在低信噪比及背景干扰下红外图像点目标和面目标的检测

Detection of Point Targets and Regional Targets of Infrared Images on the Condition of Low SNR and Background Disturbing

红外寻的导引头点目标及面目标的检测是地空导弹和航空导弹的关键技术。文中提出了高斯核自适应平滑滤波,并在此基础上依据点目标和面目标的边缘灰度值高及变化大的特点设计出提取方法,然后用区域增长恢复没有被提取出来的面目标的内部。实验表明该方法是正确和有效的。

The detection of small target image for infrared homing guidance head are the crucial techniques for ground-air and surface-air missiles. A Gaussian kernel adaptive smooth filter is proposed, on the basis of which a scheme is designed to extract information of targets according to the feature that the gray level of point targets and the edge of regional targets are high and vary greatly. Then a region growth method is used to restore the internal part of the regional targets not being extracted.