An evaluation of the UHMFS system facial image pattern recognition algorithms

The accuracy of a pattern recognition algorithm influences the degree to which application of such an algorithm is accepted. This thesis evaluates the accuracy and the sensitivity of a facial image pattern recognition algorithm. A systematic application of linear regression and analysis of variance establish both the accuracy of the algorithms and the deficiencies in the algorithms. Areas of possible improvement are defined, and recommendations for further study are presented. The approach demonstrates the application of linear regression and analysis of variance in the evaluation of experimental data.