APPLYING THE METHOD OF DIFFERENT CUMULATIVE BIN LOCAL BINARY PATTERN (DCBLBP) TO A SMALL IRIS REGION FOR FEATURES IN IRIS CLASSIFICATION

Yıl 2018, Cilt: 31 Sayı: 1, 126 - 137, 01.03.2018

Ahmad Nazri Ali

Öz

This paper presents an extended
variant of the local binary pattern (LBP) method to extract the iris feature
for iris classification system. The method is called Different Cumulative Bin
Local Binary Pattern (DCBLBP) in which the local iris information will be
projected into the binary bit and applies global characteristic for features
according to the ratio of bit transition along the horizontal axis. The
proposed DCBLBP scheme employs the majority bit decision for assigning the bit
to the reference elements where the assigned bit is unaffected by the
predetermined index number of the pixel block. The results demonstrate a good
classification score and show that the features extracted from the proposed
DCB_LBP scheme are reliable for iris classification system.

Anahtar Kelimeler

Local Binary Pattern, Support Vector Machines, Iris Classification

Kaynakça

• Reference1 R.P. Ramkumar Mahendra Institute of Technology India [e-mail]
• Reference2 Loris Nanni Department of Electronics Information and Systems, University Bologna, Italy [e-mail]
• Reference3 Izem Hamouchene LRIA Laboratory Computer Science Department USTHB University, Algeria [e-mail]