Abstract: Biometrics finds a lot of application in the modern day security aspects whether it may be in industrial sector, defense sector or in domestic sectors. In this world, without biometrics, it is very difficult to incorporate the security measures in any organization. Hence, personal identification can be very important problem in any bio-metric recognition scenarios. In this proposed technique, this biometric recognition using retinal images is done. The retinal images are preprocessed and features are extracted using image processing and KNN classifier is trained with 66 recognized images and 35 unrecognized images The proposed algorithm is tested with 20 images and the output represents if the input image is a recognized one or not.ROC curve is plotted. This accuracy of the algorithm can be extended with more number of training images. This method of biometric recognition using retinal images finds more use in this Covid pandemic situation compared to the fingerprint biometric recognition system.
Keywords:Bio-metric recognition, Retinal images, KNN classifier
|
DOI:
10.17148/IJARCCE.2021.10419
