Shape Reconstruction Using a Mirror with a Range Image Sensor – Detection of Mirror in a Range Image
In this paper, a method to reconstruct the whole shape of an object using a mirror with a range image sensor is examined. A method to obtain the pose of the mirror from the floor plane reflected in the mirror in range image is proposed. The mirror pose is estimated from the floor plane in the range image and the floor plane reflected in the mirror. The shape of the back of the object is measured from the reflected range image. Coordinate transformation is applied to the reflected back shape using the estimated mirror pose, and the whole shape of the object is obtained by registration of front and back range images. Most previous methods to obtain whole shape of an object use multiple range image sensors or use one range image sensor and moves the sensor or uses a turntable to acquire multiple range images. Compared to these methods, the proposed method does not require exact calibration in advance or precise registration method and thus is easy to use. The effectiveness of the methods is verified by experiments to estimate the mirror pose and reconstruct shape of an object using the estimated mirror plane.
KeywordsRange image measurement Mirror Three-dimensional shape reconstruction
- 1.Kim, M.-H., Park, J., Kweon, I.S.: RGBD sensor and mirrors: a practical setup for 3D reconstruction of dynamic objects. In: Proceedings of 2014 11th International Conference on Ubiquitous Robots and Ambient Intelligence (URAI), pp. 651–652 (2014)Google Scholar
- 2.Akay, A., Akgul, Y.S.: 3D reconstruction with mirrors and RGB-D cameras. In: Proceedings of 2014 International Conference on Computer Vision Theory and Applications (VISAPP), pp. 325–334 (2014)Google Scholar
- 3.Fasano, A., Callieri, M., Cignoni, P., Scopigno, R.: Exploiting mirrors for laser stripe 3D scanning. In: Proceedings of 4th International Conference on 3-D Digital Imaging and Modeling (3DIM), p. 243 (2003)Google Scholar