Gheorghe Lazea

Adaptive appearance based loop-closing in heterogeneous environments

The work described in this paper concerns the problem of detecting loop-closure situations whenever an autonomous vehicle returns to previously visited places in the navigation area. An appearance-based perspective is considered by using images gathered by the on-board vision sensors for navigation tasks in heterogeneous environments characterized by the presence of buildings and urban furniture together with pedestrians and different types of vegetation. We propose a novel probabilistic on-line weight updating algorithm for the bag-of-words description of the gathered images which takes into account both prior knowledge derived from an off-line learning stage and the accuracy of the decisions taken by the algorithm along time. An intuitive measure of the ability of a certain word to contribute to the detection of a correct loop-closure is presented. The proposed strategy is extensively tested using well-known datasets obtained from challenging large-scale environments which emphasize the large improvement on its performance over previously reported works in the literature.

3D Laser scanning system and 3D segmentation of urban scenes

This paper dwells upon the promising 3D technology for mobile robots and automation industry. The first part of the paper describes the design details of our own 3D Time of Flight (TOF) scanning system based on 2D laser range finder. The second part presents a specific segmentation technique for 3D outdoor urban environments by the common detection of plane models. In a few words, the technique separates the raw data into sparse and dense points, followed by the segmentation of the dense points into urban background and foreground objects. In the end we present some experimental results of real-world data-sets taken from the repository1,2 of the Leibniz University in Hannover, Germany.

Graph Search Techniques for Mobile Robot Path Planning

In a number of applications, the problem of determining the optimum path occurs. This applications range from finding the fastest path in a network, to determining the safest path for mobile vehicle, wandering on the surface of Mars. In this context, we shall limit our scope to the case of finding paths in Euclidean two-dimensional space. Moreover we shall limit the case study to movements along a surface that can be projected onto a directed graph. To be specific, we shall look at the case of finding the optimum path for a mobile robot moving along a flat surface, the robot’s configurations in the configuration space being the graph’s nodes while the graph’s arcs represent the cost of moving from one configuration to another. Researchers have tried to come with new and better navigation technologies in the last years. With the development of path finding, several new classical routing algorithms have been introduced to generate better routing solution. For example the Dijkstra algorithm is the most famous one, which evaluates the moving cost from one node to any other node and sets the shortest moving cost as the connecting cost of two nodes (Eklund et al., 1996). Around the same period of time, Best-First-Search algorithm is also introduced in the researchers’ community. A little different from the Dijkstra algorithm, Best-First-Search algorithm has a different approach because it estimates the distance from current position to goal position, and it chooses the step that is closer to the goal position (LaValle, 2006). The difficulty was growing with the new path finding situations so the old path finding algorithm had to be improved to meet the new introduced requirements. A new path finding algorithm was introduced and it was named the A* algorithm. The A* algorithm tries to combine the advantages offered by Dijkstra algorithm and Best-FirstSearch algorithm. This paper presents tests performed with various implementations of graph search algorithms (A*, D*, focused D*) as path planners for a mobile robot, focused on strong points and drawbacks of each implementation. 8

Laser Based Localization Techniques for Indoor Mobile Robots

The localization problem in indoor environment based on LIDAR measurements is analyzed in this paper. Practical aspects of the localization are discussed including the implementations of the state estimation and registration algorithms. The localization framework developed is sufficient generic to be used in a variety of other autonomous vehicles. The results of the proposed navigation algorithms demonstrate a reliable and accurate position estimation for autonomous vehicles operating in a variety of environments.

Use Measurement System Analysis in Order to Reduce the System Variation

One source of variation in a process is generated by measurement. This paper is presenting a case study of Measurement System Analysis (MSA) in a wood industry process. It is necessary to do an analysis of the measurement system used in a company in order to be competitive and to gain the trust of the client. The aim of the paper is to show the steps done during a measurement system analysis and how to interpret the graphic results, but also the data generated by the Software. For this paper all the simulations were done in Minitab, but similar simulations can be done with other software programs.

Classification within indoor environments using 3D perception

Making sense out of human indoor environments is an essential feature for robots. In this paper we present a system for the classification of components inside these environments, starting from our robotic platform to a simple yet robust labeling process. Our method starts by acquiring multiple point clouds which are then registered into one single dataset. An estimation of principle axes is performed and the planar surfaces are segmented out. Further on, quadrilateral-like shapes are estimated for each detected plane, by making use of edges. And finally, since our classification approach relies on physical features, the method analyses the relationship between the previously mentioned shapes, as well as their physical sizes. To validate our approach, we tested the method on different datasets, which were recorded inside our office environment.

Mobile Robot Navigation Using Graph Search Techniques over an Approximate Cell Decomposition of the Free Space

Mobile robots often operate in domains that are incompletely known. This article addresses the goal-directed navigation problem in unknown terrain where a mobile robot has to move from its current configuration to given goal configuration. We will present tests performed with various implementations of graph search algorithms (A*, D*, focused D*) as path planners for a mobile robot, focusing on the inherent strong points and drawbacks of each implementation.

Economic optimization of a thermal cracker via model predictive control technology

An ethylene plant is one of the largest chemical plants and ethylene is industrially obtained through thermal cracking of hydrocarbons. The cracking furnace is the heart of such an installation and the frequent changes in feed mix, quality and prices; and the demand for its olefin products are influencing directly the production efficiency. Each reactant in the feed produces a certain distribution of products and in order to satisfy the demand constraints at the lowest cost; one needs to optimize the amounts of each reactant.

Visual odometer system to build feature based maps for mobile robot navigation

This paper presents a visual odometer system for mobile robot position correction. The developed algorithm detects the same Speeded Up Robust Features (SURF) on the stereo pair images to obtain three dimensional point clouds at every robot location. The algorithm tracks the displacement of the identical features viewed from different positions to compute the robots positions. The displacements between the point clouds are computed with the use of the Iterative Closest Point (ICP) algorithm. The ICP is used also to register the landmarks in the feature based map of the entire environment. The results of an indoor office environment experiments are shown.

An effective method for people detection in grayscale image sequences

This paper presents a method for detecting people from images taken with a camera mounted on a robot. The purpose of the detection is avoiding people collision while robot is moving within an unknown environment. It combines two algorithms for this purpose. First, the appearance of people is learned using a set of Haar-like features and the Adaboost algorithm. This information is embedded by building a classifier to differentiate people appearances by other structures. When an image is analyzed for detecting people, regions which contain vertical structures are determined using image gradients. Those regions which have a specific aspect-ratio are selected and the classifier is applied on them. The classifier marks the regions which contain people-like structures. Because this method is desired to be integrated in an autonomous robot navigation system for a dynamic environment, particular attention is paid to increase the speed of the detection as much as possible.