Robotic Manipulation

Probabilistic Multi-View Fusion  of Active Stereo Depth Maps

fusion.png

Jun Yang, Dong Li and Steven L. Waslander. 

IEEE Robotics and Automation Letters (RA-L), 2021

The reliable fusion of depth maps from multiple viewpoints has become an important problem in many 3D reconstruction pipelines. In this work, we investigate its impact on robotic bin-picking tasks such as 6D object pose estimation. In this work, we propose a novel probabilistic framework for scene reconstruction in robotic bin-picking. Based on active stereo camera data, we first explicitly estimate the uncertainty of depth measurements, and incorporated into a probabilistic model for incrementally updating the scene. To extensively evaluate different fusion methods, we constructed ROBI (Reflective Objects in BIns), a multi-view dataset consisting of reflective objects in bin-picking scenes. ​The ROBI dataset will be released.

ROBI Dataset

  • The dataset is available here.