科研之友
微信
新浪微博
Facebook
分享链接
摘要
Images transmitted through wireless network or imaged by digital mobile devices usually contain a lot of noise against information acquisition. The improved nonlocal means (NLM) filtering algorithm proposed in this paper is suitable for digital mobile terminals in order to filter the noise mentioned above and obtain denoised images which details in them can be easily picked out for the field: the internet of things. It inherits some excellent ideas from several improved NLM denoising algorithms: abandon gaussian kernel function, utilize correlation of color components and compensate the pixels' hue for color denosing image. Besides above, we select an optimal value of the parameter h after experiments and analyzing a lot of data and we put forward a reprocessing mode based on Rough Set theory. The proposed method is intended to improve signal-to-noise ratio (SNR) of images and effectively solve the problem that the details in image are blurred. And the improved color-image denoising method can remove motle in the denoising image better than most state-of-the-art methods. The conclusion shows that it can preserve images' hue and details availably.
- 出版日期2018
- 单位上海理工大学
