一种融合多特征的UAV快速目标识别  

Fast UAV target recognition algorithm based on multiple features

作　　者：张清[1] 刘慧霞[2] 席庆彪[1,2] 

机构地区：[1]西北工业大学自动化学院,西安710072 [2]西北工业大学第365研究所,西安710065

出　　处：《计算机工程与应用》2013年第3期238-242,共5页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61074155)

摘　　要：针对UAV(Unmanned Aerial Vehicle)侦察图像快速目标识别问题,提出一种融合多特征的UAV快速目标识别算法。该算法结合图像的不变矩特征和SIFT特征,用不变矩特征构造适应度函数并利用遗传算法的全局搜索能力,在侦察图像中进行搜索,快速提取出可能包含目标的感兴趣区域(Region Of Interest,ROI)。采用尺度不变特征变换算法(Scale Invariant Feature Transform,SIFT)在ROI区域中进行匹配识别,从而精确确定目标的位置。仿真结果表明:该算法的鲁棒性较强,能有效识别特定目标并显著减少识别时间。To settle the problem of identifying the target in the UAV（Unmanned Aerial Vehicle）reconnaissance image in real time, a UAV fast target identify algorithm is proposed which is based on invariant moments and the SIFT（Scale Invariant Feature Transform）features of the image. A similarity function is designed based on invariant moments, which is taken as the fitness function of GA（Genetic Algorithm）. The image is searched globally using GA to extract the Region Of Interest（ROI） quickly which may contain targets. The object is located accurately in the ROI region based on the SIFT transform. Simulation results indicate that the method is of strong robustness and can identify the targets effectively and also reduce the recognition time significantly.

关 键 词：目标识别 无人机 感兴趣区域(ROI) 不变矩 遗传算法 尺度不变特征变换(SIFT)算法 

分 类 号：TP753[自动化与计算机技术—检测技术与自动化装置]