Title :
Neural-network-based adaptive sampling of three-dimensional-object surface elastic properties
Author :
Cretu, Ana-Maria ; Petriu, Emil M.
Author_Institution :
Sch. of Inf. Technol. & Eng., Univ. of Ottawa, Canada
fDate :
4/1/2006 12:00:00 AM
Abstract :
The paper discusses an adaptive-sampling technique for dimensionality reduction of the set of probing points in the measurement of nonuniform elastic properties of three-dimensional (3-D) objects. Two self-organizing neural-network architectures are compared for this purpose: the neural-gas network and the Kohonen self-organizing map (SOM).
Keywords :
adaptive systems; elasticity; self-organising feature maps; signal sampling; 3D object surface; Kohonen self-organizing map; adaptive systems; neural-gas network; neural-network-based adaptive sampling; nonuniform elastic properties; probing points; robot tactile systems; self-organizing neural-network architectures; Application software; Computer aided manufacturing; Design automation; Elasticity; Industrial training; Information technology; Probes; Robotic assembly; Robots; Sampling methods; Adaptive systems; neural-network applications; robot tactile systems; self-organizing feature maps; topology;
Journal_Title :
Instrumentation and Measurement, IEEE Transactions on
DOI :
10.1109/TIM.2006.870114
