Research On Video Perceptual Characteristic Parameter Measurement And Distortion Model

Traditional video coding technology has been difficult to improve coding efficiency,and the final destination of the video is the human eye and the brain,then we can improve the coding technology from the direction of visual perception.This paper proposes a distortion model based on spatio-temporal perceptual characteristics,which provides theoretical and model support for the optimization of perceptual video coding algorithms.The principal research contents are summarized as follows:1.Using the foveated masking and visual attention of HVS,this paper proposes an adaptive foveated weighting model to improve the conventional foveated JND(FJND)model.The proposed FJND model overcomes two limitations of the conventional FJND models.On the one hand,conventional FJND models can't get accurate positions of fixation points.This paper proposes a fixation prediction algorithm based on the distribution of cones to identify fixation points and regions.On the other hand,conventional FJND models did not incorporate the visual effect.By fully exploiting the visual attention and foveated masking effects,this work proposes an adaptive foveated weighting model,which is developed as a function of the fixation intensity and retinal eccentricity.Here,the fixation intensity is estimated from the saliency map which is modelled by the Gaussian Mixed Model.With the proposed weight model,a new foveated JND profile is developed.Experimental results show that the proposed model tolerates more distortion at the same perceptual image quality compared with other JND models.2.Temporal HVS characteristics are not fully exploited in conventional JND models.In this paper,we improve the spatio-temporal JND model by fully leveraging the temporal HVS characteristics.From the viewpoint of visual attention,we investigate two related factors,positive stimulus saliency and negative uncertainty.This paper measures the stimulus saliency according to two stimulus-driven parameters,relative motion and duration along the motion trajectory,and measures the uncertainty according to two uncertainty-driven parameters,global motion and residue intensity fluctuation.These four different parameters are measured with self-information and information entropy,and unified for fusion with homogeneity.As a result,a novel temporal JND adjustment weight model is proposed.Finally,we fuse the spatial FJND model and temporal JND weight to form the spatio-temporal JND model.The experiment results verify that the proposed JND model yields significant performance improvement with much higher capability of distortion concealment compared to state-of-the-art JND models.