In this talk I will present the main tools for the removal of noise in still image and image sequences.
I will briefly present the main steps of the imaging chain applied by a camera and how they affect the noise of the acquired data. The used denoising strategy might depend on which stage of the imaging/video chain we have access to the data. For example, reflex cameras permit to access the RAW acquired at the sensor while usual cameras and mobile phones give access only to the final processed and compressed image or video.
State of the art techniques dealing with real image and video denoising need of multi-scale methods, patch based models, transform thresholding and motion estimation for image sequences. Examples with real images will illustrate the quality gain obtained by using these techniques.