TY - JOUR
T1 - An efficient walkthrough from two images using view morphing and spidery mesh interface
AU - Cho, Hang Shin
AU - Kim, Chang Hun
AU - Nishihara, Seiichi
PY - 2000
Y1 - 2000
N2 - This paper presents an efficient walkthrough animation with two images of the same scene. To make animation easily and fast, view morphing uses only 2D transition but restricts its camera path on the line between two views, and Tour Into the Picture(TIP) provides the spidery mesh interface to recover simplified 3D structure from single image but lacks the reality of its foreground object in parallel moving of the viewpoint. By combining advantages of these two image-based techniques, this paper suggests a new virtual navigation technique which enables natural scene transformation when the viewpoint changes in the parallel direction as well as in the depth direction. In our method, view morphing is employed only in foreground objects, and background scene perceived carelessly is mapped into cube-like 3D model as in TIP, so as to save detail 3D reconstruction cost and improve visual realism simultaneously. To do this, we define a new camera transformation between two images from the relationship of spidery mesh transformation and its corresponding view change. The result animation demonstrates the efficiency of our method.
AB - This paper presents an efficient walkthrough animation with two images of the same scene. To make animation easily and fast, view morphing uses only 2D transition but restricts its camera path on the line between two views, and Tour Into the Picture(TIP) provides the spidery mesh interface to recover simplified 3D structure from single image but lacks the reality of its foreground object in parallel moving of the viewpoint. By combining advantages of these two image-based techniques, this paper suggests a new virtual navigation technique which enables natural scene transformation when the viewpoint changes in the parallel direction as well as in the depth direction. In our method, view morphing is employed only in foreground objects, and background scene perceived carelessly is mapped into cube-like 3D model as in TIP, so as to save detail 3D reconstruction cost and improve visual realism simultaneously. To do this, we define a new camera transformation between two images from the relationship of spidery mesh transformation and its corresponding view change. The result animation demonstrates the efficiency of our method.
UR - http://www.scopus.com/inward/record.url?scp=33750924443&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33750924443&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:33750924443
VL - 15
SP - 139
EP - 142
JO - Proceedings - International Conference on Pattern Recognition
JF - Proceedings - International Conference on Pattern Recognition
SN - 1051-4651
IS - 3
ER -