Robot vision: real time identification of simulated machine parts using features modelled from the radon space

摘要

This paper proposes a new technique that can be used to extract features in 2-Dimensional Radon space suitable for implementing fast visual systems in robots. The approach makes use of fast Radon transforms to derive features from simulated machine parts. Features so derived are highly compressed and represent simulated objects in such a way that they can be applied to a back propagation network for classification. The new approach is shown to be simple and robust and provides a cost-effective method of designing visual systems for robots.