全向凝视型多目标识别实验研究

针对目前大视场凝视红外多目标处理技术的急需,提出了基于红外鱼眼全向凝视型多目标识别数据处理系统和威胁水平排列算法。给出了全向凝视型多目标识别系统组成、硬件电路和算法实现方法,通过采集序列红外图像,实现不同目标在航路、距离、方向、速度等参数的提取。综合目标运动速度、加速度、方向等因素,对多目标威胁等级排序。通过半实物仿真红外多目标源的配合实验,证明了识别系统的可行性和处理精度。

Experimental research on omnidirectional staring multi-target identification

In order to satisfy urgent requirements for large-viewing-field staring infrared multi-target processing techniques, algorithms of data processing and threat level sequencing for the infrared fisheye omnidirectional staring multi-target identification system are proposed. The system composition, hardware circuitry and realization of algorithms are presente＆ The parameters of different targets, including track, distance, direction and velocity, are obtained through acquiring infrared image sequences; and then threat levels of different targets are sequenced based on the combination of target parameters, such as moving velocity, acceleration, direction. Finally, the experiment through semi-physical simulating infrared multi-target sources has demonstrated that the identification system is effective with high precision.