Figure 5 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Partitioning process. (a) Partitioning for selecting the different ranges of frequency information. (b) Partitioning for vessels information in different direction. (c) Final partition includes the combination of the partitions in Figures 5(a) and (b).
Source publication
Retinal image is one of the robust and accurate biometrics methods to recognize a person. In this article we present a new biometric identification system based on Fourier transform and angular partitioning of the spectrum. In this method, at first, the optical disc is localized using template matching technique and used for rotating the retinal im...
Contexts in source publication
Context 1
... solve this problem, we introduce a new partitioning based on dividing the Fourier spectrum and phase angle into several half circles with the same centre which is the centre of spectrum that include segments with same area and same degree arc as shown in Figure 5. Purposed partition in this method (Figure 5(c)) includes a combi- nation of two partitions shown in Figure 5(a) and Figure 5(b). As shown in Figure 5(a), the spectrum and phase angle are divided into different ranges of high and low frequency information. Whereas the thick vessels are of low frequency patterns and the thin vessels are of high frequency patterns, thus this partitioning shows the crite- rion of the thickness of the vessels, and while Figure 5(b) that includes the same frequency range, has information about vessels in different directions, Figure 5(c) has in- formation about the thickness and the direction of the ...
Context 2
... solve this problem, we introduce a new partitioning based on dividing the Fourier spectrum and phase angle into several half circles with the same centre which is the centre of spectrum that include segments with same area and same degree arc as shown in Figure 5. Purposed partition in this method (Figure 5(c)) includes a combi- nation of two partitions shown in Figure 5(a) and Figure 5(b). As shown in Figure 5(a), the spectrum and phase angle are divided into different ranges of high and low frequency information. Whereas the thick vessels are of low frequency patterns and the thin vessels are of high frequency patterns, thus this partitioning shows the crite- rion of the thickness of the vessels, and while Figure 5(b) that includes the same frequency range, has information about vessels in different directions, Figure 5(c) has in- formation about the thickness and the direction of the ...
Context 3
... solve this problem, we introduce a new partitioning based on dividing the Fourier spectrum and phase angle into several half circles with the same centre which is the centre of spectrum that include segments with same area and same degree arc as shown in Figure 5. Purposed partition in this method (Figure 5(c)) includes a combi- nation of two partitions shown in Figure 5(a) and Figure 5(b). As shown in Figure 5(a), the spectrum and phase angle are divided into different ranges of high and low frequency information. Whereas the thick vessels are of low frequency patterns and the thin vessels are of high frequency patterns, thus this partitioning shows the crite- rion of the thickness of the vessels, and while Figure 5(b) that includes the same frequency range, has information about vessels in different directions, Figure 5(c) has in- formation about the thickness and the direction of the ...
Context 4
... solve this problem, we introduce a new partitioning based on dividing the Fourier spectrum and phase angle into several half circles with the same centre which is the centre of spectrum that include segments with same area and same degree arc as shown in Figure 5. Purposed partition in this method (Figure 5(c)) includes a combi- nation of two partitions shown in Figure 5(a) and Figure 5(b). As shown in Figure 5(a), the spectrum and phase angle are divided into different ranges of high and low frequency information. Whereas the thick vessels are of low frequency patterns and the thin vessels are of high frequency patterns, thus this partitioning shows the crite- rion of the thickness of the vessels, and while Figure 5(b) that includes the same frequency range, has information about vessels in different directions, Figure 5(c) has in- formation about the thickness and the direction of the ...
Context 5
... solve this problem, we introduce a new partitioning based on dividing the Fourier spectrum and phase angle into several half circles with the same centre which is the centre of spectrum that include segments with same area and same degree arc as shown in Figure 5. Purposed partition in this method (Figure 5(c)) includes a combi- nation of two partitions shown in Figure 5(a) and Figure 5(b). As shown in Figure 5(a), the spectrum and phase angle are divided into different ranges of high and low frequency information. Whereas the thick vessels are of low frequency patterns and the thin vessels are of high frequency patterns, thus this partitioning shows the crite- rion of the thickness of the vessels, and while Figure 5(b) that includes the same frequency range, has information about vessels in different directions, Figure 5(c) has in- formation about the thickness and the direction of the ...
Context 6
... solve this problem, we introduce a new partitioning based on dividing the Fourier spectrum and phase angle into several half circles with the same centre which is the centre of spectrum that include segments with same area and same degree arc as shown in Figure 5. Purposed partition in this method (Figure 5(c)) includes a combi- nation of two partitions shown in Figure 5(a) and Figure 5(b). As shown in Figure 5(a), the spectrum and phase angle are divided into different ranges of high and low frequency information. Whereas the thick vessels are of low frequency patterns and the thin vessels are of high frequency patterns, thus this partitioning shows the crite- rion of the thickness of the vessels, and while Figure 5(b) that includes the same frequency range, has information about vessels in different directions, Figure 5(c) has in- formation about the thickness and the direction of the ...
Context 7
... solve this problem, we introduce a new partitioning based on dividing the Fourier spectrum and phase angle into several half circles with the same centre which is the centre of spectrum that include segments with same area and same degree arc as shown in Figure 5. Purposed partition in this method (Figure 5(c)) includes a combi- nation of two partitions shown in Figure 5(a) and Figure 5(b). As shown in Figure 5(a), the spectrum and phase angle are divided into different ranges of high and low frequency information. Whereas the thick vessels are of low frequency patterns and the thin vessels are of high frequency patterns, thus this partitioning shows the crite- rion of the thickness of the vessels, and while Figure 5(b) that includes the same frequency range, has information about vessels in different directions, Figure 5(c) has in- formation about the thickness and the direction of the ...
Citations
... Sabaghi et al. [17] have compensated the rotation of retinal image for a robust system. They have used FT (Fourier Transform) of the retinal image to extract features and called this feature vector as FSPF, and then they have used Euclidean distance to perform matching process. ...
A biometric authentication system provides an automatic person authentication based on some characteristic features possessed by the individual. Among all other biometrics, human retina is a secure and reliable source of person recognition as it is unique, universal, lies at the back of the eyeball and hence it is unforgeable. The process of authentication mainly includes pre-processing, feature extraction and then features matching and classification. Also authentication systems are mainly appointed in verification and identification mode according to the specific application. In this paper, preprocessing and image enhancement stages involve several steps to highlight interesting features in retinal images. The feature extraction stage is accomplished using a bank of Gabor filter with number of orientations and scales. Generalized Discriminant Analysis (GDA) technique has been used to reduce the size of feature vectors and enhance the performance of proposed algorithm. Finally, classification is accomplished using k-nearest neighbor (KNN) classifier to determine the identity of the genuine user or reject the forged one as the proposed method operates in identification mode. The main contribution in this paper is using Generalized Discriminant Analysis (GDA) technique to address ‘curse of dimensionality’ problem. GDA is a novel method used in the area of retina recognition.
... There are different authentication schemes described in [8,12,14,15,16,17]. These authentication schemes are very different to each other. ...
... The vascular graph is constructed using different approaches which helps to authenticate. In [14] a method has been used based on fourier transforms to evaluate retinal biometrics. The cosine values are taken into consideration. ...
The retinal pattern biometrics varies to every individual and it is one of the most efficient form of identification. Retinal scanners use infrared rays along with non-invasive color retinal cameras and are very effective in getting a retinal image. We differentiate retinas using segmentation of retinal vascular branch.Further Bifurcation and cross over points are used Steganography is the art of hiding information within another file which can be image, video, audio, text etc. After getting the retinal details, steganographic methods are used to transfer this file securely over the internet. Once received at the receiver, it can be used for verification of user authenticity
... Sabaghi et al. [24] proposed a new biometric identification system based on combination of Fourier transform and special partitioning and wavelet transform. At first, optical disc was localized using template matching technique in order to rotate the retinal image to reference position, then used wavelet transform and angular partitioning of the frequency spectrum information of retinal image for feature extraction. ...
... So, in future an enhancement for the pre-processing stage is required to raise its capability to collect the thin and small vascular segments. [24] -99.1% Qamber et al. [25] 96.29% 100% Rubaiyat et al. [26] 99.71% 100% Monisha et al. [28] 96% 97.5% Rani et al. [29] -94% Proposed method 100% 100% ...
Retinal biometric is a new methodology that increasingly being used, especially for authentication cases required high level of persons identification. Retinal recognition deals with very distinct physical property has, exceptional, very low false acceptance and false rejection rates, and the features that are determined in the retina of eye are more reliable and stable features than those found in other biometrics. This paper presents a new system for personal recognition based on retinal vascular pattern. This system is capable to compensate the effects of eyes rotation and robust to noise and brightness variations. The developed system consists of three main stages (i.e., preprocessing, feature extraction, and matching stage). Preprocessing was used (1) to enhance the retina image, and (2) to extract the vascular network (i.e., Region of Interest); then a set of discriminating local geometric features are extracted, it is a set of local average of vascular densities are proposed to define the vesicular network. Finally, Euclidean distance measure was used in the matching stage. The proposed system was evaluated on the two publicly available databases: (i) STARE (Structured Analysis of the Retina) and (ii) DRIVE (Digital Retinal Images for Vessel Extraction). The test results indicated that the attained recognition accuracy of the proposed method is 100% for both datasets.
... Köse et al. [23] proposed a retinal identification that employed a similarity measure and is capable of tolerating the transformations. In [24], Fourier transform coefficient and angular partitioning are used for feature detection. Euclidean distance is used in the matching process. ...
The expansion of automation techniques and increased risk of identity theft have led emphasis on the tremendous need of automated identification system. Due to the high recognition accuracy and robustness to changes in human physiology, retinal biometric identification system has drawn much attention in this research field. In this paper, we aim to propose an automatic fast and accurate retinal identification system for the multi-sample data set. The proposed approach uses a hybrid segmentation technique to segment out both thick/thin vessels for effectively balancing the difference of wavelet response between thick/thin blood vessels. As a result, recognition accuracy is improved. A PCA (Principle Component Analysis) based feature processing approach is proposed for efficiently reducing the dimensionality of a large number of vessels features. It significantly reduces computation time and accelerates the matching process in the retinal identification system. The proposed technique is validated on DRIVE, STARE, VARIA, RIDB, HRF, Messidor, DIARETDB0, and a large multi-sample per subject database created by authors using the images provided by Dr. Chen (Shanghai Jiao Tong University Affiliated Sixth People Hospital). Experimental results demonstrated that the proposed approach outperforms other existing techniques. Segmentation achieves an overall accuracy of 99.65% with the recognition rate of 99.40% on all these databases.
... Sabaghi et al. [27] have introduced a human identification system based on Fourier transform (FT) and angular partitioning of the spectrum of retinal images. For a robust system, the rotation of retinal image should be compensated. ...
... They have also shown that their system is robust to noise. [27] Later in [28], they have introduced another method, which is a combination of wavelet transform along with Fourier transform for FSPF extraction. With Fourier transform they extracted Fourier energy feature and with wavelet transform, they have extracted wavelet energy feature. ...
... The result of regulate retinal image to reference position: (a) Retinal image after localized center of mass and optical disk; (b) Obtain the angle of rotate; (c) Compensated image by rotation[27]. ...
... The method was examined with 480 images from 80 subjects, 40 from DRIVE and 40 from STARE databases and achieved an authentication rate of 100%. Sabaghi et al. (2011) introduced a biometric identification system based on Fourier transform (FT) and angular partitioning of the spectrum. For extracting features, first they have performed FT of retinal image and divided the Fourier spectrum using angular partitioning. ...
Abstract: In biometric authentication system, distinct set of characteristic features are
used to identify an authorized person. Retina is a stable biometric feature because of
its location and unique physiological characteristics. In this paper, we propose a texture
feature based retinal authentication system. Texture features are considered as important
features for authentication purpose. These texture features of retina are extracted using
Local Con�guration Pattern (LCP) and Radon transform technique. The LCP computes
the local structural information as well as the microscopic information of the image.
Using Radon transform on retinal images, Radon features are extracted which contains
the texture information of the blood vessels. A feature vector is formed by combining
all theses LCP and Radon features and then fed to a Feed-forward Arti�cial Neural
Network (FANN) classi�er. This stage checks whether the test image belongs to the
authorized person or not. Three general retinal databases DRIVE, HRF, Messidor and
images collected from two local eye hospitals are considered to authenticate a person.
Two retinal authentication databases RIDB and VARIA are also used for evaluating the
performance of the system. The results obtained show that the system is e�ective and
e�cient in authenticating the individuals.
... The proposed fuzzy system makes the final decision. The trained neural network can be used for decision making instead of fuzzy logic [6]. R. Ghaderi et al. projected methodology for separation of the vasculature in retinal pictures with the 2-D Morlet riffle and Neural Network. ...
... It takes smooth image patch p. A binary check is given by; (p; x,y) = (6) where, p(x) is intensity of p at point x= . The feature is defined with n binary tests; fn(p)= (p;xi,yi) ...
... Our literature survey has found only a few techniques which can reliably identify an individual using retinal vascular feature. Some prominent retinal imaging based biometric authentication schemes are described in [8,16,[18][19][20][21]. Among these methods, EyeDentify [16] is a commercially available system. ...
... A support vector machine (SVM) classifier is used to distinguish between genuine and imposter comparisons. Sabaghi et al. [18] proposed a method which is based on the Fourier transform of the retinal image and angular partitioning of the spectrum. Perceptual uniform colour space, Gabor filters and template matching based retinal image matching framework was proposed in [19]. ...
Retinal vascular network pattern is unique to each individual which can be used for person identification in biometric authentication. In this study, the authors have proposed a novel biometric authentication method using retinal vascular branch, bifurcation and crossover points (i.e. feature points). The method automatically extracts the vascular network from colour retinal images and identifies these feature points. The major blood vessels characterised by width and length are identified from the segmented vascular network. For this, a novel vessel width measurement method is applied and vessels more than certain widths are selected as major vessels following an established protocol. The geometric hashing technique is developed to compute the invariant features from these feature points. They consider the feature points from major vessels which will be less susceptible to noise for modelling a basis pair and all other points together for locations in the hash table entries. The models are invariant to rotation, translation and scaling as inherited from geometric hashing. For each person, the system is trained with the models to accept or reject a claimed identity. They have tested their method on 3010 retinal images and achieved 96.64% precision and 100% recall.
... Sabaghi et al. [15] proposed a new biometric identification system based on combination of Fourier transform and special partitioning and wavelet transform. At first, optical disc was localized using template matching technique in order to rotate the retinal image to reference position, then used wavelet transform and angular partitioning of the frequency spectrum information of retinal image for feature extraction. ...
... So, in future an enhancement for the pre-processing stage is required to raise its capability to collect the thin and small vascular segments. [12] 99.00% -Barkhoda et al. [13] 98% -Rubaiyat et al. [15] 100% 99.71% Sabaghi et al. [16] 99.1% -Monisha et al. [17] 97.5% 96% Proposed method 100% 100% ...
Retinal biometric is one of the newest biometric technologies that increasingly being used,
especially in critical areas that require tight security measures, in order that retinal is one of the
most robust and most accurate biometrics methods to recognize a person, the retinal recognition
technique is yet another step in biometrics, it deals with very distinct physical property of
exceptionally very low false acceptance and false rejection rates, and features that are found in the
retina of eye are more reliable and stable features than those found in other biometrics. Retinal
biometric is one of the newest biometric technologies that increasingly being used, especially in
critical areas that require tight security measures, in order that retinal is one of the most robust and
most accurate biometrics methods to recognize a person, the retinal recognition technique is yet
another step in biometrics, it deals with very distinct physical property of exceptionally very low
false acceptance and false rejection rates, and features that are found in the retina of eye are more
reliable and stable features than those found in other biometrics.
This paper presents a new system for personal recognition based on retinal vascular pattern. This
system is insensitive to rotation and robust to noise and brightness variations. The presented
system consists of three main stages (i.e., preprocessing, feature extraction, and matching stage).
Preprocessing is used for enhancement and segmentation the vascular network (i.e., Region of
Interest), as discriminating feature the set of local average of vascular densities have been used in feature extraction stage, finally Euclidean distance measure used in matching stage. The proposed
system is evaluated on the two publicly available databases: (i) STARE (Structured Analysis of the
Retina) and (ii) DRIVE (Digital Retinal Images for Vessel Extraction). The test results indicated that
the attained recognition accuracy of the proposed method is 100% for both datasets.
... The first method on retina identification dates back to 1976 when EyeDentify company introduced the system named EyeDentification 7.5 [12]. Since then several new methods have been reported in the literature, see for instance [1,16,6,13,18,17,7] and the references therein. The method in [13] was proposed based on the Scale Invariant Feature Transform (SIFT) and the Improved Circular Gabor Transform (ICGT). ...
... The method was composed of three principal modules including blood vessel segmentation, feature generation, and feature matching. The authors in [18] extracted the features based on Fourier transform and angular partitioning of the spectrum, and employ the Euclidean distance for feature matching. ...
Retinal fundus images are widely used for screening, diagnosis and prognosis purposes in ophthalmology. Additionally these can also be used in retinal identification/recognition systems, for identification/authentication of an identity. In this paper the aim is to explain how norms in function spaces can be used to set up, automatically, classes of different retinal fundus images. These classifications rely on crucial and unique retinal features, such as the vascular network, whose location and measurement are appropriately quantified by weighted norms in function spaces. These quantifications can be understood as retinal pattern assessments and used for improving the efficiency and speed of retinal identification/recognition frameworks. The proposed methods are evaluated in a large dataset of retinal fundus images, and, besides being very fast, they achieve a reduction of the search in the dataset (for identification/recognition purposes), by 70% on average.
