Abstract: A method of ICT image sequence segmentation based on Spatial Temporal Markov field model was presented. According to the spatial correlation of the ICT image sequence, the Markov random field(MRF) model based on spatial-temporal neighborhood system was proposed and the cost function of corresponding Gaussian mixture model was constructed. Then the maximum a posteriori(MAP) estimation was fulfilled by using the iterated conditional model(ICM) algorithms. The experimental results show that the proposed method is suitable, accurate and effective for ICT image sequence segmentation.