：采用基于内容的图像检索方法，对雷达景象匹配数据库中的图像数据进行兴趣目标的查询检索与识别．由于

雷达图像具有受斑点噪音影响明显等特点，故综合运用迭代阈值选取以及区域生长的方法，进行兴趣目标(主要是

机场)的分离；由于匹配数据库中图像数据范围比较大，包含目标多，为了在检索过程中确定检索目标在图像中的

位置，预处理时，采用对同一图像多幅子图进行特征提取的方法，最终在检索时通过子图范围来确定目标在大幅图

像上的位置．由实验分析，证明在机场目标的检索识别中，这一方法是可行的．-Content—based image retrieval method is adopted for searching and recognizing interesting

targets of images in radar scene matching image Database in this paper．Radar image is disturbed by

speckle noise seriously，therefore alternating threshold choosing and area growing algorithms are applied

to segment interesting targets(such as airdrome)．Because the radar reference image covers a large area

and includes lots of targets a new method is proposed by dividing 1arge image into several sub-images

and then feature-extracting these sub-images to determine the position of the interested target in the

large image with the bound of sub-images in the whole image．This method WaS verified feasible in air—

drome target retrieval by experiments．