的关键细节。研究人员可借助本数据集，开发并优化基于无人机与深度学习的植被入侵监测方案。

搜集汇总

背景与挑战

背景概述

VEPL是一个针对电力线路走廊植被侵占监测的语义分割数据集，包含无人机采集的RGB影像、多标签掩膜和数字表面模型，并提供原始数据及增强处理版本，用于支持深度学习模型的训练和开发。数据集旨在解决植被侵占导致的电力中断问题，推动无人机和深度学习在该领域的应用。