Performance Comparison of Edge Detection Operators for Supporting Lip Prints Biometric Identification

Abstract

In this paper, a preprocessing method for human identification implemented on lips print Biometric system has been explored. The methods used are based on Laplacian and Sobel operators for detecting edges in lips images. Experiment has been applied successfully using custom-based lip images. The steps to perform experiment can be described as follows: (i) Acquisition of lips image, (ii) Basic pre-processing tasks and (iii) Segmentation using detection operators. Not less than 90 images have been used for recognizing lips print patterns. The experimental result showed that Sobel edge detection operator is better than Laplacian operator due to multiply kernel usage in Sobel operator.