To improve the efficiency and perceptual quality in video coding,this paper proposes a model of just noticeable coding distortion (JNCD),which considering the human vis ual perception redundancy and the unreasonable factors of the existing just noticeable distortion (JND) models in codi ng process.First,we design a psychophysical experiment to analyze the just noticeable gradient difference (JNGD),and bui ld JNGD model to filter the gradient components which ar e imperceptible by human eyes.We use total variation (TV) to decompose an image i nto structural image and textural image, and calculating their gradient respectively,using JNGD t o filter out the imperceptible gradient components in each gradient image.Second,the human eye sensitivity of different gradient magnitudes is ana lyzed to model the relationship between the human eye perceptible gradient magnitude and JNCD.Finall y,considering the perceive difference of human eye in edge,flat and texture regions of an image,we adjust the JNCD value in each region and establish JNCD model of the whole im age.To verify the efficiency of the proposed JNCD model,we test the JNCD model on the high efficiency video coding (HEV C) platform.Compared with comparison schemes,the proposed model shows its advant ages in subjective visual effect,which means that it is helpful for analysis of human visual perception redundancy and the relevant perceptual video coding.