Volume 18, No. 2, 2021
Hough Transform for Distinctive Edge Detection to Images in Fingerprint Recognition Matching Transformation
Hayder G.A. Altameemi, Ahmed Abdul Azeez Ismael and Raddam Sami Mehsen
Abstract
Biometric Identification is a globally renowned procedure, which has been utilised to achieve a successful and accurate level of identification. In the sea of biometrics, fingerprints are deemed more popular when it comes to verification. This results from the presence of the ridges on the fingerprints that are completely exclusive to each individual. Besides that, fingerprints are expansively employed to ascertain and authenticate people individually. Therefore, this study had proposed to employ distinctive Edge Detection techniques together with the Hough Transform to match the images of the fingerprints in a fingerprint matching system. The Hough Transform is a superior procedure carried out to get an accurate series of finer points or lines. The finer points or lines would then distinguish the fingerprints. Nevertheless, it was still a challenge to extract finer points or lines from the fingerprints under uninhibited conditions. Therefore, this paper was organised based on four distinctive steps. First, different Edge Detection operators were employed to perform the fingerprint matching algorithm. Next, the fingerprint matching algorithm was applied twice to the same Edge Detection operators. Thirdly, the Edge Detection operators had been substituted with the Transformation Method for the same matching procedure. For example, the proposed fingerprint matching algorithm comprised of the Hough Transform and same Edge Detection operators. Finally, distinct Edge Detection operators based on the decision making algorithm were used to calculate and determine the percentage of matching. Therefore, this study proved that the prints obtained via the Prewitt Edge Detection together with Hough Transform were in an agreement.
Pages: 999-1010
DOI: 10.14704/WEB/V18I2/WEB18369
Keywords: Sobel, Canny, Prewitt Edge Detection, Hough Transforms, Fingerprint, Matching Transformation.