Abstract
Iris pattern is one of the most consistent biometric methods used for recognizing and identifying persons. Employing videos as a capturing instrument is a pretty modern style in the area of iris biometric. The use of frame by frame method provides more information and offers more suppleness compared to old-fashioned still images. Nevertheless, the size, quality and shape of the iris might differ between a frame and another. Additionally, to getting best performance it need a rapid and precise method to segment iris to amelioration rate of recognition. This work presents a method for choosing the best frames found in an iris video. This method is based on detecting motion blur and occlusion in iris videos and investigating their influence on the process of recognition. This proposed is followed by a rapid and precise method to detect pupil area, this method on the grounds of “dynamic threshold” with “Circular Hough Transform” then apply “Geodesic Active Contour” for detect outer boundary of iris. Experimental results are carried out on the MBGC NIR Iris Video datasets from the National Institute for Standards and Technology (NIST). Results show that the suggested selection method in NIR Iris Videos results in substantial enhancement in recognition efficiency. Results also indicated that the experimental evaluation of Iris segmented technique proposed in this work indicates that the precision and speed of the iris recognition via video is improved.
Original language | English |
---|---|
Pages (from-to) | 294-303 |
Number of pages | 10 |
Journal | Journal of Theoretical and Applied Information Technology |
Volume | 94 |
Issue number | 2 |
Publication status | Published - 31 Dec 2016 |
Fingerprint
Keywords
- GAC iris segmentation
- Iris biometrics
- Non-ideal iris recognition
- Pupil segmentation
- Video iris recognition
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)
Cite this
Enhance non-ideal iris recognition system from NIR Iris video. / Nseaf, Nur Khder; Shapi`i, Azrulhizam; Nseaf, Asama Kuder; Jaafar, Azizah; Jassim, Khider Nassif; Nsaif, Ahmed Khudhur.
In: Journal of Theoretical and Applied Information Technology, Vol. 94, No. 2, 31.12.2016, p. 294-303.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Enhance non-ideal iris recognition system from NIR Iris video
AU - Nseaf, Nur Khder
AU - Shapi`i, Azrulhizam
AU - Nseaf, Asama Kuder
AU - Jaafar, Azizah
AU - Jassim, Khider Nassif
AU - Nsaif, Ahmed Khudhur
PY - 2016/12/31
Y1 - 2016/12/31
N2 - Iris pattern is one of the most consistent biometric methods used for recognizing and identifying persons. Employing videos as a capturing instrument is a pretty modern style in the area of iris biometric. The use of frame by frame method provides more information and offers more suppleness compared to old-fashioned still images. Nevertheless, the size, quality and shape of the iris might differ between a frame and another. Additionally, to getting best performance it need a rapid and precise method to segment iris to amelioration rate of recognition. This work presents a method for choosing the best frames found in an iris video. This method is based on detecting motion blur and occlusion in iris videos and investigating their influence on the process of recognition. This proposed is followed by a rapid and precise method to detect pupil area, this method on the grounds of “dynamic threshold” with “Circular Hough Transform” then apply “Geodesic Active Contour” for detect outer boundary of iris. Experimental results are carried out on the MBGC NIR Iris Video datasets from the National Institute for Standards and Technology (NIST). Results show that the suggested selection method in NIR Iris Videos results in substantial enhancement in recognition efficiency. Results also indicated that the experimental evaluation of Iris segmented technique proposed in this work indicates that the precision and speed of the iris recognition via video is improved.
AB - Iris pattern is one of the most consistent biometric methods used for recognizing and identifying persons. Employing videos as a capturing instrument is a pretty modern style in the area of iris biometric. The use of frame by frame method provides more information and offers more suppleness compared to old-fashioned still images. Nevertheless, the size, quality and shape of the iris might differ between a frame and another. Additionally, to getting best performance it need a rapid and precise method to segment iris to amelioration rate of recognition. This work presents a method for choosing the best frames found in an iris video. This method is based on detecting motion blur and occlusion in iris videos and investigating their influence on the process of recognition. This proposed is followed by a rapid and precise method to detect pupil area, this method on the grounds of “dynamic threshold” with “Circular Hough Transform” then apply “Geodesic Active Contour” for detect outer boundary of iris. Experimental results are carried out on the MBGC NIR Iris Video datasets from the National Institute for Standards and Technology (NIST). Results show that the suggested selection method in NIR Iris Videos results in substantial enhancement in recognition efficiency. Results also indicated that the experimental evaluation of Iris segmented technique proposed in this work indicates that the precision and speed of the iris recognition via video is improved.
KW - GAC iris segmentation
KW - Iris biometrics
KW - Non-ideal iris recognition
KW - Pupil segmentation
KW - Video iris recognition
UR - http://www.scopus.com/inward/record.url?scp=85008151742&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85008151742&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:85008151742
VL - 94
SP - 294
EP - 303
JO - Journal of Theoretical and Applied Information Technology
JF - Journal of Theoretical and Applied Information Technology
SN - 1992-8645
IS - 2
ER -