A New Algorithm for Extracting High-Resolution Face Image from Video Sequence

How to get high-resolution face image from low-resolution video sequence with time-saving demand is an important problem. In this paper, a new framework is described to extract high-resolution face image from video sequence. Firstly, the real-time face detection based on AdaBoost algorithm and Haar like feature is implemented. The detected face area is segmented from the background to form a face image sequence of the same person with same pose. Secondly, the cubic spline interpolation and Parzen window based mutual information (MI) is used for image registration. Finally, a modified iteration algorithm for super-resolution image restoration is proposed as an extension of Irani and Peleg's method. Experimental results demonstrate the good performance of the proposed algorithm