Blind Image Denoising via Dependent Dirichlet Process Tree
IEEE Transactions on Pattern Analysis and Machine Intelligence(2017)
摘要
Most existing image denoising approaches assumed the noise to be homogeneous white Gaussian distributed with known intensity. However, in real noisy images, the noise models are usually unknown beforehand and can be much more complex. This paper addresses this problem and proposes a novel blind image denoising algorithm to recover the clean image from noisy one with the unknown noise model. To mod...
更多查看译文
关键词
Noise measurement,Image denoising,Bayes methods,Noise reduction,Mixture models,Gaussian distribution,Data models
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络