Automated conformance testing for ISO/IEC 19794-5 Standard on facial photo specifications

摘要

This paper proposes effective metrics for quantitative conformance testing for ISO/IEC 19794-5 standard on facial photo specifications. Each metric is normalised to [0, 10], as a quality score of each requirement, to conveniently utilise in facial photo quality validation systems. Furthermore, this paper proposes a robust method of extracting necessary features in the images for automated conformance testing. The proposed extraction method takes advantages of colour, intensity and edge information. Experimental results over a subset of FERET, GTAV, and FIePI databases demonstrated the effectiveness and robustness of the proposed metrics and extraction method.