Simultaneous dense scene reconstruction and object labeling

This paper presents an efficient system for simultaneous dense scene reconstruction and object labeling in real-world environments (captured with an RGB-D sensor). The proposed system starts with the generation of object proposals in the scene. It then tracks spatio-temporally consistent object proposals across multiple frames and produces a dense reconstruction of the scene. In parallel, the proposed system uses an efficient inference algorithm, where object class probabilities are computed at an object-level and fused into a voxel-based prediction hypothesis modeled on the voxels of the reconstructed scene. Our extensive experiments using challenging RGB-D object and scene datasets, and live video streams from Microsoft Kinect show that the proposed system achieved competitive 3D scene reconstruction and object labeling results compared to the state-of-the-art methods.

Authors

• Umar Asif
• Mohammed Bennamoun
• Ferdous Ahmed Sohel

Keywords

• Three-dimensional displays
• Image reconstruction
• Semantics
• Labeling
• Proposals
• Feature extraction
• Streaming media
• Labeling Density
• Object Labels
• Dense Objects
• Scene Reconstruction
• Dense Reconstruction
• Simultaneous Reconstruction
• Dense Scenes
• Dense Scene Reconstruction
• Efficient Algorithm
• 3D Reconstruction
• Object Classification
• Real-world Environments
• Multiple Frames
• Live Streaming
• Inference Algorithm
• Object Proposals
• Convolutional Neural Network
• Feature Representation
• Kernel Function
• Dimensional Vector
• Point Cloud
• Simultaneous Localization And Mapping
• Surface Normals
• Object Recognition
• Conditional Random Field
• Stream Of Work
• Semantic Map
• Bag-of-words
• Iterative Closest Point
• RGB-D Images