Iris Feature Extraction using Daubechies Wavelet Transform and its Recognition using Normalized Euclidean Distance
R. Rizal Isnanto, Achmad Hidayatno, Antonius Dwi Hartanto · 2013
Iris has a very unique pattern and different for each person, therefore it can be used as the basis for biometrics recognition system . One of methods t o identify the texture of iris is wavelet transform based on energy feature. The research purpose is to analyze the extraction as well as the recognit-ion on iris feature based on wavelet transform. In this research, Daubechies wavelet transform order 4 is used for extraction and Normalized Euclidean Distance is applied in recognition. Iris image s used in this research was processed, starts with preprocessing steps as follows: capturing eye objects using digital camera, segmenting iris from other parts of eye (i.e. pupil and sclera), and convert ing iris images into polar form. Then , the feature extraction is done using Daubechies wavelet transform order 4 for which its features obtained by energy value (approximation, horizontal, vertical, and diagonal energy) . Finally, recognition using normalized Euclidean distance is conducted . Test s consist of four types: influence of the number of sample database, influence of the decomposition level of Daubechies wavelet transform order 4, influence of different input image formats, and testing on iris images which are not in database. From the test results, it can be concluded that the highest recognition rate is 86.66% , achieved in Daubechies wavelet transform order 4 level 4 with two samples iris image stored. Meanwhile, t he lowest recognition rate is 62,5%, obtained when wavelet transform order 4 level 6 with one sample iris image stored applied . From test of influence of different input image formats, it can be concluded that the samples taken from 40 individuals which one sample is put f rom each person, the result of BMP format is equal to JPEG. Whereas, for iris images which are not in database with threshold 0.3559, the recognition level achieved is 96%.