IRIS RECOGNITION SYSTEM
Iris recognition is a standout amongst the most exact and high certainty for user verification technique that used today. The general-purpose iris recognition system is of low speed and not portable. Hence there is a need to make use of dedicated hardware for this. Most of the algorithms of this system using MATLAB do not incorporate parallel processing instead comes under sequential processing. In the proposed framework, CASIA-IRIS dataset has been used for iris recognition, here the original eye image was pre-sampled to standard pixels i.e 256*256 to crop the unneeded parts of the eye image and to decrease the time of processing during the pupil and iris boundaries detection. These eye images are processed with four major steps, they are
1. Segmentation of IRIS,
2. Normalization Process,
3. Feature Extraction and
4. Matching of templates.
Image pre-processing, segmentation of iris and normalization carried out using MATLAB software. Feature extraction algorithm (Gabour Filter) and feature matching (Hamming Distance) is simulated using Modelsim designed using Verilog HDL to achieve parallel processing and to improve the time delay. This methodology resulted in a reduction in a number of device utilization on board using Xilinx ISE and simulation time delay. This design process consisting of 5 different steps:
(i) Smoothing: convolution of image pixel values and gaussian smoothing filter values yields a blurred image, this smoothing is mainly to remove the noise of the image.
(ii) Finding edge detection using canny edge operator.
(iii) Non-maximum suppression: the pixels which are of non-maximum values will get suppressed and replaces values with '0' except points at local maxima.
(iv) Double thresholding: in order to extract only the strong edges, present in image, thresholding concept is used, which makes easier to remove false edge points
(v) Tracking of edges by hysteresis: edge points will be tracked by suppressing the most of edges that are not have a connection with the strong edge line.
Iris Recognition Framework
Reference Paper-1: Implementation of Iris Recognition System using FPGA
Author’s Name: Uma B and Pawan Kumar B
Source: IJETCSE
Year:2018
Reference Paper-2: Iris recognition using 2-D Gabor filter and XOR-SUM code
Author’s Name: Ritesh Vyas, Tirupathiraju Kanumuri, and Gyanendra Sheoran
Source: IEEE
Year:2016
Request source code for academic purpose REQUEST FORM or contact +91 7904568456 by WhatsApp or info@verilogcourseteam.com, fee applicable.
SIMULATION VIDEO DEMO