We discuss a linear image restoration algorithm using hidden Markov tree(HMT) model in complex wavelet domain based on the dual-tree complex wavelet transform(DT-CWT) and adopt a feasible and fast algorithm which avoids the time-consuming training process to estimate the HMT model parameters.The proposed fast restoration algorithm can retrieve various kinds of edges,and the restored results are improved to some extent than other some traditional restoration methods,and the running time is reduced significantly than the algorithms based on traditional HMT parameters estimation methods.