PROJECT TITLE :

CPA-SLAM: consistent plane-model alignment for direct RGB-d slam - 2016

ABSTRACT:

Planes are predominant options of artificial environments which have been exploited in several mapping approaches. During this paper, we tend to propose a real-time capable RGB-D SLAM system that consistently integrates frame-to-keyframe and frame-to-plane alignment. Our method models the environment with a global plane model and - besides direct image alignment - it uses the planes for tracking and world graph optimization. This way, our technique makes use of the dense image data on the market in keyframes for correct short-term tracking. At the identical time it uses a global model to scale back drift. Each elements are integrated consistently in an expectation-maximization framework. In experiments, we demonstrate the advantages our approach and its state-of-the-art accuracy on difficult benchmarks.