Iris recognition by using deep learning classifier and fractal feature extractor

Kua, Choon Wen (2021) Iris recognition by using deep learning classifier and fractal feature extractor. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Preview

Text (Full Text) Iris recognition by using deep learning classifier and fractal feature extractor.pdf - Submitted Version Download (1MB) | Preview

Abstract

Iris recognition system is a reliable biometric technology due to its high uniqueness and performance. This system provides high security service since the iris is unique and hard to be fake. This project is aimed to build an iris recognition system by using MATLAB and analyse the performance in terms of accuracy and recognition speed. The flow of iris recognition system starts with image acquisition and followed by edge detection. In edge detection, the centre of iris, radius of inner and outer iris boundary is extracted. Then, the information is used to segment the iris from iris image. The normalization is then used to normalize the segmented iris to rectangular block. The fractal feature which is lacunarity is extracted and inputs to the convolutional neural network. In the proposed iris recognition system, 50 iris images are trained, and another 50 iris images are tested. The proposed iris recognition system recognizes 49 iris images which is 98% of accuracy with an average recognition time of 0.2s approximately. Therefore, this proves that iris recognition system is reliable due to its high accuracy and fast recognition time.

Actions (login required)

View Item