首页|基于YOLO图像检测的激光时序控制与二次跟踪打击方法

基于YOLO图像检测的激光时序控制与二次跟踪打击方法

扫码查看
在激光打击无人机过程中,主要依靠激光在目标飞控区域灼烧击穿目标外壳击中飞控区域致其坠落,能否精确打击目标飞控区是击落目标的关键,但长时间激光打击产生的火焰、烟雾会导致基于图像的目标检测与跟踪受到严重干扰,导致激光无法精确锁定无人机飞控区域,严重影响了实战击落概率.设计了一种打击方法,采用YOLO算法对目标图像进行检测,用检测结果控制激光出光时序并进行二次跟踪.实验表明所用打击方法能够将稳定度提高59%以上,跟踪策略的改良有效提高了击落无人机的能力.
Laser Timing Control and Secondary Tracking Strike Method Based on YOLO Image Detection
In the process of laser strikes on drones,the main reliance is on the laser to burn and penetrate the target's shell in the target's flight control area,causing it to fall.Whether the target's flight control area can be accurately hit is the key to shooting down the target.However,the flames and smoke generated by long-term laser strikes can cause serious interference in image-based target detection and tracking,making it difficult for the laser to accurately lock the drone's flight control area,seriously affecting the probability of actual combat shooting down,This article designs a strike method that uses the YOLO algorithm to detect target images,and uses the detection results to control the laser output timing and perform secondary tracking.The experiment shows that the strike method used in this article can improve stability by more than 59%.The improvement of the tracking strategy effectively improves the ability to shoot down drones.

target detectiontarget trackinganti drone systemshigh-energy laser weaponsstrike strategieslaser control

葛奇鹏、杨林、李子龙、杨建强

展开 >

华中光电技术研究所—武汉光电国家研究中心,湖北武汉 430223

陆军装备部驻武汉地区军事代表局驻武汉地区第二军事代表室,湖北武汉 430000

目标检测 目标跟踪 反无人机系统 高能激光武器 打击策略 激光控制

2024

光学与光电技术
华中光电技术研究所 武汉光电国家实验室 湖北省光学学会

光学与光电技术

CSTPCD
影响因子:0.351
ISSN:1672-3392
年,卷(期):2024.22(3)