Abstract

An image-based gesture recognition module has been developed as a part of biped walking humanoid robot ASIMO’s man-machine interaction function in order to enhance its autonomy. ASIMO recognizes human gestures based on statistical analysis, using spatiotemporal variance of three-dimensional (3D) positions of the face and hand. It is able to distinguish 10 kinds of gestures and postures on left and right side each. It is also able to recognize a position pointed by an operator from the relative position of operator’s head and hand and from the direction of the arm. This is calculated by PCA (Principal Component Analysis) on a 3D-point set observed in the proximity of his arm. These components are implemented in a layered structure to reduce complexity, and in optimally designed communication structure to minimize processing resources. This effort has resulted in a comprehensive image processing system in ASIMO’s onboard computer.

References

1. (1) Hirose, M., Takenaka, N.: Development of Humanoid Robot ASIMO, Honda R&D Technical Review, Vol. 13, No. 1, p. 1-6 (2001)
2. (2) Sakagami, Y., Watanabe, R., Aoyama, C., Matsunaga, S., Higaki, N., Fujimura, K.,: The intelligent ASIMO : System overview and integration, IROS, p. 2478-2483 (2002)
3. (3) Yachida, M., Iwai, Y.: Human Motion Analysis and Recognition for Interaction, Technical Report of IEICE, PRMU99-142 (1999)
4. (4) Takahashi, M., Ogura, N., Watanabe, S.: Recognition of mixture motions using Hidden Markov Model, Technical Report of IEICE, PRMU 2000-151 (2001)
5. (5) Takahashi, K., Sakaguchi, T., Ohya, J.: Remarks on a Real-Time, Non-contact, Non-wear, 3D Human Body Posture Estimation Method, Trans. IEICE (D-II), Vol. J83-D-II, No. 5, p. 1305-1314 (2000)
6. (6) Ishii, H., Mochizuki, K., Kishino, F.: A Motion Recognition Method from Stereo Images for HumanImage Synthesis, Trans. IEICE (D-II), Vol. J76-D-II, No. 8, p. 1805-1812 (1993)
7. (7) Takahashi, K., Seki, S., Kojima, H., Oka, R.: Spotting Recognition of Human Gestures from Time-Varying Images, Trans. IEICE (D-II), Vol. J77-D-II, No. 8, p. 1552-1561 (1994)
8. (8) Ballard, D., Brown, C.: Segmented Images, Computer Vision, p. 147-205 (1987)
9. (9) Kass, M., Witkin, A., Terzopoulos, D.: Snakes: Active Contour Models, Proc.of 1st ICCV, p. 259-268 (1987)
10. (10) Muller, S., Eickeler, S., Rigoll, G.: Crane Gesture Recognition Using Preudo 3-D Hidden Markov Models, In 4th IEEE International Conference on Automatic Face and Gesture Recognition, p. 398-402 (2000)
11. (11) Kanatani, K.: Bayesian estimation, p. 149-150 (1995)

Author (organization or company)

• Nobuo HIGAKI（Wako Research Center）
• Taro YOKOYAMA（Wako Research Center）
• Kikuo FUJIMURA（Honda Research Institute USA, Inc）

Publication Information

Article of Honda R&D Technical Review Vol.15 No.1 (95 - 102)

PDF

Citation Information