Stabilization of Vibration on Archive Films by Automatic Multi-scale Rof Selection

The cost of human operation is a bottleneck in the restoration of large quantities of archive films. Eliminating the vibration of a degraded image makes a difficulty in an automatic film restoration system. The vibration is usually caused by improper film transportation during the recording, copying or the digitization process. We have developed an automatic method for image stabilization consisting of two main steps: estimating vibration then correction by drifting and rotating the whole frame. Earlier stabilization algorithms are unsuccessful in cases of multiple motions and human interaction with manual ROF (Region of Fixation) selection is needed. In contrast, the proposed method, which is based on multi-scale motion estimation, automatically selects the ROF required for the stabilization of noisy films.