Eagle-Eyes: A System for Iris Recognition at a Distance Faisal Bashir, Pablo Casaverde, David Usher, Marc Friedman Retica Systems, Inc. 201 Jones Road, Waltham, MA 02451 {fbashir, pcasaverde, dusher, mfriedman}@retica.com ABSTRACT In recent years the human iris has been established as a leading biometric. Academic research is expanding and several products have become commercially available. Conventional iris recognition systems require undue cooperation from users in terms of proximity to the device as well as movement constraints. Recent innovations in iris acquisition systems and recognition algorithms have aimed to relax these constraints. This paper unveils the design of a novel iris recognition system for long-range human identification. Eagle-Eyes is a multi-biometric system that is capable of acquiring face and both irises from multiple humans present anywhere in the capture volume. The iris acquisition system uses multiple cameras with hierarchically-ordered field of views, highly precise pan-tilt unit and large focal length zoom-focus lens. An enterprisewide solution for iris biometric processing is also presented. This solution handles enrollment, acquisition and processing of iris biometrics in a seamless device- and data- agnostic way. Experimental results are reported in an indoor office environment for multiple subject iris recognition. Index Terms—Biometrics, Iris acquisition, Iris recognition, Video surveillance. 1. INTRODUCTION Recognition of humans using unique biometric characteristics is of fundamental importance. Applications of biometrics technology are numerous and include homeland security, access control, user-specific services, etc. The iris is being established as a leading biometric trait and now has a body of empirical evidence to suggest that it is one of the most accurate. Notably, population-level performance results for iris recognition have recently been reported [3]. The iris biometric, however, has not become ubiquitous as compared to face and finger-prints. The face biometric trait has the advantage of being generally in plain sight and therefore lends itself to less constrained acquisition. Fingerprints have offered a higher accuracy solution but require contact (or near contact) with finger-print sensors. The finger-print trait is therefore not scalable in terms of

acquisition parameters, such as stand-off distance and capture volume. The iris biometric potentially offers matching accuracy exceeding that of finger-prints while sharing some of the potential advantages of face. As with the face the iris is generally in plain sight and therefore, theoretically at least, can be acquired given line-of-sight with a capture system. However, the optical resolution demands imposed by the iris biometric are significantly greater that those required for face recognition. It also requires appropriate NIR illumination. First generation iris capture systems solved these problems by imposing constraints on users during iris acquisition. An active area of research and development has the aim of relaxing these constraints. One measure of such constraints is the user’s proximity to the device; a second measure is the capture volume within which the user must place their iris. Constraints are also imposed on the motion of the user during acquisition. In most traditional iris acquisition systems, the subject has to place their eye close to the device fairly stably in a small volume for iris acquisition. This paper presents a novel prototype system to address the issues of stand-off distance, capture volume and subject motion in a scalable design. Eagle-Eyes is the first known multibiometric acquisition system that demonstrates scalable dual-eye iris recognition at a large stand-off distance (3-6 meters) and a large capture volume (3x2x3 m3). This system employs targeted laser illumination in an eye safe design that is scalable to a range of stand-off distances. The system captures the face and two irises of multiple subjects in the capture volume during acquisition phase. All of this processing is performed in real-time as the subjects pass through anywhere in the capture volume. ‘Eagle-EyesTM’ uses video surveillance techniques to acquire multiple biometrics in a scalable system design. A multi-camera system with standard VGA-resolution (752x480 pixels) digital sensors covering three hierarchical fields of view is used. The camera with the largest field of view (FOV), called scene camera, is used for wide area scene surveillance to detect and track humans. A narrower FOV camera, termed as face camera, is used to acquire higher resolution imagery of the subject’s face. The smallest FOV camera, named iris camera, uses a proprietary dualsensor camera design to acquire iris images. The Eagle-Eyes

system utilizes computer vision techniques for statistical background modeling and update, human detection, face detection and tracking. These techniques detect and track multiple human subjects. A pan-tilt unit moves the face and iris camera assembly to target the subject’s face and irises. Near infrared (NIR) laser illumination is targeted at the subject’s irises. Image acquisition software extracts acquired iris images that meet defined image quality thresholds. A suite of proprietary iris encoding and matching algorithms (EMIrisTM) are used to compare the acquired iris data with a database of enrolled dual-iris templates to search for the subject’s identity. Individuals are previous enrolled using a hand-held device for dual iris biometric enrollment (MobileEyesTM). The subject is thus identified to realize a secure physical access scenario in a minimally constrained setting. This paper is organized as follows: section 2 presents a survey of related work in the literature; section 3 outlines some application scenarios for our proposed system; section 4 enumerates some of the design challenges towards longrange iris acquisition system design; an enterprise-wide solution for iris biometric enrollment, long-range acquisition and matching is presented in section 5; the analysis of system performance is presented in section 6; finally section 7 summarizes the conclusions. 2. RELATED WORK The uniqueness of iris patterns has fueled the development of robust pattern recognition algorithms for recognition. Iris recognition methods generally process an iris image through following steps: localization, encoding and matching. For localization, Daugman [2] and Wildes et al [4] both proposed different methods of segmentation that involved estimating two non-concentric circles to mark iris and pupil boundaries. Once the iris boundaries have been localized, encoding of iris texture is used for iris representation. Daugman [2] used banks of two-dimensional Gabor filters, while Wildes et al [4] used a pyramidal representation of the Laplacian of Gaussian filter responses. Encoding method used by Daugman [2] results in binary representation of iris texture, called IrisCode, which is matched across the database using Hamming distance. In terms of long-range iris recognition, a feasibility study has been reported in [5]. They report iris recognition results from a proof-of-concept system for fixed stand-off distances of 5 and 10 meters using two custom-built telescopes with diameters of 10 and 20 cm. Although this work successfully suggested that iris recognition at larger distances was feasible, the strict constraints on subject’s eye positioning clearly reflected the fact that a better systems approach to subject’s eye localization was needed. Matey et al [6] presented an iris recognition system for moving subjects passing through a portal type structure. The system can capture images from both irises of subjects moving through a narrow portal with 20x20x10 cm3 capture volume at a stand-off distance of 3m.

Another approach to relax the constraints on subject’s location and movement during iris image capture is reported in [7]. They use a light stripe projector for depth estimation of the stationary subject in capture volume. Their system configuration processes a 1x1x1 m3 capture volume at a stand-off distance of 1.5 m. The above approaches have some major deficiencies in the context of long-range iris acquisition. One problem is that these approaches are based on non-scalable system design in terms of stand-off distance and capture volume. Also, they use very high resolution cameras for iris imaging with high cost, lower frame and reduced sensor efficiency. Situational awareness through intelligent video surveillance is an active area of research in recent literature. Hampapur et al [8] present a system with wide area video surveillance and pan-tilt-zoom functionality to capture facial biometrics of humans at a distance. Jain et al [11] present a system that uses master-slave camera assemblies to achieve wide-area surveillance and selective focus-of-attention. They report an online focus-of-attention scheme that correctly handles dynamic changes in scene and varying object depths. Hierarchical tracking with wide area surveillance and high resolution tracking is also addressed by Bashir et al [10]. A high-definition video camera with electronic pan-tiltzoom functionality, replaces a mechanical pan-tilt assembly, and is used for collaborative tracking. The object detection and tracking in low-resolution wide FOV is complemented by the tracking in high-resolution narrow FOV. Shah et al [9] present another aspect of intelligent video surveillance: distributed video surveillance across multiple cameras for wide area object tracking and behavioral analysis. Their system detects, categorizes and tracks moving objects in the scene observed by multiple cameras. 3. APPLICATION SCENARIOS A taxonomy of application environments for biometric systems is outlined in [1]. They propose a partitioning of the application scenarios into multiple categories. One important partitioning is ‘overt vs. covert’. If in an application environment, a user is aware of biometric data collection, the use is overt otherwise it is covert. Eagle-Eyes can be used in both overt and covert application scenarios. Another partition is ‘attended vs. unattended’, which refers to whether the system is operator-driven or not. Eagle-Eyes has the ability to be used as an unattended biometric acquisition system. Another partition is ‘indoor vs. outdoor’. Eagle-Eyes has been tested only in indoor environments so far. We propose that for an iris at a distance system, like Eagle-Eyes, an important classification is ‘moving vs. stationary’. If the subject has to stand still during biometric acquisition, then the application scenario is stationary subject. Eagle-Eyes is designed to deal with both stationary as well as moving (slow walking pace) subjects. We next present three

application scenarios for Eagle-Eyes operation based on the partitions of ‘over vs. covert’ and ‘stationary vs. moving’. 3.1. Overt-Stationary Subjects In this application scenario, subjects walking through an access point or corridor stand still at some non-designated location in the capture volume. They fixate their eyes in the general direction of the device and their face and two irises are acquired. The iris images are processed for authentication or verification. The specific applications include time of attendance at a facility, where the arrival of users (normally employees) is registered through their iris biometrics. This ensures that the person carrying the employee card is matched against their real identity to ensure the true identity of the person. Another specific application is physical access. In this application scenario, the access to the facility is allowed only to the individuals whose identity is proven and who have access credentials. In this context, Eagle-Eyes can be configured as a glance-andgo system. An environmentally robust version of the device can be configured to work outdoor, e.g. a check point, for drivers accessing a facility. 3.2. Overt-Moving Subjects In this application scenario, multiple moving subjects are processed for iris acquisition. The subjects are not required to stand still in the capture volume, so this scenario does not perform verification. Instead, the aim is to identify the individuals from their iris templates only. One specific application is high crowd volume processing, for example at airport terminals or train stations. In this context, the portaltype interface to realize an iris on the move system can be achieved using Eagle-Eyes by limiting the capture volume. 3.3. Covert-Moving Subjects In this application scenario, Eagle-Eyes is used to its fullest potential – as a covert, unattended multi-biometric acquisition system for multiple subjects in an unconstrained environment. The large capture volume under surveillance is monitored for the humans present in the capture volume. Individual humans are then targeted for face and iris acquisition in a covert manner. A specific application in this context is highly robust biometric identification in video surveillance for homeland security. Eagle-Eyes can be the recognition module within a larger video surveillance network where recognized subjects are handed off to the larger multi-camera video surveillance network for seamless tracking of humans in the full surveillance perimeter.

4. DESIGN CHALLENGES Some of the major deficiencies of the existing iris recognition systems lie in the system parameters of stand-off distance, capture volume, and subject motion constraints. These constraints exist because the human iris is a small organ in the body. The average human iris is around 11 mm in diameter. Imaging this small organ with appropriate illumination to bring out the intricate structure for robust representation is a daunting task. Design and implementation of systems to provide better accessibility for iris biometrics at large distances poses even more stringent angular accuracy requirements on the imaging system. The constraints on the above-mentioned system parameters directly map to relaxed requirements on system design issues. One such design issue is the NIR illumination of the iris region for image acquisition. For optimal illumination, the system has to be designed such that it generates uniform illumination profile across the iris region. This facilitates the generation of images of acceptable quality with high enough contrast. A conventional iris acquisition system achieves this by placing a set of NIR LEDs in close proximity to the imaging device. This type of illumination is depth-limited in nature and illuminates only a narrow capture volume at a small stand-off distance. Hence the limitations on stand-off distance and capture volume. Another design issue is the number of pixels needed across the iris region. The iris image data interchange standard [12] rates an image with 200 pixels across the iris region as high quality iris image. This standard puts a lower limit on the number of pixels across iris region of 100 pixels under low quality iris image. A conventional iris acquisition system takes advantage of the proximity of the iris with imaging system; a constraint that dramatically reduces the requirements on both lens size and number of pixels (resolution) of the imaging sensor. Finally, the combination of constraints on stand-off distance and subject motion results in the fact that during acquisition process, subject’s iris is placed stably in the capture volume. This is due to the fact that optical design parameters of focal length and depth of focus can be traded off to yield only fully focused images at a certain distance from the iris acquisition system. This constraint relaxes the requirements on the algorithms for iris region extraction from the image sequence. A long-range iris acquisition system relaxes the constraints on stand-off distance, capture volume and subject motion. These relaxed constraints for higher accessibility of the system come at the price of more demanding system design. Firstly, a uniform NIR illumination of the iris region can not be achieved by a set of low-power LED-based illumination as in conventional iris acquisition systems. Power requirements for the illumination increase with standoff distance and capture volume. Eye safety considerations restrict the irradiance on the surface of iris under increasing power levels. A related design consideration is that of

illuminating a fixed region in the capture volume or moving the illumination region inside the capture volume with the subject. Targeted illumination that tracks with the subject is potentially more scalable but can suffer from problems associated with retro-reflections. The second major design issue of pixel resolution on the iris region can not be addressed with inexpensive off the shelf lenses and sensors as in conventional iris acquisition systems. The focal length and magnification requirements for large distance iris acquisition need to be considered. This issue can be addressed by a combination of zoom lens and sensor resolution Relaxing constraints imposed on the motion of the subject, within a non-portal large standoff distance scenario, introduces the need for dynamic subject motion estimators and accurate motion control of the imaging hardware. A final design issue deals with the fact that in conventional iris acquisition systems, the imaging of non-eye regions can be reasonably minimized. For a large capture volume iris acquisition system, the potential of non-eye iris regions getting imaged increases significantly. 5. ENTERPRISE-WIDE IRIS BIOMETRIC SOLUTION This section provides an enterprise-level solution developed at Retica Systems for iris biometrics processing. The solution developed at Retica includes a hand-held device for iris biometric enrollment (Mobile-EyesTM); a device-agnostic software development kit (SDK) to process iris data (EMIrisTM); and finally a system for iris acquisition at a distance (Eagle-EyesTM). This iris processing solution developed at Retica addresses all the aspects of application scenarios for long-range iris recognition outlined in section 3.

5.1. Iris Enrollment Enrollment of a subject’s iris biometric is achieved using a hand-held device developed at Retica Systems. MobileEyesTM is a hand-held device which performs simultaneous dual-iris acquisition. This device uses NIR LEDs to project a uniform illumination profile on subject’s iris region. An innovative optical system design uses a set of lenses and VGA-resolution imaging sensors to generate wellilluminated images of a subject’s irises at video rate. The device has a user-friendly binocular-like form-factor and weights about 2.5 lbs. Two cameras in the device simultaneously generate video streams of images that meet the high quality specifications of iris image standard, ISO/IEC 19794-6 [12]. This device operates in an open, non-proprietary way and images captured by this device can be used with any third-party iris processing software solutions. 5.2. Iris Encoder-Matcher SDK The SDK features a set of software components that can be used in a device-agnostic way to process the iris images captured using any iris acquisition device. Iris image data in several image formats is accepted; the iris region is segmented from the image; the segmented region is encoded in to Retica’s iris template; the template data is matched against records in a database. The encoding and matching algorithms have been demonstrated to have one of the top match rate performances in the industry. On NIST-ICE data set, these algorithms generated a TAR (True Accept Rate) performance of 99.96% at an FAR (False Accept Rate) of 10-4 using dual-eye fusion. The algorithms also generate iris image quality scores. The iris image quality is measured in terms of focus as well as coverage, i.e. how much of the iris is occluded by eyelids, eyelashes, and reflections. This functionality helps select better images for enrollment when used with Mobile-EyesTM device. In the currently highlighted enterprise-wide solution, this SDK is used to encode and store iris biometrics acquired using the MobileEyes device. This SDK is also used as iris recognition engine in the Eagle-Eyes system as explained in the next section. 5.3. Eagle-Eyes Architecture

Figure 1: Enterprise-wide solution architecture for iris biometric enrollment with Mobile-Eyes and long-range multi-biometric acquisition using Eagle-Eyes. Iris processing is performed through EMIris SDK.

The Eagle-Eyes system forms the final piece of the solution for long-range iris recognition highlighted in this paper. This system has been designed to acquire face and two irises from multiple subjects in a large capture volume. Our system overcomes the design challenges outlined in section 4 using an innovative opto-mechanical system design. The designed system is integrated with novel video surveillance algorithms to acquire face and iris biometrics at a large stand-off distance. This system also employs Retica’s

device-agnostic EMIris SDK for iris processing as shown in Figure 1. The innovative opto-mechanical system design of EagleEyes is shown in Figure 2. The figure shows multiple cameras with their respective fields of view. The fixed scene camera is used for wide area scene surveillance to detect and track humans. A scheduler ranks faces in the scene FOV and directs the PTU in sequence to all active subjects. A list of previous acquisition events prevents repeat acquisitions of the same subject. The face camera, which has a narrower FOV is used to acquire a higher resolution image of the subject’s face. This camera is mounted on the pan-tilt unit assembly. Images generated from this camera are also used to locate and track the subject’s eyes. A target point is located on the subject’s face mid-way between their two eyes. The problem of long-distance iris illumination is solved using an innovative design of a laser illuminator. The designed laser illumination propagates a collimated beam that maintains its uniform illumination profile up to large distances. This is in stark contrast to the existing longerrange iris acquisition approaches, such as [6] and [7], that illuminate a fixed region in the capture volume. Iris image resolution requirements are addressed using a scalable optical design. A larger focal length zoom lens is used in conjunction with a custom-designed dual sensor iris camera. The dual sensor iris camera is made up of two standard VGA-resolution image sensors. This again, is in contrast with the design of other systems where imaging sensors are scaled up in terms of pixel resolution [6] and [7]. The motion of the subject is accounted for by a subject-servoloop that tracks both the motion of the PTU assembly and the subject. Eye tracking in the face camera is combined with range information using an optical rangefinder. The velocity of the PTU assembly and subject are matched such that their relative positions coincide at predicted future times/locations. The zoom and focus of the iris lens are controlled to match the subject’s range trajectory. The issue of the acquisition of non-iris images is addressed by a threelevel hierarchical processing. The first level processes information at the human level, so non-human regions can be filtered from further processing. The second level process information at the level of subject’s face. This processing makes sure that non-eye regions are not processed for iris localization. The third level of iris localization is then activated only when the system has been positioned to acquire a subject’s irises. The images acquired from the NIR iris cameras are then processed using the EMIris SDK. The iris images are segmented for pupil and iris localization. The captured iris images are encoded to generate iris templates, which are matched against existing enrolled iris templates for ‘1:N’ comparison. The subject is thus identified to realize an iris recognition scenario in a minimally constrained setting. A publish/subscribe software architecture is employed whereby processing models publish signals that are subscribed to by

relevant connected processing modules. Communication via TCP-IP allows distribution across multiple PCs. This software architecture is scalable and adaptable to multiple configurations servicing different scenarios.

Figure 2: System schematic for Eagle-Eyes. 6. SYSTEM PERFORMANCE ANALYSIS The enterprise-wide biometric solution with iris recognition at a distance presented in this paper has been tested in various indoor settings over the past year. The system has been tested with multiple subjects in capture volume for face and iris acquisition. The system reliably locates subjects present in the capture volume and targets one subject after the other. Preliminary experiments have aimed to measure acquisition times and iris match rates. For single stationary subjects acquisition times averaged 6.1 seconds. For three subjects three sequential acquisitions averaged X.X seconds. Sample frames from face and dual-iris cameras are shown in Figure 3.

Figure 3: Sample acquired multi-biometric data from a subject at a stand-off distance of 4.5 m.

This figure shows the detected and tracked locations of face and both eyes in the face camera. Also, the geometries of both pupil and iris are shown in the two images from dualiris camera. Table 1 shows the matching rates for the EagleEyes acquisition of thirteen subjects as matched against a database of Mobile-Eyes acquisitions. True-Match-Rates (TMR) are quoted for first acquisition attempts, given three successive attempts and averages over all 65 acquisitions (5 repeat acquisitions x 13 subjects). TMR 1st attempt 3 attempts all attempts 3.5m 10/13 (77%) 12/13 (92%) 51/65 (78%) 4.5m 10/13 (77%) 12/13 (92%) 52/65 (80%) Table 1. True-Match-Rates (TMR) for iris acquisitions at 3.5 and 4.5 meters. 7. CONCLUSIONS This paper presents a system for long-range multi-biometric acquisition. Our system addresses the issues of small standoff distance and capture volume in conventional iris recognition systems. The system is capable of acquiring face and both irises from multiple humans present anywhere in the capture volume. The iris acquisition system uses multiple cameras with hierarchically-ordered fields of views, highly precise pan-tilt unit and large focal length zoom-focus lens. An enterprise-wide solution for iris enrollment, long-range multi-biometric acquisition and iris data processing is presented as developed by Retica Systems. The system is tested in indoor environment for multiple subject iris recognition. System performance results, in terms of dualiris recognition times, at different stand-off distances are presented to highlight fast system performance anywhere in the capture volume. Sample acquired face and iris images are also presented to highlight the unprecedented multibiometric acquisition performance of the system. 8. ACKNOWLEDGEMENTS The authors would like to acknowledge the efforts of management, research and development teams at Retica Systems for making Eagle-Eyes a reality. REFERENCES [1] J. Wayman, A. Jain, D. Maltoni, D. Maio (eds.), “Biometric Systems: Technology, Design and Performance Evaluation”, Springer-Verlag, 2005. ISBN: 1852335963. page 3. [2] J. Daugman, “How iris recognition works”, IEEE Trans. On Circuits and Systems for Video Technology, CSVT, Vol. 14(1), January 2004, pp. 21 - 30. [3] J. Daugman, “Probing the Uniqueness and Randomness of IrisCodes: Results from 200 Billion Iris Pair Comparisons”, Proceedings of the IEEE, Vol. 94(11), November 2006. pp. 1927-1935.

[4] R. Wildes, “Iris Recognition”, In: J. Wayman, A. Jain, D. Maltoni, D. Maio (Eds.) Biometrics Systems: Technology, Design and Performance Evaluation. Springer, pp. 63-95. [5] C. Fancourt, L. Bogoni, K. Hanna, Y. Guao, R. Wildes, N. Takahashi, U. Jain, “Iris Recognition at a Distance”, In T. Kanade, A. Jain, N. Ratha (eds.) AVBPA 2005. LNCS, Vol. 3546, pp. 1-13. Springer, Heidelberg (2005). [6] J. Matey, O. Naroditsky, K. Hanna, R. Kolczynski, D. Loiacono, S. Magru, M. Tinker, T. Zappia, W. Zhao, “Iris on the Move: Acquisition of Images for Iris Recognition in Less Constrained Environments”, Proceedings of the IEEE, Vol. 94(11), Nov. 2006. [7] S. Yoon, H. Jung, J. Suhr, J. Kim, “Non-intrusive Iris Image Capturing System Using Light Stripe Projection and PanTilt-Zoom Camera”, IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2007. [8] A. Hampapur, L. Brown,H. Connell, A. Ekin, N. Haas, M. Lu, H. Merkl, S. Pankanti, A. Senior, C. Shu, Y. Tian, “Smart Video Surveillance: Exploring the concept of multi-scale spatiotemporal tracking”, IEEE Signal Processing Magazine, March 2005, pp. 38-51. [9] M. Shah, O. Javed, K. Shafique, “Automated Visual Surveillance in Realistic Scenarios”, IEEE Multimedia, Vol. 14(1), Jan-March 2007, pp. 30-39. [10] F. Bashir, F. Porikli, “Collaborative Tracking of Objects in EPTZ Cameras”, Visual Communications and Image Processing, VCIP 2007, Vol. 6508 (1). [11] A. Jain, D. Kopell, K. Kakligian, Y. Wang, “Using Stationary-Dynamic Camera Assemblies for Wide-area Video Surveillance and Selective Attention”, IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2006. [12] “Information Technoloty. Biometric Data Interchange Formats. Iris Image Data”, ISO/IEC 19794-6:2005. [13] D. Sliney, M. Wolbarsht, “Safety with Laser and Other Optical Sources”, New York: Plenum, 1980.

Eagle-Eyes: A System for Iris Recognition at a Distance

has the advantage of being generally in plain sight and therefore lends ... dual-eye iris recognition at a large stand-off distance (3-6 meters) and a ... Image acquisition software extracts acquired iris images that .... Hence the limitations on stand-off distance and ... addressed with inexpensive off the shelf lenses and sensors.

908KB Sizes 3 Downloads 275 Views

Recommend Documents

Eagle-Eyes: A System for Iris Recognition at a Distance
novel iris recognition system for long-range human identification. Eagle-Eyes is a ... physical access scenario in a minimally constrained setting. This paper is ...

Review of Iris Recognition System Iris Recognition System Iris ... - IJRIT
Abstract. Iris recognition is an important biometric method for human identification with high accuracy. It is the most reliable and accurate biometric identification system available today. This paper gives an overview of the research on iris recogn

Review of Iris Recognition System Iris Recognition System Iris ...
It is the most reliable and accurate biometric identification system available today. This paper gives an overview of the research on iris recognition system. The most ... Keywords: Iris Recognition, Personal Identification. 1. .... [8] Yu Li, Zhou X

An Effective Segmentation Method for Iris Recognition System
Biometric identification is an emerging technology which gains more attention in recent years. ... characteristics, iris has distinct phase information which spans about 249 degrees of freedom [6,7]. This advantage let iris recognition be the most ..

Recognition at a Long Distance: Very Low Resolution ...
This work is supported in part by the Ministry of Science and Technology of ... Transactions on Pattern Analysis and Machine Intelligence,. 2012. [5] S. Biswas et ...

Healing at a Distance
comparisons, repeated words, cause-and-effect, and emphasized words. Particularly notice imperatives (commands) and verbs (action words), which are like tree limbs. .... Heavenly Father, I give You my life, my future, and my family today. Please help

EF-45 Iris Recognition System - CMITECH.pdf
Whoops! There was a problem loading more pages. EF-45 Iri ... ITECH.pdf. EF-45 Iris ... MITECH.pdf. Open. Extract. Open with. Sign In. Details. Comments. General Info. Type. Dimensions. Size. Duration. Location. Modified. Created. Opened by me. Shari

Efficient Small Template Iris Recognition System Using ...
in illumination, pupil size and distance of the eye from the camera. ..... With a pre-determined separation Hamming distance, a decision can be made as to ...

A Review: Study of Iris Recognition Using Feature Extraction ... - IJRIT
INTRODUCTION. Biometric ... iris template in database. There is .... The experiments have been implemented using human eye image from CASAI database.

A Review: Study of Iris Recognition Using Feature Extraction ... - IJRIT
analyses the Iris recognition method segmentation, normalization, feature extraction ... Keyword: Iris recognition, Feature extraction, Gabor filter, Edge detection ...

Sparse Distance Learning for Object Recognition ... - Washington
objects, we define a view-to-object distance where a novel view is .... Google 3D Warehouse. ..... levels into 18 (0◦ −360◦) and 9 orientation bins (0◦ −180◦),.

accent tutor: a speech recognition system - GitHub
This is to certify that this project prepared by SAMEER KOIRALA AND SUSHANT. GURUNG entitled “ACCENT TUTOR: A SPEECH RECOGNITION SYSTEM” in partial fulfillment of the requirements for the degree of B.Sc. in Computer Science and. Information Techn

Working at a Cynical Distance: Implications for Power ...
customers into a sale even when the agent knows that they could ..... Opportunities for .... He has recently joined the Warwick Business School as a lecturer.

A System for Recognition of Biological Patterns in ...
A System for Recognition of Biological Patterns in Toxins Using. Computational Intelligence. Bernardo Penna Resende de Carvalho, Thais Melo Mendes, Ricardo de Souza Ribeiro,. Ricardo Fortuna, José Marcos Veneroso and Maurıcio de Alvarenga Mudado. A

Recognition of qualification obtained through Distance Education.PDF
Recognition of qualification obtained through Distance Education.PDF. Recognition of qualification obtained through Distance Education.PDF. Open. Extract.

Recognition of qualification obtained through Distance Education.PDF ...
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. Recognition of ...

A Distributed Speech Recognition System in Multi-user Environments
services. In other words, ASR on mobile units makes it possible to input various kinds of data - from phone numbers and names for storage to orders for business.

Face Authentication /Recognition System For Forensic Application ...
Graphic User Interface (GUI) is a program interface item that allows people to interact with the programs in more ways than just typing commands. It offers graphical icons, and a visual indicator, as opposed to text-based interfaces, typed command la

BigTable: A System for Distributed Structured ... - Research at Google
2. Motivation. • Lots of (semi-)structured data at Google. – URLs: • Contents, crawl metadata ... See SOSP'03 paper at http://labs.google.com/papers/gfs.html.

A Collaborative Tool for Synchronous Distance Education
application in a simulated distance education setting. The application combines video-conference with a networked virtual environment in which the instructor and the students can experiment ..... Virtual Campus: Trends for Higher Education and. Train