红外技术, 2013, 35 (10): 638, 网络出版: 2013-11-01
高斯曲面拟合在对空红外点目标检测中的应用
Application of Gaussian Quadrics Fitting in the Infrared Point Targets Detection in Sky Background
目标检测 高斯曲面拟合 偏态系数 峰度系数 targets detection Gaussian quadrics fitting skewness coefficient kurtosis coefficient
摘要
针对天空背景红外点目标检测时云层边缘产生的虚警, 根据点目标的高斯分布特性, 提出采用高斯曲面拟合算法, 首先利用传统目标检测算法检测得到目标和云层边缘, 然后获取其高斯曲面拟合参数, 最后计算偏态系数和峰度系数, 根据偏态和峰度系数的差异, 区别目标和云层。实验结果证明, 该方法能有效地区别目标和云层边缘干扰, 从而降低目标检测算法的虚警率。
Abstract
Aiming at the false alarm from the cloud edge when detecting infrared point targets in sky background, this paper proposed the method of Gaussian quadrics fitting according to the point targets’ Gaussian distributing characteristic. Firstly, the targets and the cloud edge were got by traditional arithmetic of targets detection. Then the parameters of Gaussian quadrics fitting were obtained. Finally the skewness coefficient and kurtosis coefficient were calculated. Then the targets and cloud edge could be differentiated based on the difference of the skewness coefficient and kurtosis coefficient. The experimental results show that this method of Gaussian quadrics fitting can effectively differentiate the targets and cloud edge, so can reduce the false alarm probability.
李国强. 高斯曲面拟合在对空红外点目标检测中的应用[J]. 红外技术, 2013, 35(10): 638. LI Guo-qiang. Application of Gaussian Quadrics Fitting in the Infrared Point Targets Detection in Sky Background[J]. Infrared Technology, 2013, 35(10): 638.