Gianluca Sforza

Biometric Recognition in Automated Border Control: A Survey

The increasing demand for traveler clearance at international border crossing points (BCPs) has motivated research for finding more efficient solutions. Automated border control (ABC) is emerging as a solution to enhance the convenience of travelers, the throughput of BCPs, and national security. This is the first comprehensive survey on the biometric techniques and systems that enable automatic identity verification in ABC. We survey the biometric literature relevant to identity verification and summarize the best practices and biometric techniques applicable to ABC, relying on real experience collected in the field. Furthermore, we select some of the major biometric issues raised and highlight the open research areas.

A Texture-Based Image Processing Approach for the Description of Human Oocyte Cytoplasm

The purpose of this paper is to develop a diagnostic tool that can analyze light microscope images of human oocytes and derive a description of the oocyte cytoplasm that is useful for quality assessment in assisted insemination. The proposed approach includes three main phases: 1) segmentation; 2) feature extraction; and 3) clustering. In the segmentation phase, a region of interest inside the cytoplasm is extracted through morphological operators and the Hough transform. In the second phase, regions that result from segmentation are processed through a multiresolution texture analysis to extract a set of features that describe different levels of cytoplasm granularity. To this aim, we evaluate some statistics in the Haar wavelet transform domain. Finally, the extracted features are used to cluster oocytes according to different levels of granularity. This approach is made by fuzzy clustering. Experimental results on a collection of microscope images of oocytes are reported to show the effectiveness of the proposed approach. In addition, comparison with alternative methods for feature extraction and clustering is performed.

Using Adaptive Thresholding and Skewness Correction to Detect Gray Areas in Melanoma In Situ Images

The incidence of melanoma in situ (MIS) is growing significantly. Detection at the MIS stage provides the highest cure rate for melanoma, but reliable detection of MIS with dermoscopy alone is not yet possible. Adjunct dermoscopic instrumentation using digital image analysis may allow more accurate detection of MIS. Gray areas are a critical component of MIS diagnosis, but automatic detection of these areas remains difficult because similar gray areas are also found in benign lesions. This paper proposes a novel adaptive thresholding technique for automatically detecting gray areas specific to MIS. The proposed model uses only MIS dermoscopic images to precisely determine gray area characteristics specific to MIS. To this aim, statistical histogram analysis is employed in multiple color spaces. It is demonstrated that skew deviation due to an asymmetric histogram distorts the color detection process. We introduce a skew estimation technique that enables histogram asymmetry correction facilitating improved adaptive thresholding results. These histogram statistical methods may be extended to detect any local image area defined by histograms.

Automatic Classification of Acquisition Problems Affecting Fingerprint Images in Automated Border Controls

Automated Border Control (ABC) systems are technologies designed to increase the speed and accuracy of the identity verifications performed at international borders. A great number of ABCs deployed in different countries use fingerprint recognition techniques because of their high accuracy and user acceptability. However, the accuracy of fingerprint recognition methods can drastically decrease in this application context due to user-sensor interaction factors. This paper presents two main contributions. The first of them consists in an experimental evaluation performed to search the main negative aspects that could affect the usability and accuracy in ABCs based on fingerprint biometrics. The mainly considered aspects consists in the presence of luggage and cleanness of the finger skin. The second contribution consists in a novel approach for automatically identifying the type of user-sensor interaction that caused quality degradations in fingerprint samples. This method uses a specific feature set and computational intelligence techniques to detect non-idealities in the acquisition process and to suggest corrective actions to travelers and border guards. To the best of our knowledge, this is the first method in the literature designed to detect problems in user-sensor interaction different from improper pressures on the acquisition surface. We validated the proposed approach using a dataset of 2880 images simulating different scenarios typical of ABCs. Results shown that the proposed approach is feasible and can obtain satisfactory performance, with a classification error of 0.098.

Adaptive ECG biometric recognition: a study on re-enrollment methods for QRS signals

The diffusion of wearable and mobile devices for the acquisition and analysis of cardiac signals drastically increased the possible applicative scenarios of biometric systems based on electrocardiography (ECG). Moreover, such devices allow for comfortable and unconstrained acquisitions of ECG signals for relevant time spans of tens of hours, thus making these physiological signals particularly attractive biometric traits for continuous authentication applications. In this context, recent studies showed that the QRS complex is the most stable component of the ECG signal, but the accuracy of the authentication degrades over time, due to significant variations in the patterns for each individual. Adaptive techniques for automatic template update can therefore become enabling technologies for continuous authentication systems based on ECG characteristics.

Towards touchless pore fingerprint biometrics: A neural approach

Touchless fingerprint recognition systems are being increasingly used for a fast, hygienic, and distortion-free recognition. However, due to the greater complexity of the algorithms required for processing touchless fingerprint samples, currently only Level 1 and Level 2 features are being used for recognition, and Level 3 features are used only in touch-based optical devices with about 1000 ppi resolution. In this paper, we propose the first innovative method in the literature able to extract Level 3 features, in particular sweat pores, from fingerprint images captured with a touchless acquisition using a commercial off-the-shelf camera. The method uses image processing algorithms to extract a set of candidate sweat pores. Then, computational intelligence techniques based on neural networks are used to learn the local features of the real pores, and select only the actual sweat pores from the set of candidate points. The results show the validity of the proposed methodology, with the majority of the pores correctly extracted, indicating that a touchless fingerprint recognition using Level 3 features is feasible.

Shape annotation for intelligent image retrieval

Annotation of shapes is an important process for semantic image retrieval. In this paper, we present a shape annotation framework that enables intelligent image retrieval by exploiting in a unified manner domain knowledge and perceptual description of shapes. A semi-supervised fuzzy clustering process is used to derive domain knowledge in terms of linguistic concepts referring to the semantic categories of shapes. For each category we derive a prototype that is a visual template for the category. A novel visual ontology is proposed to provide a description of prototypes and their salient parts. To describe parts of prototypes the visual ontology includes perceptual attributes that are defined by mimicking the analogy mechanism adopted by humans to describe the appearance of objects. The effectiveness of the developed framework as a facility for intelligent image retrieval is shown through results on a case study in the domain of fish shapes.

Multiresolution texture analysis for human oocyte cytoplasm description

In this work we present an approach based on image texture analysis to obtain a description of oocyte cytoplasm which could aid the medical expert in the selection of oocytes to be used for assisted insemination. More specifically, we describe some characteristics such as different levels of uniformity and/or granularity in the oocyte cytoplasm, using multiresolution texture analysis applied to light microscope images. To this aim, we evaluate some statistical measures in the wavelet transform domain of image regions and classify them according to different levels of granularity. Preliminary experimental results on a collection of light microscope images of oocytes are reported to show the effectiveness of the proposed approach.

Improving OSB wood panel production by vision-based systems for granulometric estimation

Oriented Strand Board (OSB) is a kind of engineered wood particle board widely adopted in manufacturing, construction and logistics. The production of OSB panels requires rectangular-shaped wood strands of specific size, arranged in layers to form the so-called “mattress” (mat) and bonded together with glue. The structural properties of the panel rely directly on the layer forming. In particular, the size distribution - namely granulometry - of the strands should fulfill standard measures to reach the mechanical properties expected from the panel. Offline granulometry of particle materials is the most commonly procedure used to evaluate the production process, but it is prone to several drawbacks owing to the manual intervention of human operators. Vision-based systems, instead, allow for performing granulometric analyses in an automatic and contactless manner. We propose a computer vision approach to estimate the granulometry of wood strands. The designed framework analyzes images of a mat of strands placed over a moving conveyor belt, and provides useful information and measurements to enhance the production of OSB panels. Because of the very large amount of wood strands on the mat, particle-overlapping is frequent and represents a main issue for measurement algorithms. In order to overcome this problem, our framework joins image processing and computational intelligence methods, such as edge detection and fuzzy color clustering. We tested the framework with real and synthetic images, useful to variate the conditions of the materials shape. The obtained results demonstrate the ability of our approach to evaluate the granulometry of the strands in real conditions, and robustness against the simulated variations of the production process.

Enhancing the Performance of Multimodal Automated Border Control Systems

Biometric recognition in Automated Border Control (ABC) systems is performed in response to an increased worldwide traffic, by automatically verifying the identity of the passenger during border crossing. Currently, ABC systems seldom use methods for multimodal biometric fusion, which have been proved to increase the recognition accuracy, due to technological and privacy limitations. This paper proposes a framework for the biometric fusion in ABC systems, with the features of being technology-neutral and privacy- compliant, by performing an analysis of the most suitable biometric fusion techniques for ABC systems and considering the current technical and legal limitations.