In recent years, the market for automatized surveillance and use of unmanned ground vehicles (UGVs) has increased considerably. In order for unmanned vehicles to operate autonomously, high level algorithms of artificial intelligence need to be developed and accompanied by some way to make the robots perceive and interpret the environment. The purpose of this work is to investigate methods for real-time obstacle detection using stereo vision and implement these on an existing UGV platform. To reach real-time processing speeds, the algorithms presented in this work are designed for parallel processing architectures and implemented using programmable graphics hardware. The reader will be introduced to the basics of stereo vision and given an overview of the most common real-time stereo algorithms in literature along with possible applications. A novel wide-baseline real-time depth estimation algorithm is presented. The depth estimation is used together with a simple obstacle detection algorithm, producing an occupancy map of the environment allowing for evasion of obstacles and path planning. In addition, a complete system design for autonomous navigation in multi-UGV systems is proposed.

Today, there are industries that utilize AGVs to transport goods and materials from one location to another. For smaller scale industries it is costly to have a custom made AGV for their manufacturing unit, so they modify the shape of an AGV to accommodate the necessity of carrying and transporting goods. When the shape of an AGV is modified the built-in sensors will not detect the change in shape of the AGV. Hence, there is a risk that the AGV may collide with objects. Also in some AGVs floor sensors are missing to detect the presence of floor/no floor in front of the AGV, which can be a hazardous situation as there are chances of the AGV falling off from the surface. An example of such an AGV is the Patrolbot which can travel around in an industrial premise wirelessly, but needs addition of such sensors to avoid collisions with the modified structure. A Patrolbot has been used in this thesis work and ultrasonic sensors are utilised for obstacle detection with a modified structure and a built-in laser scanner is studied for mapping purpose. The results of this master thesis was that the ultrasonic sensors were tested under various conditions and results were derived. To obtain the same level of results every time it is required to maintain the conditions on which the ultrasonic sensors rely.

A drone is an unmanned aerial vehicle with multiple forms of usage. Drones can be programmed to fly with different degrees of autonomous flight. Autonomous controlled flight makes it possible for the drone to fly without human involvement and it is then controlled solely by software. The goal of this project is to program the micro drone Crazyflie 2.1 to autonomously fly through an obstacle course in the shortest amount of time and a predetermined direction. The nature and placement of the obstacles are unknown beforehand. The obstacles are detected and avoided by using the obstacle detection sensor Multi- ranger. To achieve autonomous flight two possible navigation systems were tested, the Loco Positioning System and Flow deck. Flying the Crazyflie while using Flow deck as positioning system performed best, managing to fly through the obstacle course avoiding all obstacles.

This diploma thesis deals with complete designing and implementation of local navigation of a robot which travels according to set GPS. This work contains a study of contemporary used principles of the sensors which can be used as the components of the robot. There are sensors for measuring the surroundings of the robot and orientation of the robot in the environs. The equipment includes an instructing control unit with the ATmega8 microcontroller. This equipment collects data from the sensors and transfers them to the other systems of the robot. The following part of the thesis deals with the draft of the buses and communication protocols, which are necessary for connection between the sensors and the control unit. The I2C and RS232 buses were chosen as the most suitable. The communication protocol was borrowed from the semester project written by Bc. Michal Sitta. The main task of this control unit is reading out the data which have been measured from connected sensors. The realization of interface between the sensor bus and the major system is another core goal of the control unit. The system for a local navigation is designed with the approach which allows the widest range of extensibility. During the designing of the control unit the future possibility of insertion other sensors was taken into consideration.

La détection, localisation d'obstacles et la reconstruction de carte d'occupation 2D sont des fonctions de base pour un robot navigant dans un environnement intérieure lorsque l'intervention avec les objets se fait dans un environnement encombré. Les solutions fondées sur la vision artificielle et couramment utilisées comme SLAM (simultaneous localization and mapping) ou le flux optique ont tendance a être des calculs intensifs. Ces solutions nécessitent des ressources de calcul puissantes pour répondre à faible vitesse en temps réel aux contraintes. Nous présentons une architecture matérielle pour la détection, localisation d'obstacles et la reconstruction de cartes d'occupation 2D en temps réel. Le système proposé est réalisé en utilisant une architecture de vision sur FPGA (field programmable gates array) et des capteurs d'odométrie pour la détection, localisation des obstacles et la cartographie. De la fusion de ces deux sources d'information complémentaires résulte un modèle amelioré de l'environnement autour des robots. L'architecture proposé est un système à faible coût avec un temps de calcul réduit, un débit d'images élevé, et une faible consommation d'énergie / Obstacle detection, localization and occupancy map reconstruction are essential abilities for a mobile robot to navigate in an environment. Solutions based on passive monocular vision such as simultaneous localization and mapping (SLAM) or optical flow (OF) require intensive computation. Systems based on these methods often rely on over-sized computation resources to meet real-time constraints. Inverse perspective mapping allows for obstacles detection at a low computational cost under the hypothesis of a flat ground observed during motion. It is thus possible to build an occupancy grid map by integrating obstacle detection over the course of the sensor. In this work we propose hardware/software system for obstacle detection, localization and 2D occupancy map reconstruction in real-time. The proposed system uses a FPGA-based design for vision and proprioceptive sensors for localization. Fusing this information allows for the construction of a simple environment model of the sensor surrounding. The resulting architecture is a low-cost, low-latency, high-throughput and low-power system.

Self-driving cars have recently become a challenging research topic, with the aim of making transportation safer and more efficient. Current advanced driving assistance systems (ADAS) allow cars to drive autonomously by following lane markings, identifying road signs and detecting pedestrians and other vehicles. In this thesis work we improve the robustness of autonomous cars by designing an on-road obstacle detection system. The proposed solution consists on the low-level fusion of radar and lidar through the occupancy grid framework. Two inference theories are implemented and evaluated: Bayesian probability theory and Dempster-Shafer theory of evidence. Obstacle detection is performed through image processing of the occupancy grid. Last, the Dempster-Shafer additional features are leveraged by proposing a sensor performance estimation module and performing advanced conflict management. The work has been carried out at Volvo Car Corporation, where real experiments on a test vehicle have been performed under different environmental conditions and types of objects. The system has been evaluated according to the quality of the resulting occupancy grids, detection rate as well as information content in terms of entropy. The results show a significant improvement of the detection rate over single-sensor approaches. Furthermore, the Dempster-Shafer implementation may slightly outperform the Bayesian one when there is conflicting information, although the high computational cost limits its practical application. Last, we demonstrate that the proposed solution is easily scalable to include additional sensors.

No description available.

Scene perception and traversability analysis are real challenges for autonomous driving systems. In the context of off-road autonomy, there are additional challenges due to the unstructured environments and the existence of various vegetation types. It is necessary for the Autonomous Ground Vehicles (AGVs) to be able to identify obstacles and load-bearing surfaces in the terrain to ensure a safe navigation (McDaniel et al. 2012). The presence of vegetation in off-road autonomy applications presents unique challenges for scene understanding: 1) understory vegetation makes it difficult to detect obstacles or to identify load-bearing surfaces; and 2) trees are usually regarded as obstacles even though only trunks of the trees pose collision risk in navigation. The overarching goal of this dissertation was to study traversability analysis in unstructured forested terrains for off-road autonomy using LIDAR data. More specifically, to address the aforementioned challenges, this dissertation studied the impacts of the understory vegetation density on the solid obstacle detection performance of the off-road autonomous systems. By leveraging a physics-based autonomous driving simulator, a classification-based machine learning framework was proposed for obstacle detection based on point cloud data captured by LIDAR. Features were extracted based on a cumulative approach meaning that information related to each feature was updated at each timeframe when new data was collected by LIDAR. It was concluded that the increase in the density of understory vegetation adversely affected the classification performance in correctly detecting solid obstacles. Additionally, a regression-based framework was proposed for estimating the understory vegetation density for safe path planning purposes according to which the traversabilty risk level was regarded as a function of estimated density. Thus, the denser the predicted density of an area, the higher the risk of collision if the AGV traversed through that area. Finally, for the trees in the terrain, the dissertation investigated statistical features that can be used in machine learning algorithms to differentiate trees from solid obstacles in the context of forested off-road scenes. Using the proposed extracted features, the classification algorithm was able to generate high precision results for differentiating trees from solid obstacles. Such differentiation can result in more optimized path planning in off-road applications.

No description available.

A system is described for identifying obstacle-free landing sites for a vertical-takeoff-and-landing (VTOL) semi-autonomous unmanned aerial vehicle (UAV) from point cloud data obtained from a stereo vision system. The relatively inexpensive, commercially available Bumblebee stereo vision camera was selected for this study. A "point cloud viewer" computer program was written to analyze point cloud data obtained from 2D images transmitted from the UAV to a remote ground station. The program divides the point cloud data into segments, identifies the best-fit plane through the data for each segment, and performs an independent analysis on each segment to assess the feasibility of landing in that area. The program also rapidly presents the stereo vision information and analysis to the remote mission supervisor who can make quick, reliable decisions about where to safely land the UAV. The features of the program and the methods used to identify suitable landing sites are presented in this thesis. Also presented are the results of a user study that compares the abilities of humans and computer-supported point cloud analysis in certain aspects of landing site assessment. The study demonstrates that the computer-supported evaluation of potential landing sites provides an immense benefit to the UAV supervisor. / Master of Science