This document summarizes a research paper on a finger vein extraction and authentication system for ATMs. The system uses repeated line tracking during feature extraction to improve the analysis of 256 pixels in finger vein images. During preprocessing, images undergo binarization, edge detection to isolate the finger region of interest, and enhancement. Features are then extracted using the repeated line tracking before classification with support vector machines. The system was tested on images from 30 subjects and achieved a peak signal to noise ratio of 78.1443 for identification, demonstrating its potential for biometric authentication applications like ATMs.