A robust image mosaicing technique capable of creating integratedpanoramas

摘要

Existing featureless image mosaicing techniques do not pay enoughattention to the robustness of the image registration process, and arenot able to combine multiple video sequences into an integratedpanoramic view. These problems have certainly restricted applications ofthe existing methods for large-scale panorama composition, video contentoverview and information visualization. In this paper we propose amethod that is able to create an integrated panoramic view for a virtualcamera from multiple video sequences which each records a part of a vastscene. The method further enables the user to visualize the integratedpanoramic view from an arbitrary viewpoint and orientation by alteringthe parameters of the virtual camera. To ensure a robust and accuratepanoramic view synthesis from long video sequences, we attach a globalpositioning system (GPS) to the video camera, and utilize its outputdata to provide initial estimates for the camera's translationalparameters, and to prevent the camera parameter recovery process fromfalling into spurious local minima. Our proposed method is not onlysuitable for video content overview but also applicable to the areas ofinformation visualization, team collaborations, disastrous rescues, etc.The experimental results demonstrate the effectiveness of the proposedmethod