A color saturation adjustment algorithm of video coding in MPEG-2 is given in this paper. This algorithm divides the color difference signals of current frame into four color saturation regions of very high

high

low

very low region by using neural networks and derive the modification factor to adjust the color saturation of the current frame. Simulation results indicate that this method can improve the subjective and objectiv equality of video decoding frame in terms of decoding peak signal-to-noise ratio and color saturation.