Research On Video Denoising Algorithm Based On Temporal-spatial Domain

Noise in the digital video sequence does not only degrade the subjective visual quality of video, but also lead adverse effects to the post-processing such as compressing, encoding, transporting and storinging and so on . So the research on the denoising of image sequence has significant meaning and far-reaching influence on promoting the development of the whole digital media industry.On the basis of describering the types of noises in digital images and commonly used algorithms for video denoising, this paper proposes a new image quality assessment algorithm based on the human visual system, which can solve the problem that the objective evaluation of image quality is inconsistent with the subjective evaluation.Through the research of the common digital image and video denoising algorithms, a new video denoising algorithm based on temporal-spatial domain has been proposed. The algorithm makes use of the correlation characteristics of spatial domain and temporal domain of the image sequences. First, spatial filtering is made for the image sequences, and on this basis, the temproral filtering is carried out to get the final denoised image sequences. The spatial processing adopts direction spatial filtering algorithm based on human visual system. The spatial filtering template is got by the direction detection at first, and then the templates are adjusted by the HVS characteristics, so the filtering results according to the human subjective visual are achieved finally. In the temporal-domain filtering , kalman algorithm is adopted to remove noises.Finally, the proposed algorithm in this paper has been simulated, and the results show that it can remove the noise of video sequences effectively and improve the objective and subjective quality of the video images. .