Vossiek M, Magori V, Ermert H (1994)
Publication Language: English
Publication Type: Conference contribution
Publication year: 1994
Publisher: IEEE
Edited Volumes: Proceedings of the IEEE Ultrasonics Symposium
City/Town: Piscataway, NJ, United States
Book Volume: 2
Pages Range: 1293-1297
Conference Proceedings Title: IEEE International Ultrasonics Symposium
Event location: Cannes, France
DOI: 10.1109/ULTSYM.1994.401820
This paper describes an ultrasonic sensor system for application in air that is capable of recognizing workpieces with size up to 20 mm, independently of their position and orientation, with high reliability. The 4x4 Transducer Matrix signals are first processed with a simplified reconstruction algorithm that leads to position invariant axial profiles of the object reflectivity. A broadband holographic reconstruction is then performed for those object planes that are reflecting the main signal energy. The Zernike Moments of the resulting cross-sectional patterns and features of the reflectivity profiles are fed in a 3-layered perceptron network. The system has been tested with a set of measuring data from 25 typical workpieces. Even though some objects have very similar reflection properties, it is possible to identify the objects in an area of 25x25 mm2 independently of their position and orientation with a recognition rate higher than 99%.
APA:
Vossiek, M., Magori, V., & Ermert, H. (1994). An Ultrasonic Multielement Sensor System for Position Invariant Object Identification. In IEEE International Ultrasonics Symposium (pp. 1293-1297). Cannes, France: Piscataway, NJ, United States: IEEE.
MLA:
Vossiek, Martin, V. Magori, and Helmut Ermert. "An Ultrasonic Multielement Sensor System for Position Invariant Object Identification." Proceedings of the Proceedings of the 1994 IEEE Ultrasonics Symposium. Part 1 (of 3), Cannes, France Piscataway, NJ, United States: IEEE, 1994. 1293-1297.
BibTeX: Download