A collection of modelling approaches for visual inspection in industry

Abstract

The paper outlines the consequences of inspection error for visual statistical quality control schemes and goes on to discuss various approaches that have been adopted to try to estimate the proportion of good items rejected and the proportion of defective items accepted. Brief discussions of those approaches are included, with some criticisms. It is said that most of the approaches used so far have not contributed much towards a model of inspector performance, though there are some recent developments that look a little more encouraging.