Fingerprint quality and validity analysis

This paper discusses methods to estimate the quality as well as validity of a fingerprint image. Orientation certainty is used to certify the localized texture pattern of the fingerprint images while ridge to valley structure is analyzed to detect invalid images. Global uniformity and continuity ensures that the image is valid as a whole. 150 images with various qualities are evaluated using the proposed algorithm and quality benchmark we defined. A monotonic relationship is found indicating that the proposed algorithm is feasible in detecting low quality as well as invalid fingerprint images.