Authors:
Rui Nóbrega
and
Nuno Correia
Affiliation:
Universidade Nova de Lisboa, Portugal
Keyword(s):
Computer Vision, Augmented Reality, Computer Graphics.
Related
Ontology
Subjects/Areas/Topics:
Augmented, Mixed and Virtual Environments
;
Computer Vision, Visualization and Computer Graphics
;
Interactive 3D Graphics for Mobile Devices Like Smartphones, PDAs and UMPCs
;
Interactive Environments
;
Non-Desktop Interfaces
Abstract:
This paper proposes a framework for the creation of interactive multimedia applications that take advantage of detected features from user-captured photos. The goal is to create games, architectural and space planning applications that interact with visual elements in the images such as walls, floors and empty spaces. The framework takes advantage of a semi-automatic algorithm to detect scene elements and camera parameters. Using the detected features, virtual objects can be inserted in the scene. In this paper several example applications are presented and discussed, and the reliability of the detection algorithm is compared with other systems. The presented solution analyses the photos using graph-cuts for segmentation, vanishing point detection and line analysis to detect the scene elements. The main advantage of the proposed framework is the semi-automatic creation of the tri-dimensional model to be used in mixed reality applications. This enables scenarios where the user can be
responsible for the input scene without much prior knowledge or experience. The current implemented examples include a furniture positioning system and a snake game with a user-built maze in the real world. The proposed system is ideal for multimedia mobile applications where interaction is combined with the back camera of the device.
(More)