The robust tracking in the abrupt motion is a challenging task in computer vision due to its large motion uncertainly. While various tracking method by using particle filters and by using Markov-Chain Monte Carlo(MCMC) method have been proposed for visual tracking, these are often suffer from local-trap problem. In this paper, we introduce the Stochastic Approximation Monte Carlo (SAMC) sampling method into the Bayesian filter tracking framework to effective handling the local-trap problem. In addition to improve the sampling efficiency, we propose a new MCMC sampler with intensive adaptation. This is done by combining the SAMC sampling with a density-grid-based predictive model. The proposed method is effective and computationally efficient in addressing the abrupt motion problem. The proposed method was named as IA-MCMC.