基于方向滤波器消除遥感图像孤立条带噪声的方法

REMOVAL OF RANDOM STRIPE NOISES IN REMOTE SENSING IMAGE BY DIRECTIONAL FILTER

条带噪声是影响遥感图像质量的重要因素之一.在分析遥感CCD图像孤立条带噪声产生原因和噪声模型的基础上,提出了基于二维方向滤波器抑制条带噪声的新方法.该方法利用了窄带方向滤波器良好的图像纹理方向的频率选择特性,将条带噪声与图像其它信息分离到不同的方向滤波器子带内,并在水平信息子带内采用均值补偿方法去除条带噪声.实验结果证明它不仅能有效去除CCD图像中的随机孤立条带噪声,还能保持原始图像的辐射入射幅度,保留细节信息,较传统的方法相比,性能有较大的提高.

Stripe noise is one of the most important factors which influence the qualities of remote sensing digital images. A new destriping method based on 2-D directional filtering was proposed based on the analysis of the main causes and characteristics of the stripe noises in remote sensing images. The directional filter can isolate the directional information of the stripe noise from the noised.image, so it is easy to uniform the gray value of the image along column by mean equalization. The results prove that this method can effectively remove the stripe noise while keep the detailed information of the original image.