6D pose estimation of textureless shiny objects using random ferns for bin-picking

We address the problem of 6D pose estimation of a textureless and shiny object from single-view 2D images, for a bin-picking task. For a textureless object like a mechanical part, conventional visual feature matching usually fails due to the absence of rich texture features. Hierarchical template matching assumes that few templates can cover all object appearances. However, the appearance of a shiny object largely depends on its pose and illumination. Furthermore, in a bin-picking task, we must cope with partial occlusions, shadows, and inter-reflections. In this paper, we propose a purely data-driven method to tackle the pose estimation problem. Motivated by photometric stereo, we build an imaging system with multiple lights where each image channel is obtained under different lightning conditions. In an offline stage, we capture images of an object in several poses. Then, we train random ferns to map the appearance of small image patches into votes on the pose space. At runtime, each patch of the input image votes on possible pose hypotheses. We further show how to increase the accuracy of the object poses from our discretized pose hypotheses. Our experiments show that the proposed method can detect and estimate poses of textureless and shiny objects accurately and robustly within half a second.