The Smart Borders initiative will have an mayor impact on the perspective border crossing performance in Schengen. In October 2014 the European Commission published the "Technical Study " aiming at identifying and assessing the most suitable and promising options and solutions for the implementation. In 2015 the proof of concept shall be finalized by an European pilot project led by the IT agency Eu-LISA. Germany will start its operation in the context of this pilot project on the 22nd of June 2015. The presentation will give an insight on the first findings of this evaluation.

• projective transformation
• face recognition
• forensic scenario
• pitch deviation

• Face Tracking
• Likelihood Map
• Optical Flow
• AdaBoost Cascade Classifier

• biometrics
• gait recognition
• silhouette quality
• feature fusion

• iris recognition
• segmentation
• fusion

• Palm and finger vein authentication
• Near infrared palm and finger images
• Prototypes
• Databases

Dense frequency and orientation maps are needed, for low-quality image processing in general, image enhancement in fingerprints in particular. Such dense maps open up also a novel research question since a good estimation of both, allows continuous ridge counting. The interrelationship of both maps, and how to estimate them when none is known initially will be the theoretical focus of this presentation whereas the applications will be drawn from forensic fingerprints.

We will present the idea of a logarithmic scale space driven by the trace of the Structure Tensor, which has been utilized to estimate dense orientation maps since 3 decades, to obtain the dense frequency map itself. In contrast to the common scale space, the log-scale space is non-linear and absolute frequency estimation turns out to be an orientation estimation in it.

Both frequency and orientation maps of the same picture can successively improve each other, combined in an image enhancement scheme.  In fingerprints, such an image enhancement scheme can be obtained simply by projections of local images on Gabor filters, dynamically calculated (or pointed at) by using the dense frequency and orientation maps. Analytic and experimental comparisons with emphasis on understanding the relationship of the novel scheme to state of the art on forensic fingerprints,  where the ground truth is known, as test bed will be discussed. Among others examples of ridge-counting,  without ridge detection, thinning, etc, and at fractions of the local fingerprint period along any path (not only between minutiae) will be discussed.

• Fingerprint
• Quality
• Forensics

• Fingermark
• Identification
• Mobile phone capture
• Sufficient evidential value

• Fingerprint
• security
• cryptography

• alignment-free
• fingerprint features
• improved fuzzy vault scheme
• unlinkable
• quantization scheme
• clustering

• biometric template protection
• honey object
• irreversibility
• face recognition

Human hands are easier to present, convenient to be imaged and can reveal variety of features representing anatomy of humans for the biometric identification. Hand-based biometrics has been employed for over 2000 years and continues to reveal new features, when images under multiple spectrums and imaging resolutions, to uniquely identify individuals. This talk will detail on two recent advancements on hand-based biometrics, i.e. finger knuckle and 3D fingerprint identification, which can initiate new applications in law-enforcement, immigration, consumer and mobile biometrics. I will also outline fundamental significance of such research in significantly altering the theoretically believed limits on the accuracy of identifying the human population on planet using the most popular hand biometric modality today.

• iris segmentation
• eyelid segmentation
• structured learning

• fingerprint
• fingerprint indexing
• identification system
• minutia

• Fingerprint Classification
• Directional Patterns
• Orientation Fields
• Singular Points
• Region Segmentation