A Novel Deep Learning Lacunarity Texture Analysis System Using Mid-Point ROI Extraction Algorithm for Palmprint Recognition System

The palmprint comprises multiple unique patterns that are distinct in detecting human identity. There are numerous algorithms proposed by past researches for recognizing Two-dimensional Palmprint Region of Interest (2DPROI) images. In this research, an innovative Deep Learning Lacunarity Texture Analysis System (D2LTA) is developed for recognizing the accredited persons at higher recognition rate. To impart the D2LTA model, Two-dimensional palmprint hands’ ROI images are segmented using Mid-point ROI generation algorithm, produced a peculiar feature vector using lacunarity approach in a state-of-the-art manner, and then Deep Learning ConvNet classifier is proposed for D2LTA system to justify the accredit person. The key principle of the Mid-point ROI generation approach is to determine the perfect straight line on the center of the palm. Based on the straight line in the palm, determine the pixel values of the ROI’s rectangular box. To catch the perfect straight line, line mid-point method is used. To do this research, 2D-palm hands are procured from three different datasets such as BMPD, CASIA and IIT palm datasets and 2DPROI images are secured from PolyU, Hong Kong Polytechnic University, Hong Kong. The proposed model has been assessed with diverse dimensions to prove the acquirement 99.25% of higher precious authentication rate.