Treatment of systematic errors II: Fusion of ultrasound and visual sensor data
In this work we present a methodology for the fusion of ultrasound and visual sensor data as acquired by a mobile robot. The objective of the methodology was the reduction of systematic errors which arise in the processing of the data in the individual sensor domains. In the initial processing of the ultrasound scan, rectilinear (Cartesian map) and polar (strings) data structures were built. In the initial processing of the CCD camera image, vertical edge segments were identified and labeled according to their connectivity. The systematic errors treated included ultrasound distortions in size, and visual ambiguities in discriminating depth discontinuities from intensity gradients generated by other details in the image. These systematic errors were first flagged by comparing the ultrasound strings and visual vertical edges to one another. The ranges, spatial orientation of the camera, and geometric information extracted from least-squares fits were then used in the fusion stage processing of the visual image. Vertical edge information was used in the subsequent fusion stage processing of the ultrasound data. By the end of this feedback-like fusion process the data structures in each sensor domain carried some information from the other domain. We had identified the vertical edges of interest, tagged them with range information, and removed the distortions from the Cartesian navigation maps. 32 refs., 11 figs.
- Research Organization:
- Oak Ridge National Lab., TN (USA)
- Sponsoring Organization:
- DOE/ER; DOE/NE
- DOE Contract Number:
- AC05-84OR21400
- OSTI ID:
- 6697953
- Report Number(s):
- ORNL/TM-11349; CESAR-89/38; ON: DE90014780
- Country of Publication:
- United States
- Language:
- English
Similar Records
World modelling and multi-sensor integration for a mobile robot
Performance of visual and ultrasound sensing by an autonomous robot
Related Subjects
ROBOTS
DATA PROCESSING
DATA ACQUISITION
ERRORS
IMAGE PROCESSING
LEAST SQUARE FIT
NAVIGATION
STATISTICS
ULTRASONIC WAVES
VISION
MATHEMATICS
MAXIMUM-LIKELIHOOD FIT
NUMERICAL SOLUTION
PROCESSING
SOUND WAVES
990200* - Mathematics & Computers