Advanced Driver Assistance System (ADAS) was widely learned nowadays. As crucial parts of ADAS, lane markings detection, as well as other objects detection, have become more popular than before. However, most methods implemented in such areas cannot perfectly balance the performance of accuracy versus efficiency, and the mainstream methods (e.g. Machine Learning) suffer from several limitations which can hardly break the wall between partial autonomous and fully autonomous driving. This thesis proposed a real-time lane marking detection framework for ADAS, which included 4-extreme points set descriptor and a rule-based cascade classifier. By analyzing the behavior of lane markings on the road surface, a characteristic of markings was discovered, i.e., standard markings can sustain their shape in the perpendicular plane of the driving direction. By employing this feature, a 4-extreme points set descriptor was applied to describe the shape of each marking first. Specifically, after processing Maximally Stable Extremal Region (MSER) and Hough transforms on a 2-D image, several contours of interest are obtained. A bounding box, with borders parallel to the image coordinate, intersected with each contour at 4 points in the edge, which was named 4-extreme points set. Afterward, to verify consistency of each contour and standard marking, some rules abstracted from construction manual are employed such as Area Filter, Colour Filter, Relative Location Filter, Convex Filter, etc. To reduce the errors caused by changes in driving direction, an enhanced module was then introduced. By tracking the vanishing point as well as other key points of the road net, a method for 3-D reconstruction, with respect to the optical axis between vanishing point and camera center, is possible. The principle of such algorithm was exhibited, and a description about how to obtain the depth information from this model was also provided. Among all of these processes, a key-point based classification method is the main contribution of this paper because of its function in eliminating the deformation of the object caused by inverse perspective mapping. Several experiments were conducted in highway and urban roads in Ottawa. The detection rate of the markings by the proposed algorithm reached an average accuracy rate of 96.77% while F1 Score (harmonic mean of precision and recall) also attained a rate of 90.57%. In summary, the proposed method exhibited a state-of-the-art performance and represents a significant advancement of understanding.

Den här rapporten har utrett möjligheten att automatiskt identifiera diken frånflygburet insamlat LiDAR-data. Den metod för identifiering som har valts harförst skapat en höjdbild från LiDAR-data. Därefter har den tagit fram kandidatertill diken genom att vektorisera resultatet från en linjedetektering. Egenskaper-na för dikeskandidaterna har sedan beräknats genom en analys av höjdprofilerför varje enskild kandidat, där höjdprofilerna skapats utifrån ursprungliga data.Genom att filtrera kandidaterna efter deras egenskaper kan dikeskartor med an-vändarspecificerade mått på diken presenteras i ett vektorformat som underlättarvidare användning. Rapporten beskriver hur algoritmen har implementerats ochpresenterar också exempel på resultat. Efter en analys av algoritmen samt förslagpå förbättringar presenteras den viktigaste behållningen av rapporten; Att det ärmöjligt med automatisk detektering av diken. / This Master’s thesis is investigating the possibility of automatically identifyingditches in airborne collected LiDAR data. The chosen approach to identificationcommences by creating an elevation picture from the LiDAR data. Then it usesthe result of a line detection to exhibit candidates for ditches. The properties forthe various candidates are calculated through an analysis of the elevation profile forthe candidates, where the elevation profiles are created from the original data. Byfiltering the candidates according to their calculated properties, maps with ditchesconforming to user-specified limits are created and presented in vector format.This thesis describes how the algorithm is implemented and gives examples ofresults. After an analysis of the algorithm and a proposal for improvements, itis suggested that automatic detection of ditches in LiDAR collected data is anachievable objective.

No description available.

In this study, a calibration setup for laser beam positioning systems used in visual inspection applications in industry is designed and manufactured. The laser positioning systems generate movable parallel laser lines on the projection surface. There are several translational and angular error sources affecting the positioning accuracy of the laser lines on the projection surface. Especially, since the laser line positioning error caused by angular error sources increases with the distance between the laser system and the projection surface, angular parameters of the laser sources should be measured and adjusted precisely. The calibration setup developed in this study detects the laser line positions at two different projection distances by means of laser sensing cameras which are positioned precisely along the laser lines and laser positioning axis which is perpendicular to these lines. Cameras detect the positions of the laser lines which are directed to the camera sensors with micrometer repeatability by means of some special imaging algorithms. The precise positioning of the cameras requires a special camera positioning system. For this reason, the disturbances like temperature changes and vibration should be minimized. In order to provide a suitable environment for the calibration system, special tests are conducted and a special calibration room is constituted. Construction inside the room is also made by considering the required ambient parameters. Finally, several verification tests of the calibration system are conducted.

This thesis is carried out in the context of Advanced Driver Assistance Systems, and especially autonomous vehicles. Its aim is to propose a method to enhance localization of vehicles on roads. I suggests using a camera to detect lane markings, and to match these to a map to extract the corrected position of the vehicle.The thesis is divided in three parts dealing with: the map, the line detector and the evaluation. The map is based on the OpenStreetMap data. The line detector is a based on ridge detection. The results are compared with an Iterative Closest Point algorithm. It also focuses on implementing the components under a real-time constraint. Technologies such as ROS, for synchronization of the data, and CUDA, for parallelization, are used.

No description available.

Document page segmentation is one of the most crucial steps in document image analysis. It ideally aims to explain the full structure of any document page, distinguishing text zones, graphics, photographs, halftones, figures, tables, etc. Although to date, there have been made several attempts of achieving correct page segmentation results, there are still many difficulties. The leader of the project in the framework of which this PhD work has been funded (*) uses a complete processing chain in which page segmentation mistakes are manually corrected by human operators. Aside of the costs it represents, this demands tuning of a large number of parameters; moreover, some segmentation mistakes sometimes escape the vigilance of the operators. Current automated page segmentation methods are well accepted for clean printed documents; but, they often fail to separate regions in handwritten documents when the document layout structure is loosely defined or when side notes are present inside the page. Moreover, tables and advertisements bring additional challenges for region segmentation algorithms. Our method addresses these problems. The method is divided into four parts:1. Unlike most of popular page segmentation methods, we first separate text and graphics components of the page using a boosted decision tree classifier.2. The separated text and graphics components are used among other features to separate columns of text in a two-dimensional conditional random fields framework.3. A text line detection method, based on piecewise projection profiles is then applied to detect text lines with respect to text region boundaries.4. Finally, a new paragraph detection method, which is trained on the common models of paragraphs, is applied on text lines to find paragraphs based on geometric appearance of text lines and their indentations. Our contribution over existing work lies in essence in the use, or adaptation, of algorithms borrowed from machine learning literature, to solve difficult cases. Indeed, we demonstrate a number of improvements : on separating text columns when one is situated very close to the other; on preventing the contents of a cell in a table to be merged with the contents of other adjacent cells; on preventing regions inside a frame to be merged with other text regions around, especially side notes, even when the latter are written using a font similar to that the text body. Quantitative assessment, and comparison of the performances of our method with competitive algorithms using widely acknowledged metrics and evaluation methodologies, is also provided to a large extend.(*) This PhD thesis has been funded by Conseil Général de Seine-Saint-Denis, through the FUI6 project Demat-Factory, lead by Safig SA

Lire automatiquement le texte présent dans les documents permet de rendre accessible les informations qu'ils contiennent. Pour réaliser la transcription de pages complètes, la localisation des lignes de texte est une étape cruciale. Les méthodes traditionnelles de détection de lignes, basées sur des approches de traitement d'images, peinent à généraliser à des jeux de données hétérogènes. Pour cela, nous proposons dans cette thèse une approche par réseaux de neurones profonds. Nous avons d'abord proposé une approche de segmentation mono-dimensionnelle des paragraphes de texte en lignes à l'aide d'une technique inspirée des modèles de reconnaissance, où une classification temporelle connexionniste (CTC) est utilisée pour aligner implicitement les séquences. Ensuite, nous proposons un réseau qui prédit directement les coordonnées des boîtes englobant les lignes de texte. L'ajout d'un terme de confiance à ces boîtes hypothèses permet de localiser un nombre variable d'objets. Nous proposons une prédiction locale des objets afin de partager les paramètres entre les localisations et, ainsi, de multiplier les exemples d'objets vus par chaque prédicteur de boîte lors de l'entraînement. Cela permet de compenser la taille restreinte des jeux de données utilisés. Pour récupérer les informations contextuelles permettant de prendre en compte la structure du document, nous ajoutons, entre les couches convolutionnelles, des couches récurrentes LSTM multi-dimensionnelles. Nous proposons trois stratégies de reconnaissance pleine page qui permettent de tenir compte du besoin important de précision au niveau des positions et nous montrons, sur la base hétérogène Maurdor, la performance de notre approche pour des documents multilingues pouvant être manuscrits et imprimés. Nous nous comparons favorablement à des méthodes issues de l'état de l'art. La visualisation des concepts appris par nos neurones permet de souligner la capacité des couches récurrentes à apporter l'information contextuelle. / Being able to automatically read the texts written in documents, both printed and handwritten, makes it possible to access the information they convey. In order to realize full page text transcription, the detection and localization of the text lines is a crucial step. Traditional methods tend to use image processing based approaches, but they hardly generalize to very heterogeneous datasets. In this thesis, we propose to use a deep neural network based approach. We first propose a mono-dimensional segmentation of text paragraphs into lines that uses a technique inspired by the text recognition models. The connexionist temporal classification (CTC) method is used to implicitly align the sequences. Then, we propose a neural network that directly predicts the coordinates of the boxes bounding the text lines. Adding a confidence prediction to these hypothesis boxes enables to locate a varying number of objects. We propose to predict the objects locally in order to share the network parameters between the locations and to increase the number of different objects that each single box predictor sees during training. This compensates the rather small size of the available datasets. In order to recover the contextual information that carries knowledge on the document layout, we add multi-dimensional LSTM recurrent layers between the convolutional layers of our networks. We propose three full page text recognition strategies that tackle the need of high preciseness of the text line position predictions. We show on the heterogeneous Maurdor dataset how our methods perform on documents that can be printed or handwritten, in French, English or Arabic and we favourably compare to other state of the art methods. Visualizing the concepts learned by our neurons enables to underline the ability of the recurrent layers to convey the contextual information.

Schistosoma mansoni é o parasita causador da esquistossomose mansônica que, de acordo com o Ministério da Saúde do Brasil, afeta atualmente vários milhões de pessoas no país. Uma das formas de diagnóstico da esquistossomose é a detecção de ovos do parasita através da análise de lâminas microscópicas com material fecal. Esta tarefa é extremamente cansativa, principalmente nos casos de baixa endemicidade, pois a quantidade de ovos é muito pequena. Nesses casos, uma abordagem computacional para auxílio na detecção de ovos facilitaria o trabalho de diagnóstico. Os ovos têm formato ovalado, possuem uma membrana translúcida, apresentam uma espícula e sua cor é ligeiramente amarelada. Porém nem todas essas características são observadas em todos os ovos e algumas delas são visíveis apenas com uma ampliação adequada. Além disso, o aspecto visual do material fecal varia muito de indivíduo para indivíduo em termos de cor e presença de diversos artefatos (tais como partículas que não são desintegradas pelo sistema digestivo), tornando difícil a tarefa de detecção dos ovos. Neste trabalho investigamos, em particular, o problema de detecção das linhas que contornam a borda de vários dos ovos. Propomos um método composto por duas fases. A primeira fase consiste na detecção de estruturas do tipo linha usando operadores morfológicos. A detecção de linhas é dividida em três etapas principais: (i) realce de linhas, (ii) detecção de linhas, e (iii) refinamento do resultado para eliminar segmentos de linhas que não são de interesse. O resultado dessa fase é um conjunto de segmentos de linhas. A segunda fase consiste na detecção de subconjuntos de segmentos de linha dispostos em formato elíptico, usando um algoritmo baseado na transformada Hough. As elipses detectadas são fortes candidatas a contorno de ovos de S. mansoni. Resultados experimentais mostram que a abordagem proposta pode ser útil para compor um sistema de auxílio à detecção dos ovos. / Schistosoma mansoni is one of the parasites which causes schistosomiasis. According to the Brazilian Ministry of Health, several million people in the country are currently affected by schistosomiasis. One way of diagnosing it is by egg identification in stool. This task is extremely time-consuming and tiring, especially in cases of low endemicity, when only few eggs are present. In such cases, a computational approach to help the detection of eggs would greatly facilitate the diagnostic task. Schistosome eggs present oval shape, have a translucent membrane and a spike, and their color is slightly yellowish. However, not all these features are observed in every egg and some of them are visible only with an adequate microscopic magnification. Furthermore, the visual aspect of the fecal material varies widely from person to person in terms of color and presence of different artifacts (such as particles which are not disintegrated by the digestive system), making it difficult to detect the eggs. In this work we investigate the problem of detecting lines which delimit the contour of the eggs. We propose a method comprising two steps. The first phase consists in detecting line-like structures using morphological operators. This line detection phase is divided into three steps: (i) line enhancement, (ii) line detection, and (iii) result refinement in order to eliminate line segments that are not of interest. The output of this phase is a set of line segments. The second phase consists in detecting subsets of line segments arranged in an elliptical shape, using an algorithm based on the Hough transform. Detected ellipses are strong candidates to contour of S. mansoni eggs. Experimental results show that the proposed approach has potential to be effectively used as a component in a computer system to help egg detection.

Cette thèse s’inscrit dans le cadre d’une convention industrielle de formation par la recherche (CIFRE) entre le laboratoire Hubert Curien et l’entreprise Pattyn Bakery Division. L’objectif de ces travaux est le développement d’un système de détection de trait laser sur FPGA (Field Programmable Gate Array) qui soit plus performant que système actuel de l’entreprise. Dans l’industrie, les concepteurs de systèmes de vision doivent pouvoir créer et modifier facilement leurs systèmes afin de pouvoir les adapter aux besoins de leurs clients et aux évolutions technologiques. Ainsi les opérateurs développés doivent être génériques afin de permettre aux concepteurs de modifier le système de vision sans nécessairement avoir de compétences matérielles. Les concepteurs doivent également pouvoir être en mesure d’estimer quelles seront les ressources utilisées par l’opérateur en cas modifications du système : paramètres de l’application, capteur, famille de FPGA... Dans ce manuscrit, les principaux algorithmes de détection de trait laser ainsi que leurs propriétés ont été étudiés. Un opérateur de détection de trait laser a été choisi et développé. L’implantation de cet opérateur sur une caméra-FPGA du marché a permis d’obtenir un premier prototype fonctionnel. Les performances temporelles de ce nouveau système sont quatre fois supérieures à celles du système actuellement utilisé par l’entreprise. Le nouveau système est capable de traiter jusqu’à 2500 images par seconde. Enfin, les modèles de la consommation des ressources permettent de dimensionner une architecture à partir d’un ensemble de paramètres prédéfinis de manière rapide et sans faire de synthèses. Le paramètre auquel les concepteurs doivent prêter le plus d’attention est le niveau de parallélisme des données. Ce paramètre permet d’exploiter les capacités de parallélisme du FPGA en consommant plus de ressources. Cependant, les ressources du FPGA sont limitées et augmenter le niveau de parallélisme peut induire la nécessité de changer de FPGA. Le système et les données fournies permettront à l’entreprise d’adapter le système de vision selon les besoins futurs des clients en les guidant vers le choix du matériel / This thesis is part of an industrial research training agreement (CIFRE) between the Hubert Curien laboratory and the company Pattyn Bakery Division. The goal of this work is the development of an FPGA laser line detection system that is more efficient than the current system of the company. In the industry, vision system designers need to be able to easily create and modify their systems in order to adapt them to their customers’ needs and technological developments. Thus developed operators must be generic to allow designers to change the vision system without necessarily having material skills. Designers must also be able to estimate what resources will be used by the operator in case of system changes : application parameters, sensor, family of FPGAs ... In this manuscript, the main laser line detection algorithms and their properties have been studied. A laser line detection operator was chosen and developed. The implementation of this operator on an FPGA-camera from market has resulted in a first functional prototype. The time performance of this new system is four times that of the system currently used by the company. The new system is able to process up to 2500 frames per second. Finally, resource consumption models makes it possible to size an architecture from a set of predefined parameters quickly and without synthesizing. The parameter to which designers must pay the most attention is the level of parallelism of the data. This parameter makes it possible to exploit the parallelism capabilities of the FPGA by consuming more resources. However, the resources of the FPGA are limited and increasing the level of parallelism can induce the need to change the family of FPGAs. The system and the data provided will enable the company to adapt the vision system to the future needs of customers by guiding the choice of equipment.