摘要
针对红外目标类型识别问题,提出了一种基于灰关联分析的红外目标识别算法。该算法在对红外图像进行边缘检测后,提取目标的奇异值特征作为关联参数,通过比较特征参数的灰关联值实现目标识别。该算法根据目标各个特征指标稳定性的不同,采用了一种灰关联加权系数的方法,仿真实验验证了该方法在红外目标识别问题上的有效性,并对目标旋转和平移有一定的适应能力。
Aimed at the problem of the infrared target classification, a recognition method based on gray correlation analysis algorithm is proposed throught extracting the singular value feature from the target image after making edge detection to an infratred image. Target recognition is implemented by contrasting gray correlation value of features. One approache determining the weighed corefl＇icients are also proposed based on stadility of the features of the target. The example proves that this method can be used for solving the problem of the infrared target classification effectively, and it is adaptive to the rotated target and romoved target.
出处
《激光与红外》
CAS
CSCD
北大核心
2008年第6期609-611,618,共4页
Laser & Infrared
关键词
红外
目标识别
灰关联
边缘检测
inifrared
target classification
gray correlation analysis
edge detection
