Robust method of digital image stabilization

Abstract

A relative motion between the camera and the scene during the image integration time results in a degradation of the image quality, known as motion blur. The objective of image stabilization is either to prevent or to remove this type of image degradation. In this paper we propose an image stabilization method based on fusing multiple short exposed image frames of the same scene. Due to their short exposure, the individual frames are corrupted by sensor noises (e.g. photon-shot noise, readout noise), but, on the other hand, they are less affected by motion blur. By registering and fussing the individual frames we obtain the equivalent of a long exposed image of the scene without motion blur. The paper addresses the main requirements for each component of an image stabilization system, presenting our own solution. Various natural images acquired with camera phones have been used to evaluate the proposed image stabilization system. The results reveal the ability of the system to improve the image quality by simulating longer exposure times. In addition the system has the ability to reduce the effect of noise, out of focus blur, and outliers present in the individual short exposed frames.