基于纹理特征的SAR图像水上桥梁目标检测  

Target detection of bridge over the water in SAR images based on texture feature

作　　者：童涛[1] 杨桄[1] 崔震[1] 叶怡[1] 王寿彪[1] 

机构地区：[1]空军航空大学特种专业系,长春130022

出　　处：《中国体视学与图像分析》2012年第2期107-111,共5页Chinese Journal of Stereology and Image Analysis

基　　金：国家自然科学基金项目(40901096)

摘　　要：针对SAR影像进行水上桥梁目标检测问题,提出了一种纹理特征与关联特征相结合的方法。该方法首先通过去噪及增强预处理改善图像质量,并采用阈值分割方法快速提取初始目标;然后使用灰度共生矩阵对目标及其周边背景纹理进行差异对比分析,从而去除虚警目标。实验结果表明,该方法能够准确地检测出水上桥梁目标。A combined method is presented to detect bridge target in SAR images using the texture features and referencing features. Firstly, de-noising and enhancement preprocessing operators are used to improve the image quality, and threshold method is adopted to extract initial objects; And then , gray co- occurrence matrix is used to differentiate the textures of objects and background; Finally, reference fea- tures of bridge are used to get rid of the clutter. Experimental results show that the method can detect the bridge target precisely.

关 键 词：SAR图像 灰度共生矩阵 纹理特征 桥梁检测 

分 类 号：TN957.52[电子电信—信号与信息处理]