Analisi degli errori di estrazione di minuzie in impronte di bassa qualita

Lo scopo di questa tesi è quello di analizzare i casi in cui un particolare algoritmo di estrazione delle minuzie da immagini di impronte digitali compie notevoli errori, e di migliorarne le prestazioni.

Acquisizione e riconoscimento di impronte digitali mediante smartphone

Nel presente studio è stato affrontato il problema di effettuare l’acquisizione di impronte digitali mediante la fotocamera di uno Smartphone. Si tratta di un’applicazione potenzialmente molto interessante in quanto l’utilizzo di uno Smartphone renderebbe molto più semplice l’acquisizione delle impronte, non essendo necessari dispositivi specifici come, ad esempio, scanner d’impronte digitali. D’altra parte, l’utilizzo di una fotocamera per l’acquisizione delle impronte introduce diverse problematiche, fra cui individuare l’area del dito corrispondente all'impronta, valutare la qualità di un'immagine e determinare quali minuzie estratte corrispondano effettivamente a quelle di interesse. Questo studio conferma la fattibilità di un sistema del genere che risulta essere in grado di fornire buone prestazioni di riconoscimento biometrico.

Analisi della qualità locale per aumentare l'accuratezza nel riconoscimento di impronte digitali

Il riconoscimento di impronte digitali viene largamente utilizzato in molteplici ambiti, dai sistemi di sicurezza alle applicazioni forensi. La maggior parte dei sistemi automatici di riconoscimento si basa sull'estrazione e il confronto delle minuzie nelle impronte digitali, ma l’accuratezza di queste operazioni peggiora sensibilmente in presenza di impronte di scarsa qualità, fino ad arrivare anche a compromettere il riconoscimento stesso. In questo lavoro di tesi ci si è posto come obiettivo il miglioramento delle prestazioni di un algoritmo di estrazione delle minuzie, attraverso l’utilizzo di varie tecniche per l’analisi di qualità locale delle impronte digitali.

Evaluation of two-view geometry methods with automatic ground-truth generation

A large number of methods have been published that aim to evaluate various components of multi-view geometry systems. Most of these have focused on the feature extraction, description and matching stages (the visual front end), since geometry computation can be evaluated through simulation. Many data sets are constrained to small scale scenes or planar scenes that are not challenging to new algorithms, or require special equipment. This paper presents a method for automatically generating geometry ground truth and challenging test cases from high spatio-temporal resolution video. The objective of the system is to enable data collection at any physical scale, in any location and in various parts of the electromagnetic spectrum. The data generation process consists of collecting high resolution video, computing accurate sparse 3D reconstruction, video frame culling and down sampling, and test case selection. The evaluation process consists of applying a test 2-view geometry method to every test case and comparing the results to the ground truth. This system facilitates the evaluation of the whole geometry computation process or any part thereof against data compatible with a realistic application. A collection of example data sets and evaluations is included to demonstrate the range of applications of the proposed system.

Calibration Method for Underwater Visual Ground-Truth System

This work presents an automatic calibration method for a vision based external underwater ground-truth positioning system. These systems are a relevant tool in benchmarking and assessing the quality of research in underwater robotics applications. A stereo vision system can in suitable environments such as test tanks or in clear water conditions provide accurate position with low cost and flexible operation. In this work we present a two step extrinsic camera parameter calibration procedure in order to reduce the setup time and provide accurate results. The proposed method uses a planar homography decomposition in order to determine the relative camera poses and the determination of vanishing points of detected lines in the image to obtain the global pose of the stereo rig in the reference frame. This method was applied to our external vision based ground-truth at the INESC TEC/Robotics test tank. Results are presented in comparison with an precise calibration performed using points obtained from an accurate 3D LIDAR modelling of the environment.

From Craton to Rift: Empirically Based Ground-Truth Criteria for Local Events Recorded on Regional Networks

Region-specific empirically based ground-truth (EBGT) criteria used to estimate the epicentral-location accuracy of seismic events have been developed for the Main Ethiopian Rift and the Tibetan plateau. Explosions recorded during the Ethiopia-Afar Geoscientific Lithospheric Experiment (EAGLE), the International Deep Profiling of Tibet, and the Himalaya (INDEPTH III) experiment provided the necessary GT0 reference events. In each case, the local crustal structure is well known and handpicked arrival times were available, facilitating the establishment of the location accuracy criteria through the stochastic forward modeling of arrival times for epicentral locations. In the vicinity of the Main Ethiopian Rift, a seismic event is required to be recorded on at least 8 stations within the local Pg/Pn crossover distance and to yield a network-quality metric of less than 0.43 in order to be classified as EBGT5(95%) (GT5 with 95% confidence). These criteria were subsequently used to identify 10 new GT5 events with magnitudes greater than 2.1 recorded on the Ethiopian Broadband Seismic Experiment (EBSE) network and 24 events with magnitudes greater than 2.4 recorded on the EAGLE broadband network. The criteria for the Tibetan plateau are similar to the Ethiopia criteria, yet slightly less restrictive as the network-quality metric needs to be less than 0.45. Twenty-seven seismic events with magnitudes greater than 2.5 recorded on the INDEPTH III network were identified as GT5 based on the derived criteria. When considered in conjunction with criteria developed previously for the Kaapvaal craton in southern Africa, it is apparent that increasing restrictions on the network-quality metric mirror increases in the complexity of geologic structure from craton to plateau to rift. Accession Number: WOS:000322569200012

Seasonal cycles of monthly means of ground truth wind speed of METAR station GVAC and of area-averaged TMI wind speed at 37 GHz for the Cape Verde Island Sal

In the present paper ground truth and remotely sensed datasets were used for the investigation and quantification of the impact of Saharan dust on microwave propagation, the verification of theoretical results, and the validation of wind speeds determined by satellite microwave sensors. The influence of atmospheric dust was verified in two different study areas by investigations of single dust storms, wind statistics, wind speed scatter plots divided by the strength of Saharan dust storms, and wind speed differences in dependence of microwave frequencies and dust component of aerosol optical depth. An increase of the deviations of satellite wind speeds to ground truth wind speeds with higher microwave frequencies, with stronger dust storms, and with higher amount of coarse dust aerosols in coastal regions was obtained. Strong Saharan dust storms in coastal areas caused mean relative errors in the determination of wind speed by satellite microwave sensors of 16.3% at 10.7 GHz and of 20.3% at 37 GHz. The mean relative errors were smaller in the open sea area with 3.7% at 10.7 GHz and with 11.9% at 37 GHz.

A Ground-Truth Video Dataset for the Development and Evaluation of Vision-based Sense-and-Avoid systems

The importance of vision-based systems for Sense-and-Avoid is increasing nowadays as remotely piloted and autonomous UAVs become part of the non-segregated airspace. The development and evaluation of these systems demand flight scenario images which are expensive and risky to obtain. Currently Augmented Reality techniques allow the compositing of real flight scenario images with 3D aircraft models to produce useful realistic images for system development and benchmarking purposes at a much lower cost and risk. With the techniques presented in this paper, 3D aircraft models are positioned firstly in a simulated 3D scene with controlled illumination and rendering parameters. Realistic simulated images are then obtained using an image processing algorithm which fuses the images obtained from the 3D scene with images from real UAV flights taking into account on board camera vibrations. Since the intruder and camera poses are user-defined, ground truth data is available. These ground truth annotations allow to develop and quantitatively evaluate aircraft detection and tracking algorithms. This paper presents the software developed to create a public dataset of 24 videos together with their annotations and some tracking application results.

Detecting outliers and asserting consistency in agriculture ground truth information by using temporal VI data from modis.

Collecting ground truth data is an important step to be accomplished before performing a supervised classification. However, its quality depends on human, financial and time ressources. It is then important to apply a validation process to assess the reliability of the acquired data. In this study, agricultural infomation was collected in the Brazilian Amazonian State of Mato Grosso in order to map crop expansion based on MODIS EVI temporal profiles. The field work was carried out through interviews for the years 2005-2006 and 2006-2007. This work presents a methodology to validate the training data quality and determine the optimal sample to be used according to the classifier employed. The technique is based on the detection of outlier pixels for each class and is carried out by computing Mahalanobis distances for each pixel. The higher the distance, the further the pixel is from the class centre. Preliminary observations through variation coefficent validate the efficiency of the technique to detect outliers. Then, various subsamples are defined by applying different thresholds to exclude outlier pixels from the classification process. The classification results prove the robustness of the Maximum Likelihood and Spectral Angle Mapper classifiers. Indeed, those classifiers were insensitive to outlier exclusion. On the contrary, the decision tree classifier showed better results when deleting 7.5% of pixels in the training data. The technique managed to detect outliers for all classes. In this study, few outliers were present in the training data, so that the classification quality was not deeply affected by the outliers.

Fusione di impronte: analisi dei problemi di sicurezza e possibili contromisure

In questa tesi si esamineranno alcune possibili vulnerabilità dei sistemi di riconoscimento di impronte digitali e si tenterà di migliorare la loro sicurezza nei confronti di una tipologia specifica di attacco che utilizza impronte digitali "artificiali" per permettere ad un criminale di utilizzare il documento di un complice. È stata infatti recentemente dimostrata la possibilità di inserire in un documento elettronico caratteristiche biometriche che lo rendono utilizzabile da due diverse persone. Questa problematica di sicurezza è alla base dell’attacco che verrà analizzato in questa tesi e per il quale si cercheranno contromisure efficaci.