Publication:
Dorsal side of fingers acquisition with image enhancement on FPGA platform

Loading...
Thumbnail Image
Date
2023-08
Authors
Tan, Shu Han
Journal Title
Journal ISSN
Volume Title
Publisher
Research Projects
Organizational Units
Journal Issue
Abstract
Due to Covid-19, the demand for contactless biometric systems was sparked by the critical need to reduce the spread of the coronavirus as it can be transmitted through touching objects. Finger-knuckle-print has received a great attention by the biometric research community. This project aims to develop a system of image acquisition and image pre-processing for further finger-knuckle-print authentication process. This project is divided into three phases which are image acquisition, image enhancement, and evaluation. For image acquisition, a box with a USB webcam is set up as the acquisition device. Python and OpenCV are used to control the camera. For image enhancement, there are four algorithms which are Histogram Equalization (HE), Adaptive Histogram Equalization (AHE), Sharp Adaptive Histogram Equalization (SAHE), and Contrast-Limited Adaptive Histogram Equalization (CLAHE) are considered. All the image enhancement algorithms are performed in MATLAB, whereas in FPGA DE1-SoC, only HE is applied using Verilog language. The comparison of the results is carried out in two categories, which are comparison between the performance of four algorithms in MATLAB, and comparison between the performance of algorithm in between MATLAB and FPGA for the case of HE algorithm. In MATLAB, the algorithm with the best performance is CLAHE, which achieved a PSNR value of 16.64dB. From the experimental works, it is found that with proper coding process, the implementation of image enhancement algorithm can be applied in FPGA and the outcome of the image is the same with the implementation in MATLAB, which achieve a PSNR value of 13.43dB.
Description
Keywords
Citation