Abstract

In traditional panoramic photography, moving objects or as referred in this paper - the strangers, in the background should be eliminated since those strangers obscure the scenery that we want to retain in our photograph. This paper discusses a novel method to remove the strangers whose face data is not available in the face database of camera from the background of the focused area and to compose a panoramic image. In the proposed system the object segmentation, motion inpainting of background and, effective panoramic creation steps are used to compose a panorama. If the face data of the stranger is not in the face database of the camera, stranger is automatically removed from the panorama during the step of background repair which is known as inpainting. Panorama creation of the developed system is fully automatic except that the user required marking the unidentified moving objects in the object segmentation phase. Experimental results demonstrate that the proposed system can be effectively used to generate panorama photographs without appearing moving objects in the background.

Operation of the proposed tool
Stranger detection and segmentation
Face recognition and GrabCut algorithm is used to detect and segment strangers
Background inpainting

Video inpainting is used to inpaint the disoccluded background information in the current frame from the background information in the frames prior or after.

Regular video inpainting is not suitable for the proposed application as the video follows a circular tragectory, hence CDHS is used to calculate the movement direction of relevant frames and disoccluded gaps are inpainted.

Selected results
Sample 01
Sample 02
Sample 03
Sample 04
Sample 05
Copyright 2014 - Shih-Ming Chang and MINElab, Tamkang University and National Central University