Machine-Awareness in Indoor Environment: A Pseudo-3D Vision-Based Approach Combining Multi-Resolution Visual Information
Résumé
The present paper describes a dual approach using pseudo-3D vision for Machine-Awareness in indoor environment. Provided by color and depth cameras of the a Kinect system, the aforementioned duality presents an appealing solution for robots' 3D-vision. Placing the human-robot and in a more general way the human-machine interactions as a key outcome of the expected visual Machine-Awareness, the proposed vision-system aims proffering the machine the self-reliance in awareness about the surrounding environment in which the machine is supposed to evolve. Blend pseudo-3D vision and salient objects' detection algorithm, the investigated approach seeks an autonomous detection of relevant items in 3D environment. The pseudo-3D perception leads to reducing computational complexity inborn to the 3D vision context into a 2D computational task by processing 3D visual information within a 2D-images' framework. The statistical foundation of the investigated approach proffers it a solid and comprehensive theoretical basis, holding out a bottom-up nature making the issued system unconstrained regarding prior hypothesis. We provide experimental results validating the proposed system.