地空背景下红外图像弱小飞机目标检测跟踪数据集

红外弱小目标检测跟踪是远程精确打击、空天攻防对抗和遥感情报侦察等**应用中的重要研究内容。针对当前红外目标探测识别领域仿真数据真实性不足、实测数据样本匮乏的情况，本数据集面向低空飞行的弱小飞机目标检测跟踪应用，通过外场实地拍摄和数据准备加工，提供了一套以一架或多架固定机翼无人机目标为探测对象的算法测试数据集。数据集获取场景涵盖了天空、地面等背景以及多种场景，共计22段数据、30条航迹、16177帧图像、16944个目标，每个目标对应一个标注位置，每段数据对应一个标注文件。本数据集可为弱小目标探测、精确制导和红外目标特性等研究提供基础数据。

Infrared small and weak target detection and tracking is a critical research topic in applications such as long-range precision strike, aerospace offensive and defensive confrontation, and remote sensing intelligence reconnaissance. To tackle the current challenges of insufficient authenticity of simulation data and scarcity of real-world measured data samples in the field of infrared target detection and recognition, this dataset is tailored for the application of detecting and tracking small and weak aircraft targets in low-altitude flight. Collected via on-site field shooting and subsequent data preparation and processing, it provides an algorithm test dataset that takes one or more fixed-wing UAV targets as detection objects. The acquisition scenarios of the dataset cover backgrounds including sky, ground and various other scenarios, comprising a total of 22 data segments, 30 flight tracks, 16177 image frames, and 16944 targets. Each target corresponds to one annotated position, and each data segment corresponds to one annotation file. This dataset can provide foundational data for research on small and weak target detection, precision guidance, and infrared target characteristics.

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于地空背景下红外图像中弱小飞机目标检测与跟踪的数据集，包含2017-2019年在陕西省宝鸡市眉县采集的22个图像序列、16177帧图像和16944个目标，每个目标均有标注位置。数据集基于实际外场记录和数据处理，提供真实可靠的红外辐射特性，适用于军事应用如远程精确打击和战场侦察的算法测试与研究。