高建, 张斌, 张飞艳, 秦前清. 一种遥感影像Min/max流去噪方法的研究[J]. 武汉大学学报 ( 信息科学版), 2012, 37(1): 31-34.
引用本文: 高建, 张斌, 张飞艳, 秦前清. 一种遥感影像Min/max流去噪方法的研究[J]. 武汉大学学报 ( 信息科学版), 2012, 37(1): 31-34.
GAO Jian, ZHANG Bin, ZHANG Feiyan, QIN Qianqing. Surface Min/max Flow Method for Remote Sensing Image Denoising[J]. Geomatics and Information Science of Wuhan University, 2012, 37(1): 31-34.
Citation: GAO Jian, ZHANG Bin, ZHANG Feiyan, QIN Qianqing. Surface Min/max Flow Method for Remote Sensing Image Denoising[J]. Geomatics and Information Science of Wuhan University, 2012, 37(1): 31-34.

一种遥感影像Min/max流去噪方法的研究

Surface Min/max Flow Method for Remote Sensing Image Denoising

  • 摘要: 提出一种曲面Min/max流去噪方法,利用图像曲面曲率代替经典方法曲线曲率并修改开关机制,增强对复杂噪声环境的适应性,同时加入梯度权控制边缘曲面演化速度,在噪声压制的同时尽量保持边缘。实验表明,该方法能够在去除噪声的同时保持比较清晰的图像边缘。

     

    Abstract: As image surface is introduced for the surface Min/max flow denoising method,the curvature of isolux curve in classical Min/max flow is replaced by the one of image surface and improves switching mechanism for better adaptability on complex noise,while adding gradient weight to control evolving speed of image surface for better edge preserving.The proposed method makes Center-Weighted Mean filtering as preprocessing in order to eliminate remarkable impulse noise which could make inefficiency in Min/max flow.In the denoising experiment of Remote Sensing image with artificial noise,the resultant image indicate that the proposed method removes noise effectively while preserving edge well,and the statistical data for PSNR,SSIM and detail preserving index is good or better than some other popular denoising methods such as BM3D,wavelet.

     

/

返回文章
返回