Electrode Classification and Retrieval Using Supported Vector Machine
Due to the similarities between electrode model and their CNC machining process, process design could be finished efficiently using the electrode automatic classification system as well as the existed standard process template. This study developed an automatic classification retrieval system of electrode model by applying a statistical approach, namely SVM model, to the classification of electrode model, where 3D Polar-Radius Surface Moment was used to extract the feature vector of the electrode model. Experiments showed a promising result with an average classification accuracy up to 85.72% in addition to the high efficiency and usability. Most important, the developed approach is capable of reusing existing knowledge and experience and as a result it makes the CNC programming process easier.
Guojun Zhang and Jessica Xu
H. C. Wang et al., "Electrode Classification and Retrieval Using Supported Vector Machine", Advanced Materials Research, Vols. 160-162, pp. 743-749, 2011