Francisco Assis da Silva

A Step Towards the Automated Diagnosis of Parkinson's Disease: Analyzing Handwriting Movements

Parkinsons disease (PD) has affected millions of people world-wide, being its major problem the loss of movements and, consequently, the ability of working and locomotion. Although we can find several works that attempt at dealing with this problem out there, most of them make use of datasets composed by a few subjects only. In this work, we present some results toward the automated diagnosis of PD by means of computer vision-based techniques in a dataset composed by dozens of patients, which is one of the main contributions of this work. The dataset is part of a joint research project that aims at extracting both visual and signal-based information from healthy and PD patients in order to go forward the early diagnosis of PD patients. The dataset is composed by handwriting clinical exams that are analyzed by means of image processing and machine learning techniques, being the preliminary results encouraging and promising. Additionally, a new quantitative feature to measure the amount of tremor of an individuals handwritten trace called Mean Relative Tremor is also presented.

A new computer vision-based approach to aid the diagnosis of Parkinson's disease

BACKGROUND AND OBJECTIVE Even today, pointing out an exam that can diagnose a patient with Parkinsons disease (PD) accurately enough is not an easy task. Although a number of techniques have been used in search for a more precise method, detecting such illness and measuring its level of severity early enough to postpone its side effects are not straightforward. In this work, after reviewing a considerable number of works, we conclude that only a few techniques address the problem of PD recognition by means of micrography using computer vision techniques. Therefore, we consider the problem of aiding automatic PD diagnosis by means of spirals and meanders filled out in forms, which are then compared with the template for feature extraction. METHODS In our work, both the template and the drawings are identified and separated automatically using image processing techniques, thus needing no user intervention. Since we have no registered images, the idea is to obtain a suitable representation of both template and drawings using the very same approach for all images in a fast and accurate approach. RESULTS The results have shown that we can obtain very reasonable recognition rates (around ≈67%), with the most accurate class being the one represented by the patients, which outnumbered the control individuals in the proposed dataset. CONCLUSIONS The proposed approach seemed to be suitable for aiding in automatic PD diagnosis by means of computer vision and machine learning techniques. Also, meander images play an important role, leading to higher accuracies than spiral images. We also observed that the main problem in detecting PD is the patients in the early stages, who can draw near-perfect objects, which are very similar to the ones made by control patients.

ALPRS - A New Approach for License Plate Recognition Using the SIFT Algorithm

This paper presents a new approach for the automatic license plate recognition, which includes the SIFT algorithm in step to locate the plate in the input image. In this new approach, besides the comparison of the features obtained with the SIFT algorithm, the correspondence between the spatial orientations and the positioning associated with the keypoints is also observed. Afterwards, an algorithm is used for the character recognition of the plates, very fast, which makes it possible its application in real time. The results obtained with the proposed approach presented very good success rates, so much for locating the characters in the input image, as for their recognition.

Intrusion Detection System Based On Flows Using Machine Learning Algorithms

The use of technology information and communication by different types of devices generates a large quantity of data packets that contains of confidential and personal information. The traffic of data packet can be summarized in network flow. Due this reason, it is necessary to use computer security tools, such as Intrusion Detection Systems (IDS). This work presents an IDS that can perform the flow- based analysis (netflow). This research conducted an analysis on flows previously collected and properly detected of three different types of attacks. The flows were organized to be processed by machine learning methods. The results obtained by proposed approach were very promising. Also, this work aimed at building a public dataset to be used by researchers worldwide in order to foster IDS-related research.

A New Method for Automatic Vehicle License Plate Detection

License plate recognition has been widely studied, and the advance in image capture technology helps enhance or create new methods to achieve this objective. In this work is presented a method for real time detection and segmentation of car license plates based on image analyzing and processing techniques. The results show that the computational cost and accuracy rate considering the proposed approach are acceptable to real time applications, with an execution time under 1 second. The proposed method was validated using two datasets (A and B). It was obtained over 92% detection success for dataset A, 88% in digit segmentation for datasets A and B, and 95% digits classification accuracy rate for dataset B.

StitchingPHm--A new algorithm for panoramic images

In this paper it is presented a new algorithm for the construction of panoramic images (including 360 panoramas), which has as main characteristic to avoid the distortion that occurs by joining of several successive images in a sequence. We used the SIFT and RANSAC algorithms to find overlap areas between pairs of images, as well as a Blend algorithm for smoothing the joints. As the proposed algorithm doesn’t cause distortions, the subsequent correction is not necessary, contributing to a better performance. The results of experiments using commercial software and also the proposed algorithm were compared through a visual analysis. In addition, a quantitative analysis was done using numeric measures calculated on a panoramic image, generated from an image sequence of a region mapped and georeferenced by Google Earth.

USO DO APRENDIZADO DE MÁQUINA NO DIAGNÓSTICO MÉDICO DE PATOLOGIAS

A tecnologia se desenvolve rapidamente em diferentes areas do conhecimento, uma destas areas e a medicina. Este trabalho apresenta uma proposta de automatizacao de diagnostico de patologias utilizando tecnicas e metodos classificadores de Aprendizagem de Maquina (AM). Atraves dos metodos que serao explanados e implementados, as informacoes contidas nas bases de dados serao analisadas e classificadas, gerando resultados. Com a utilizacao dessas tecnicas, atraves de maquinas para o diagnostico medico, patologias poderao ser detectadas num estagio menos avancado da doenca e com maior precisao, quando comparado a diagnostico inteiramente humano sofre influencias de fatores externos, o que pode afetar no diagnostico do paciente. Neste trabalho busca-se analisar dados e classifica-los de acordo com os metodos a serem citados, e por fim pode ser definido o metodo aplicado mais viavel e eficaz.

IDENTIFICAÇÃO DE SCANNERS A PARTIR DE IMAGENS/DOCUMENTOS DIGITAIS

Em diversas areas do conhecimento o uso de imagens digitais e de grande importância, e conhecer a sua origem e autenticidade e algo primordial; pois falsificacoes podem ser feitas facilmente. A criacao de metodos que possibilitem identificar dispositivos geradores de imagens se torna fundamental. Este trabalho apresenta um ferramental matematico que identifica e classifica imagens provenientes de um determinado scanner. A construcao do metodo foi feita utilizando a transformada wavelet e algumas tecnicas de thresholding e filtragem. Paralelamente foi criada uma base de imagens digitalizadas, utilizando tres modelos de scanners, para testar e validar a ferramenta desenvolvida. Os resultados dos testes realizados foram considerados satisfatorios, pois obtiveram a acuracia media de 87,5% de acertos.

VISUALIZAÇÃO DE AMBIENTES INTERNOS ATRAVÉS DE INTERATIVIDADE COMPUTACIONAL PROVIDA POR IMAGENS PANORÂMICAS

This paper presents a way to view and interact in indoor environments, using sequences of panoramic images with positioning registered in the environment during the way used to perform the capture of images. The viewing of indoor environments can be used for various purposes, such as cultural, commercial, among others. A camera LadyBug 5 was used with a rate of 15 frames per second to capture the image sequences. A tool to perform the cataloging of the captured images was developed, forming a segment for each imaged way. We used these segments with the images in a Web application that we developed using HTML5, CSS3, JavaScript, and WebGL, which allows viewing and interactivity on the images of environment.

UM NOVO MÉTODO PARA IDENTIFICAÇÃO DE CÂMERA A PARTIR DE IMAGENS DIGITAIS

The identification of the camera responsible for the capture of a given image may be used as indicia of various kinds of crimes. This identification can occur with the use of a scientific method that can determine which camera is the generator of a particular image. Each camera generates images with some peculiarities due to their own features. Within this context, this work was based on the creation of a new generating camera identification method from images, through studies and implementation of digital image processing algorithms. A new database was also generated for conducting training and testing of the proposed method.