长视距空战中可疑目标的快速识别模型仿真

Long Range Air Quickly Identify Suspicious Target in Simulation Model

研究长视距空战中可疑目标的快速识别问题,由于在长视距空战中,可疑目标与空中识别设备之间存在较长的距离,采集目标比较弱小,特征模糊。目标一旦过于接近,会形成堆积盲区,传统的识别方法在遇到远距离可疑目标时,由于距离太远,会出现目标堆积盲区,导致对目标分割出现错误,无法准确识别目标。为解决上述问题,提出采用主成分聚类算法的长视距空战中可疑目标的快速识别方法。针对可疑目标采集大量的计算机视觉图像,提取计算机视觉图像中可疑目标的特征,将上述特征作为可疑目标识别的依据。针对所有的可疑目标识别特征,进行可疑目标的快速识别,并对识别到的可疑目标进行累加运算,实现可疑目标的智能识别。实验结果表明,利用改进算法进行长视距空战中可疑目标的快速识别,能够极大的提高识别的准确性。

Study the problems of quickly identify suspicious targets in long-horizon air combat. The paper used methods of principal component analysis and cluster algorithm to quickly identify suspicious targets in long-horizon air combat. For a large number of computer visual images of suspicious target, we extracted computer visual image features of suspicious targets as a basis for identifying suspicious targets. For all of the recognition features of suspicious targets, we quickly identified the suspicious targets and carried out the accumulation for identified suspicious targets to achieve the intelligent recognition of suspicious targets. The experimental results show that the improved algorithm can greatly improve the accuracy of recognition.