3D filtering color image contaminated by mixed noise using sparse representation

Filtering image has different applications, such as computer vision, multimedia tools, telemedicine, and satellite imaging, where the objective is restoring the lossed information due to the presence of noise. The noise present in an image can be modeled as a stochastic process. There are diverse reasons why the noise appears, such as: non-uniform lighting, random fluctuations in an object's surface orientation and texture, sensor limitations, non-ideal transmission, and interference. Noise affects not only the performance of an image in a specific problem but also its perceived quality. In this paper, a novel framework is presented for denoising colour images corrupted by a mixture of additive and impulsive noise. The proposed method could be described in three stages: 1)impulsive noise filtering; 2)additive noise filtering; and 3)post-processing. In the first stage, a pixel contaminated by impulsive noise should be detected, the detection is performed using the values of the pixels in local and interchannel way. Also, the restoration of corrupted pixels is performed using a filter based on the summing of distance vectors. In the next stage, filtering of additive noise is based on behavior of the additive noise on the discrete cosine transform domain, sparse representation and 3D-processing. Finally, the post-processing stage increases filtering quality using a Wiener filter.