反舰导弹智能化目标检测应用研究

Research on Intelligent Target Detection Application of Anti-Ship Missiles

将AI目标检测算法应用于对海上舰船目标的检测识别,选用YOLOv4目标检测网络,通过自建海上舰船目标数据集验证网络检测性能和有关实验,证实YOLOv4网络能够实现反舰导弹对遮蔽目标及小目标的精准高效检测,通过NVIDIA Jetson TX2嵌入式平台建立任务型计算机辅助雷达、红外等传统制导方式,验证该网络在海上实景图像中的性能,为实现反舰导弹精细化目标选择提供了可行技术途径。

Aiming at the combat demand of realizing anti-ship missiles’accurate strike on the shielding and small targets,AI target detection algorithms were applied to the detection of ship targets on the sea by using YOLOv4 network and verifying the network’s performance by self-built data sets,which concludes that YOLOv4 network can meet the accuracy and speed to detect shielding and small targets,the network’s performance was verified further by NVIDIA Jetson TX2 development board.The experimental results show that the network can improve the detection accuracy about shielding and small targets through the embedded platform by establishing a mission computer to aid radar,infrared and other guidance modes,which provides a feasible technical approach to realize delicacy target selection of anti-ship missiles.