多装备远程智能协同调度及控制系统地质灾害灾后现场数据集

该数据系基于重大自然灾害灾后复杂现场、次生灾害不可预见的非结构化救援现场的正负障碍物场景感知实验训练数据，数据集涵盖该场景类型下的不同种类的障碍物类型（救援人员，建筑废料，土堆，石块，凹坑，壕沟等），数据样本量共计10个，每个数据样本以点级细粒度严格标注了所在场景中障碍物的类型和范围。

This dataset comprises experimental training data for positive and negative obstacle scene perception in complex post-disaster sites of major natural disasters and unstructured rescue sites facing unpredictable secondary disasters. It includes various obstacle types under such scenarios, such as rescue personnel, construction waste, soil piles, stones, pits, trenches, and more. The dataset consists of 10 data samples in total, with each sample having strict point-level fine-grained annotations for the category and extent of obstacles present in the corresponding scene.

搜集汇总

背景与挑战

背景概述

该数据集是面向地质灾害灾后现场的非结构化救援场景数据，包含10个样本，以点级细粒度标注了多种障碍物类型（如救援人员、建筑废料、土堆等），主要用于激光雷达场景语义分割研究。数据集由东南大学创建，属于国家重点研发计划项目，旨在支持多装备远程智能协同调度及控制系统的开发。