A comparison of rule-based, k-nearest neighbor, and neural net classifiers for automated industrial inspection

Abstract

As classifiers for use in automated industrial inspection, the rule-based, k-nearest-neighbor, and neural-network approaches are discussed. These approaches were implemented and tested for label verification in a machine vision system for hardwood lumber inspection. The test results, together with other considerations, have led to the selection of neural networks as the preferred method for doing the label verification in this machine vision system.<>