if you want the project pls call @8125424511
IRIS DATASET
ABSTRACT
This work shows the applicability and feasibility of
different machine learning techniques on iris recognition from smartphone
captured eye images. First, the iris is localized using the popular Daugman’s
method and the eyelids are suppressed with canny edge detection technique. Then
normalization of the extracted iris region is performed in a novel way by
setting an adaptive threshold. Next, the normalized image is decomposed using
Haar wavelets to obtain the feature vectors. Histogram equalization is
performed for better classification accuracy. After that, different classifiers
are trained using the extracted feature vectors which yield about 99.7%
accuracy for training and 97% accuracy for testing. Finally, the results are
compared with other previously applied methods on the same dataset and it is
found that the proposed method outperforms most of them.
INTRODUCTION
Human iris is well known for its
uniqueness, stability and non-evasiveness [1]. Hence iris recognition is a very
popular problem for the researchers in the field of bioinformatics,
cryptography, computational intelligence etc. Many successful approaches have
been taken so far. This approaches can be classified into two categories based
whether they used machine learning or not. A significant thing to notice that
the datasets which are used in these approaches are iris images captured by NIR
(Near Infrared) camera as they offer very good visibility of iris texture, even
for heavily pigmented regions [2]. As a result, the extracted iris region
provides more accurate information implying better chances for recognition.
However, the setup complexity with the above mentioned camera is difficult
especially when the issue of portability and simplicity arises. On the other
hand, smartphones with cameras are within everyone’s reach now a days. The only
problem with these cameras is that they capture images in visible light
spectrum resulting in less detailed iris images compared to the NIR cameras. So
the usual question arises “Are they good enough for iris recognition?” A good
number of studies [2,8] replied positively towards the question. The one thing
to notice here that almost all the above mentioned approaches paid a little or
no effort to state the feasibility or usability of machine learning techniques
in case of the smartphone captured iris database. This is important because
machine learning techniques have provided very good results in case of NIR
camera captured datasets [3]. As iris images in visible light are likely to
offer relatively less details, the question of applicability of ML techniques
in this case still remain unanswered. However, one studyRaja et. al[4] used
Sparse Reconstruction Classifier with K-means clustering which gave a very low
EER percentage (Equal Error Rate).It is in other words a very good indication
but it does not compare or give any further info on the feasibility of other
machine learning techniques. In our work, we investigate further on the use of
machine learning techniques on iris recognition using smartphone captured iris
images in visible light spectrum. In order to do so we develop a complete
segmentation and feature extraction technique and try to use same set of
extracted features to train different classifier. Finally, we compare the
classification accuracy of the trained classifiers and decide whether the
machine learning techniques are feasible in case of smartphone captured
database.
MODULES
A.
Image Pre-processing: The images provided in the dataset were in RGB
format. We had to convert it into a single channel to proceed. While
converting, red channel was used as wavelengths corresponding to red light
(closest to near infrared) are the longest in our visible spectra, the best
iris pattern should be visible this way.
B.
Iris Localization:
For extracting the iris region first, the classic Daugman’s
Integro-differential operator is used.
C.
Eyelid Suppression: The visible portion of the iris part is not exactly
circular. It is partly covered by the eyelids which needs to be suppressed. To
do so, we followed an approach inspired from Masek [18]. The total search
region was divided into two parts, upper eyelid and lower eyelid. The width of
the search region is exclusively the difference between the iris and the pupil
radius.
D.
Normalization:
So far we have successfully segmented the iris part and suppressed eyelids. Now
we have to transform it into fixed dimensions for further processing. To do
that, we used the very popular homogenous rubber sheet model introduced by
Daugman [17]. In the homogenous rubber sheet model, each point within the iris
region is remapped to a pair of polar coordinates (𝑟, 𝜃) where r is on
the interval [0, 1] and 𝜃 is angle in the
range [0, 2𝜋].
E.
Eyelash Removal:
Even though eyelash removal is a part of noise cancellation, it was done after
the normalization in our work. Developing a method to do so was a tough job as
the eyelashes differ largely from image to image. The most obvious option was
applying a threshold. But if a hard threshold value is applied, there is no
guarantee that it will work for every image. In some images the iris region was
darker than the others. So we had to develop an adaptive algorithm to set the
threshold value in each image separately.
F.
Histogram Equalization: Once the iris region is segmented, normalized and
noise has been removed, the relevant texture and intensity information needs to
be extracted to train a classifier. But before doing that a histogram
equalization was performed on the normalized images. This is because the
histogram analysis of the normalized image revealed that the image intensities
were congested in a very small region making it harder for the classifier to
differentiate.
G.
Feature Extraction: A typical iris consists of lots of complex patterns
such as arching ligaments, furrows, ridges, crypts, rings, corona, freckles and
a zigzag collarette. These complex patterns are very much complicated to
extract. That is why we chose to train with the image itself. So far we have a
normalized image of size 64 x 512 pixels. If we want to train the classifier
with this amount of data, it will be too heavy and training will take a very
long period of time. So we need to scale it. But scaling may result in loss of
important information.
H.
Training
Classifier: In the given database, we had eye images of 70 people. For
training the model we took 5 images for each person and rest of the images were
kept aside for testing the classifier. For training we used 5-fold cross
validation method so that each image in the training set can be tested once
against the others. We tried several classifier and among them support vector
machines, knearest neighbour, linear discriminants etc. showed great promise.
EXISTING
SYSTEM
Several works
have been done with the publicly available datasets UBIRISv1 [5], UBIRISv2 [6],
MICHE [7] etc. containing iris images in visible light spectrum. The challenges
of iris recognition associated with unconstrained iris images in visible light
were discussed by Proenca et al [8]. Noisy iris images and independent
segmentation and noise recognition procedures are most likely the sources of
errors. Santos et al. [9] explored best illumination configurations for visible
light iris images. gorithms offered more than 95% accuracy for the dataset.
Machine learning techniques have also been proved to be very successful in iris
recognition. A study from De Marsicoet al. [3] compared different machine
learning techniques in iris recognition. In these studies, they used mostly
CASIA-Iris [15] dataset which is created from images taken with NIR camera.
Among different approaches, Rai and Yadav [16] were able to obtain 99% accuracy
with a combination of Support Vector Machines and Hamming distance.
PROPOSED
SYSTEM
In the proposed system we investigate
further on the use of machine learning techniques on iris recognition using
smartphone captured iris images in visible light spectrum. In order to do so we
develop a complete segmentation and feature extraction technique and try to use
same set of extracted features to train different classifier. Finally, we
compare the classification accuracy of the trained classifiers and decide
whether the machine learning techniques are feasible in case of smartphone
captured database.
CONCLUSION
In this paper a machine learning based
approach on iris recognition from smartphone captured images is proposed. With
the results above, this paper successfully showed that in case of smartphone
captured visible spectrum iris images, the machine learning techniques are
equally as good as the other ones, in some cases even better. Still accuracy
can be further improved. And in our findings, accuracy largely depends on
accurate segmentation. So some robust approaches may be taken to improve the
segmentation result. In our approach we tried to stick to some basic
segmentation approaches. This was done keeping in mind their easy
implementation. As smartphones of todays’ are equipped with very good camera,
the whole recognition system shows great promise to be implemented on these
smartphones for recognition, security and identification purpose. Already
Samsung® [19] has developed a built in iris scanner which works for the user
who is using it. Our next task would be to develop a cloud based server where
iris data can be easily sent through the smartphone. The classifier will run on
the server and the sent data would be matched and verified. Thus by just using
the smartphones, it will be possible to develop a full security system.
BIBLIOGRAPHY
[1] J. Daugman,
"How iris recognition works.," in IEEE Transactions on circuits and
systems for video technology, 2004.
[2] M.
Trokielewicz, "Iris Recognition with a Database of Iris Recognition with a
Database of Iris Images Obtained in Visible Light Using Smartphone
Camera," in The IEEE International Conference on Identity, Security and
Behavior Analysis (ISBA 2016), Sendai, Japan, 2016/02.
[3] M. D.
Marsico, A. Petrosino and S. Ricciardi, "Iris recognition through machine
learning techniques: A survey," Pattern Recognition Letters, vol. 82, pp.
106-115, 2016.
thank you for your comment
pls call me on 8125424511