遥感图像常用去噪方法

The research of remote sensing image denoising methods

遥感图像在获取和传输的过程中,受各种噪声影响,使图像的边缘纹理等细节模糊,质量降低。为获得清晰的、高质量的遥感图像必须进行降噪预处理。该文就遥感图像去噪的邻域平均法、中值滤波法、维纳滤波法及小波变换法算法原理进行了研究和比较分析并进行了仿真实验。结果表明:对受不同噪声影响的遥感图像选择不同滤波算法均能取得较好的效果,但在噪声模型未知的情况下,小波去噪效果更佳。

In the acquisition and transmission process of remote sensing images, all kinds of noise made the images edge details fuzzy, this lower images quality. In order to get a clear, high quality remote sensing images, de-noising must be conducted. In this paper, we studied the remote sensing image de-noising principle such as neighborhood average, median filtering, wiener filtering and wavelet transform method , and carried out the comparative analysis and the simulation experiment. The results shows that, for the noised remote sensing images select different filtering algorithm can obtain a better result, but when the noise model is unknown, wavelet transform method denoising effect is better.