入侵目标视觉检测与识别的研究进展

Review on Vision Based Intruder Detection and Recognition

摘要:

入侵目标视觉检测与识别是无人机感知与规避技术领域中重点研究的课题,关键任务是无人机在飞行过程中通过机载传感器获取光学机载图像中并判断是否存在入侵目标,并对入侵目标进行检测识别和定位。入侵目标视觉检测与识别是将无人机安全集成到国家空域并保证无人机和有人机飞行安全的关键技术之一。本文主要围绕无人机感知与规避技术中的入侵目标视觉检测与识别技术方面,分析检测与识别入侵目标所面临的一些难点问题,综述当前入侵目标视觉检测与识别的主要处理方法,并指出了该领域存在的尚未解决的问题和展望未来的发展趋势。

Abstract:

Intruder detection and recognition are the important topics of UAV in sense and avoid technology. The key task is to detect whether there is an intruder in Optical airborne images obtained by airborne sensors during the flight of the UAV, and identify and locate the intruder. This is one of the key technologies for the safe integration of UAVs into the national airspace and for ensuring the safety of UAVs and manned aircraft. This paper focuses on the detection and recognition of intruder in sense and avoid technology, analyzes the difficulties in detection and recognition of intruder, summarizes the main methods of visual detection and recognition of intruder, and points out the existing problems in this field future development trends.