An Active Robotic Vision System with a Pair of Moving and Stationary Cameras
Vision is one of the main potential sources of information for robots to understand their surroundings. For a vision system, a clear and close enough view of objects or events, as well as the viewpoint angle can be decisive in obtaining useful features for the vision task. In order to prevent performance drops caused by inefficient camera orientations and positions, manipulating cameras, which falls under the domain of active perception, can be a viable option in a robotic environment.
In this paper, a robotic object detection system is proposed that is capable of determining the confidence of recognition after detecting objects in a camera view. In the event of a low confidence, a secondary camera is moved toward the object and performs an independent detection round. After matching the objects in the two camera views and fusing their classification decisions through a novel transferable belief model, the final detection results are obtained. Real world experiments show the efficacy of the proposed approach in improving the object detection performance, especially in the presence of occlusion.
KeywordsActive perception Active vision Robotics PR2 Dual-camera Transferable belief model Dempster-Shafer Occlusion
This work has been supported by Office of Naval Research Award #N00014-16-1-2312.
- 4.Mattamala, M., Villegas, C., Yáñez, J.M., Cano, P., Ruiz-del-Solar, J.: A dynamic and efficient active vision system for humanoid soccer robots. In: Almeida, L., Ji, J., Steinbauer, G., Luke, S. (eds.) RoboCup 2015. LNCS (LNAI), vol. 9513, pp. 316–327. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-29339-4_26CrossRefGoogle Scholar
- 7.Dalal, N., Triggs, B.: Histograms of oriented gradients for human detection. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, San Diego (2005)Google Scholar
- 8.CIE 015:2018: Colorimetry. 4th edn., International Commission on IlluminationGoogle Scholar
- 9.Kalal, Z., Mikolajczyk, K., Matas, J.: Forward-backward error: Automatic detection and tracking failures. In: International Conference on Pattern Recognition, Istanbul (2010)Google Scholar
- 10.Hoseini A., S.P., Nicolescu, M., Nicolescu, M.: Active object detection through dynamic incorporation of Dempster-Shafer fusion for robotic applications. In: International Conference on Vision, Image and Signal Processing (ICVISP), Las Vegas (2018)Google Scholar
- 11.Hoseini A., S.P., Nicolescu, M., Nicolescu, M.: Handling ambiguous object recognition situations in a robotic environment via dynamic information fusion. In: IEEE Conference on Cognitive and Computational Aspects of Situation Management, Boston (2018)Google Scholar
- 14.Denoeux, T.: A neural network classifier based on Dempster-Shafer theory. IEEE Trans. Syst. Man Cybern. Part A Syst. Hum. 30(2), 131–150 (2000)Google Scholar
- 15.Blankenburg, J., et al.: A distributed control architecture for collaborative multi-robot task allocation. In: International Conference on Humanoid Robotics, Birmingham (2017)Google Scholar