一种基于改进联合双边滤波的雷达图像去噪方法  

A method of radar image denoising based on improved joint bilateral filtering

作　　者：孟凡 贾倩茜 杨光 MENG Fan;JIA Qian-qian;YANG Guang(No. 8 Research Academy of CSSC, Nanjing 211153)

机构地区：[1]中国船舶集团有限公司第八研究院,南京211153

出　　处：《雷达与对抗》2021年第4期24-27,共4页Radar & ECM

摘　　要：提出一种基于改进联合双边滤波的雷达图像去噪方法,该方法把雷达幅度数据映射成为图像数据,从图像处理的角度引入雷达多普勒速度作为引导图,通过计算此引导图的全局阈值,并以此阈值作为门限,与原始幅度数据相结合,计算值域权重得到值域核,与空间核相乘得到卷积模版,从而实现检测区域增强。经实验验证,该方法可以有效保护目标边缘并弱化噪声,显著提高信噪比。A radar image denoising method is proposed based on improved joint bilateral filtering.First,radar amplitude data are mapped into image data.From the perspective of image processing,radar Doppler velocity is introduced as a guidance map,and the global threshold of the guidance map is calculated.Combined with the original amplitude data,the range weight is calculated to obtain the range kernel.The convolution template is obtained by multiplying the spatial kernel to enhance the detection area.The test results show that the algorithm can effectively protect target edges,weaken the noise,and improve the SNR.

关 键 词：图像去噪 联合双边滤波 多普勒速度 目标边缘 

分 类 号：TN957.52[电子电信—信号与信息处理]