MACHINE VISION FOR AUTOMATICVISUAL INSPECTION OF WOODENRAILWAY SLEEPERS USING UNSUPERVISED NEURAL NETWORKS

The motivation for this thesis work is the need for improving reliability of equipment and quality of service to railway passengers as well as a requirement for cost-effective and efficient condition maintenance management for rail transportation. This thesis work develops a fusion of various machine vision analysis methods to achieve high performance in automation of wooden rail track inspection.The condition monitoring in rail transport is done manually by a human operator where people rely on inference systems and assumptions to develop conclusions. The use of conditional monitoring allows maintenance to be scheduled, or other actions to be taken to avoid the consequences of failure, before the failure occurs. Manual or automated condition monitoring of materials in fields of public transportation like railway, aerial navigation, traffic safety, etc, where safety is of prior importance needs non-destructive testing (NDT).In general, wooden railway sleeper inspection is done manually by a human operator, by moving along the rail sleeper and gathering information by visual and sound analysis for examining the presence of cracks. Human inspectors working on lines visually inspect wooden rails to judge the quality of rail sleeper. In this project work the machine vision system is developed based on the manual visual analysis system, which uses digital cameras and image processing software to perform similar manual inspections. As the manual inspection requires much effort and is expected to be error prone sometimes and also appears difficult to discriminate even for a human operator by the frequent changes in inspected material. The machine vision system developed classifies the condition of material by examining individual pixels of images, processing them and attempting to develop conclusions with the assistance of knowledge bases and features.A pattern recognition approach is developed based on the methodological knowledge from manual procedure. The pattern recognition approach for this thesis work was developed and achieved by a non destructive testing method to identify the flaws in manually done condition monitoring of sleepers.In this method, a test vehicle is designed to capture sleeper images similar to visual inspection by human operator and the raw data for pattern recognition approach is provided from the captured images of the wooden sleepers. The data from the NDT method were further processed and appropriate features were extracted.The collection of data by the NDT method is to achieve high accuracy in reliable classification results. A key idea is to use the non supervised classifier based on the features extracted from the method to discriminate the condition of wooden sleepers in to either good or bad. Self organising map is used as classifier for the wooden sleeper classification.In order to achieve greater integration, the data collected by the machine vision system was made to interface with one another by a strategy called fusion. Data fusion was looked in at two different levels namely sensor-level fusion, feature- level fusion. As the goal was to reduce the accuracy of the human error on the rail sleeper classification as good or bad the results obtained by the feature-level fusion compared to that of the results of actual classification were satisfactory.

A low cost one-camera optical tracking system for indoor wide-area augmented and virtual reality environments

In the last years the number of industrial applications for Augmented Reality (AR) and Virtual Reality (VR) environments has significantly increased. Optical tracking systems are an important component of AR/VR environments. In this work, a low cost optical tracking system with adequate attributes for professional use is proposed. The system works in infrared spectral region to reduce optical noise. A highspeed camera, equipped with daylight blocking filter and infrared flash strobes, transfers uncompressed grayscale images to a regular PC, where image pre-processing software and the PTrack tracking algorithm recognize a set of retro-reflective markers and extract its 3D position and orientation. Included in this work is a comprehensive research on image pre-processing and tracking algorithms. A testbed was built to perform accuracy and precision tests. Results show that the system reaches accuracy and precision levels slightly worse than but still comparable to professional systems. Due to its modularity, the system can be expanded by using several one-camera tracking modules linked by a sensor fusion algorithm, in order to obtain a larger working range. A setup with two modules was built and tested, resulting in performance similar to the stand-alone configuration.

Solução rápida das ambigüidades GPS para aplicações no posicionamento relativo de linhas de base curtas

Pós-graduação em Ciências Cartográficas - FCT

Monitoramento das reações fisiológicas em sujeitos autistas decorrentes da influência dos sons: um estudo em Engenharia biomédica

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Glossário bilíngue dos termos fundamentais do setor feirístico: busca de equivalências em inglês

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

ZTD em tempo quase real via Estações GNSS Terrestres: estratégia de processamento para o GIPSY-OASIS II e combinação das séries temporais

Pós-graduação em Ciências Cartográficas - FCT

Análise das tensões em próteses implantossuportadas esplintadas, variando o tipo de conexão protética e a inclinação do implante: estudo pelo método dos elementos finitos tridimensionais não linear

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Dental CT imaging as a screening tool for dental profiling: advantages and limitations

The use of dental processing software for computed tomography (CT) data (Dentascan) is described on postmortem (pm) CT data for the purpose of pm identification. The software allows reconstructing reformatted images comparable to conventional panoramic dental radiographs by defining a curved reconstruction line along the teeth on oblique images. Three corpses that have been scanned within the virtopsy project were used to test the software for the purpose of dental identification. In every case, dental panoramic images could be reconstructed and compared to antemortem radiographs. The images showed the basic component of teeth (enamel, dentin, and pulp), the anatomic structure of the alveolar bone, missing or unerupted teeth as well as restorations of the teeth that could be used for identification. When streak artifacts due to metal-containing dental work reduced image quality, it was still necessary to perform pm conventional radiographs for comparison of the detailed shape of the restoration. Dental identification or a dental profiling seems to become possible in a noninvasive manner using the Dentascan software.

Computer-aided operation planning and 3-dimensional osteotomy guidance for alveolar distraction

Objectives: In alveolar distraction, in cases of severe atrophy in particular, it is often difficult to perform osteotomies in order to make a transport segment in optimal size and shape. Moreover care must be taken, not to damage the closely locating anato- mical structures such as the maxillary sinus, the inferior alveolar nerve, and the roots of the neighboring teeth. For setting ideal osteotomy lines exactly, we have developed a CT-based preoperative planning tool. Methods: 3-dimensional visual reconstruction of the jaw is created from the preoperative CT scans (1.0-mm slice thick- ness). Using the image-processing software Mimics (Materialise, Yokohama, Japan), various procedures of virtual cutting are simulated first to determine optimal osteotomy lines and to design an ideal transport segment. After the computer planning, data from the virtual solid model are transferred to a rapid prototype model, and a guiding splint is made to transfer the planned surgical simulation to the actual surgery. Results: The method was used in a case of severe atrophy of the anterior maxilla. The patient had a large maxillary sinus requir- ing a precise osteotomy in this critical area. Using the splint allowing a 3-dimensional guidance, alveolar osteotomies were easily done to achieve a transport segment in sufficient dimen- sion as planned, and any perforation of the maxillary sinus could be avoided. Finally the alveolar distraction of 10mm has suc- cessfully been performed. Conclusion: The preoperative planning method and the guiding splint described here are useful in problematic cases requiring an extremely precise osteotomy due to lack of bony space.

THE HOLOGRAPHIC DETECTOR FOR CLOUDS FROM TRL 5 TO TRL 8

Clouds are one of the most influential elements of weather on the earth system, yet they are also one of the least understood. Understanding their composition and behavior at small scales is critical to understanding and predicting larger scale feedbacks. Currently, the best method to study clouds on the microscale is through airborne in situ measurements using optical instruments capable of resolving clouds on the individual particle level. However, current instruments are unable to sufficiently resolve the scales important to cloud evolution and behavior. The Holodec is a new generation of optical cloud instrument which uses digital inline holography to overcome many of the limitations of conventional instruments. However, its performance and reliability was limited due to several deficiencies in its original design. These deficiencies were addressed and corrected to advance the instrument from the prototype stage to an operational instrument. In addition, the processing software used to reconstruct and analyze digitally recorded holograms was improved upon to increase robustness and ease of use.

Quality in emergency departments: a study on 3,285,440 admissions

INTRODUCTION: A multi-centre study has been conducted, during 2005, by means of a questionnaire posted on the Italian Society of Emergency Medicine (SIMEU) web page. Our intention was to carry out an organisational and functional analysis of Italian Emergency Departments (ED) in order to pick out some macro-indicators of the activities performed. Participation was good, in that 69 ED (3,285,440 admissions to emergency services) responded to the questionnaire. METHODS: The study was based on 18 questions: 3 regarding the personnel of the ED, 2 regarding organisational and functional aspects, 5 on the activity of the ED, 7 on triage and 1 on the assessment of the quality perceived by the users of the ED. RESULTS AND CONCLUSION: The replies revealed that 91.30% of the ED were equipped with data-processing software, which, in 96.83% of cases, tracked the entire itinerary of the patient. About 48,000 patients/year used the ED: 76.72% were discharged and 18.31% were hospitalised. Observation Units were active in 81.16% of the ED examined. Triage programmes were in place in 92.75% of ED: in 75.81% of these, triage was performed throughout the entire itinerary of the patient; in 16.13% it was performed only symptom-based, and in 8.06% only on-call. Of the patients arriving at the ED, 24.19% were assigned a non-urgent triage code, 60.01% a urgent code, 14.30% a emergent code and 1.49% a life-threatening code. Waiting times were: 52.39 min for non-urgent patients, 40.26 min for urgent, 12.08 for emergent, and 1.19 for life-threatening patients.

Radiological factors related to pre-operative hearing levels in patients with vestibular schwannomas

The pathogenetic mechanism of hearing loss in patients with vestibular schwannomas (VS) remains unclear. Our aim was to determine the radiological and clinical parameters that might be related to hearing. The radiological images and charts of 99 patients were reviewed. Image processing software was used to analyse the maximal tumor diameter in three planes; its volume; its extension cranially, caudally, anteriorly and posteriorly; the width and length of the intrameatal tumor portion, its shape and consistency; and the tumor-fundus distance. These parameters were correlated with the patient's pre-operative hearing range. The degree of hearing correlated significantly with the tumor size, volume and coronal diameter, the degree of intrameatal tumor growth, and the distance between the lateral tumor end and the fundus (p < 0.05). No correlation was found regarding tumor extension, shape and consistency, the presence of hydrocephalus, or the extent of erosion of the internal auditory canal. Loss of hearing in the VS appears to be multifactorial. Determining the radiological parameters related to the hearing level can help to clarify the pathophysiological mechanisms involved.

Noninvasive pressure difference mapping derived from 4D flow MRI in patients with unrepaired and repaired aortic coarctation.

PURPOSE To develop a method for computing and visualizing pressure differences derived from time-resolved velocity-encoded three-dimensional phase-contrast magnetic resonance imaging (4D flow MRI) and to compare pressure difference maps of patients with unrepaired and repaired aortic coarctation to young healthy volunteers. METHODS 4D flow MRI data of four patients with aortic coarctation either before or after repair (mean age 17 years, age range 3-28, one female, three males) and four young healthy volunteers without history of cardiovascular disease (mean age 24 years, age range 20-27, one female, three males) was acquired using a 1.5-T clinical MR scanner. Image analysis was performed with in-house developed image processing software. Relative pressures were computed based on the Navier-Stokes equation. RESULTS A standardized method for intuitive visualization of pressure difference maps was developed and successfully applied to all included patients and volunteers. Young healthy volunteers exhibited smooth and regular distribution of relative pressures in the thoracic aorta at mid systole with very similar distribution in all analyzed volunteers. Patients demonstrated disturbed pressures compared to volunteers. Changes included a pressure drop at the aortic isthmus in all patients, increased relative pressures in the aortic arch in patients with residual narrowing after repair, and increased relative pressures in the descending aorta in a patient after patch aortoplasty. CONCLUSIONS Pressure difference maps derived from 4D flow MRI can depict alterations of spatial pressure distribution in patients with repaired and unrepaired aortic coarctation. The technique might allow identifying pathophysiological conditions underlying complications after aortic coarctation repair.

A consistent combination of GNSS and SLR with minimum constraints

In this article, the realization of a global terrestrial reference system (TRS) based on a consistent combination of Global Navigation Satellite System (GNSS) and Satellite Laser Ranging (SLR) is studied. Our input data consists of normal equation systems from 17 years (1994– 2010) of homogeneously reprocessed GPS, GLONASS and SLR data. This effort used common state of the art reduction models and the same processing software (Bernese GNSS Software) to ensure the highest consistency when combining GNSS and SLR. Residual surface load deformations are modeled with a spherical harmonic approach. The estimated degree-1 surface load coefficients have a strong annual signal for which the GNSS- and SLR-only solutions show very similar results. A combination including these coefficients reduces systematic uncertainties in comparison to the singletechnique solution. In particular, uncertainties due to solar radiation pressure modeling in the coefficient time series can be reduced up to 50 % in the GNSS+SLR solution compared to the GNSS-only solution. In contrast to the ITRF2008 realization, no local ties are used to combine the different geodetic techniques.We combine the pole coordinates as global ties and apply minimum constraints to define the geodetic datum. We show that a common origin, scale and orientation can be reliably realized from our combination strategy in comparison to the ITRF2008.

Morphologic predictors of aortic expansion in chronic type B aortic dissection.

AIM To identify morphologic factors affecting aortic expansion in patients with uncomplicated type B aortic dissections. METHODS Computed tomography data of 24 patients (18 male; median age: 61 years), diagnosed with acute uncomplicated type B aortic dissections between 2002 and 2013, were retrospectively reviewed. All patients had at least two computed tomography angiography scans and six months of uneventful follow-up. Computed tomography scans were assessed by two independent readers with regard to presence and number of entry tears. Thoracic and abdominal aortic diameters were derived using image processing software. RESULTS Twenty-two of 24 patients showed aortic expansion over a median computed tomography angiographic follow-up of 33.2 months. Annual rates showed an increase of 1.7 mm for total aortic diameter, 2.1 mm for the false and a decrease of -0.4 mm for the true lumen. In three patients (12.5%), aortic diameter exceeded 60 mm during follow-up, and all three patients underwent thoracic endovascular aortic repair. Patients with a maximum aortic diameter <4 cm at baseline showed a significantly higher expansion rate compared to cases with an initial maximum aortic diameter of ≥4 cm (p=0.0471). A median of two entries (range: 1-5) was recognized per patient. Presence of more than two entry tears (n = 13) was associated with faster overall diameter expansion (mean annual rates: 2.18 mm vs. 1.16 mm; p = 0.4556), and decrease of the cross-sectional surface of the true lumen over time (annual rate for > 2 entries vs. ≤2 entries: -7.8 mm(2) vs. +37.5 mm(2); p = 0.0369). Median size of entry tears was 12 mm (range: 2-53 mm). CONCLUSIONS The results presented herein suggest that uncomplicated type B aortic dissection patients with more than two entry tears and/or an initial maximum aortic diameter of<4 cm are at risk for aortic dilatation and, therefore, may require stricter follow-up including the possible need for early intervention.