Recognition and Pose Estimation of Primitive Shapes from Depth Images for Spatial Augmented Reality

In this paper, we propose a method for recognition and pose estimation of primitive shapes from depth images for spatial augmented reality (SAR). To use SAR in our everyday life, technology to recognize and estimate the pose of projected objects in the room is necessary. However, it is not a simple task to recognize primitive shapes because of their low 3D feature values. Hence, we focused on the gradient of normal vector map to extract surfaces and used the information of the surfaces of each object to recognize target objects. With our method, it becomes possible to recognize and estimate the pose of target objects in various scenes. Additionally, we projected an image onto each of the surfaces of the physical objects.