Graph-based representations and techniques for image processing and image analysis

Sanfeliu, Alberto; Alquézar, René; Andrade, Jasbon; Climent, Joan Mahiques; Serratosa, Francesc; Vergès, Josep-María

doi:10.1016/s0031-3203(01)00066-8

Cited by 81 publications

(39 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Results from this part of our research were presented at ICPR 2000 in Barcelona [8], and in a collective work on the use of graphs as topological data structures for computer vision, which appeared in the March 2002 issue of the journal Pattern Recognition [172].…”

Section: Dynamic Environmentsmentioning

confidence: 99%

Environment Learning for Indoor Mobile Robots

Andrade‐Cetto¹

2006

Springer Tracts in Advanced Robotics

View full text Add to dashboard Cite

This thesis focuses on the various aspects of autonomous environment learning for indoor service robots. Particularly, on landmark extraction from sensor data, autonomous map building, and robot localization.To univocally identify landmarks from sensor data, we study several landmark representations, and the mathematical foundation necessary to extract the features that build them from images and laser range data. The features extracted from just one sensor may not suffice in the invariant characterization of landmarks and objects, pushing for the combination of information from multiple sources. We present a new algorithm that fuses complementary information from two low level vision modules into coherent object models that can be tracked and learned in a mobile robotics context. Illumination conditions and occlusions are the most prominent artifacts that hinder data association in computer vision. By using photogrammetric and geometric constraints we restrict the search for landmark matches in successive images, and by locking our interest in one or a set of landmarks in the scene, we track those landmarks along successive frames, reducing considerably the data association problem. We concentrate on those tools from the geometry of multiple views that are relevant to the computation of initial landmark location estimates for coarse motion recovery; a desirable characteristic when odometry is not available or is highly unreliable.Once landmarks are accurately extracted and identified, the second part of the problem is to use these observations for the localization of the robot, as well as the refinement of the landmark location estimates. We consider robot motion and sensor observations as stochastic processes, and treat the problem from an estimation theoretic point of view, dealing with noise by using probabilistic methods.The main drawback we encounter is that current estimation techniques have been devised for static environments, and that they lack robustness in more realistic situations. To aid in those situations in which landmark observations might not be consistent in time, we propose a new set of temporal landmark quality functions, and show how iii by incorporating these functions in the data association tests, the overall estimationtheoretic approach to map building and localization is improved. The basic idea consists on using the history of data association mismatches for the computation of the likelihood of future data association, together with the spatial compatibility tests already available.Special attention is paid in that the removal of spurious landmarks from the map does not violate the basic convergence properties of the localization and map building algorithms already described in the literature; namely, asymptotic convergence and full correlation.The thesis also gives an in depth analysis of the fully correlated model to localization and map building from a control systems theory point of view. Considering the fact that the Kalman filter is nothing else but an optimal observer,...

show abstract

Section: Dynamic Environmentsmentioning

confidence: 99%

Environment Learning for Indoor Mobile Robots

Andrade‐Cetto¹

2006

Springer Tracts in Advanced Robotics

View full text Add to dashboard Cite

show abstract

“…Wong and You [13] proposed the First-Order Random Graphs (FORGS) with strong simplifications of the GRG, specifically they introduce three assumptions about the probabilistic independence between vertices and arcs which restrict too much the applicability of these graphs to object recognition. Later, our group introduced a new class of graphs called Function-Described Graphs (FDG) [14][15] to overcome some of the problems of the FORG. The FDG also considers some independence assumptions, but some useful 2 º order functions are included to constrain the generalisation of the structure.…”

Section: Learning and Recognising 3d Objects Represented By Multiple mentioning

confidence: 99%

Robot Vision for Autonomous Object Learning and Tracking

Sanfeliu

2003

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this paper we present a summary of some of the research that we are developing in the Institute of Robotics of the CSIC-UPC, in the field of Learning and Robot Vision for autonomous mobile robots. We describe the problems that we have found and some solutions that have been applied in two issues: tracking objects and learning and recognition of 3D objects in robotic environments. We will explain some of the results accomplished.

show abstract

“…In this paper we present some results of the use of graph techniques [16] for the process of identification of objects and scenes in indoor environments for mobile robots.…”

Section: Introductionmentioning

confidence: 99%

The Use of Graph Techniques for Identifying Objects and Scenes in Indoor Building Environments for Mobile Robots

Sanfeliu

2004

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this paper we present some of the common issues that appear when we try to recognize objects in indoor scenes of a building, and we describe some strategies for recognizing them by using graph techniques. These scene images are captured by the colour cameras of a mobile robot, which in the learning phase, learn the objects by taken a set of 2D images of the projective object views. Then afterwards, the robot must identify the objects once its moves through the area that has been used to learn the objects. We describe two strategies to use graph techniques for object and scene recognition, some algorithms and preliminary results.

show abstract

Graph-based representations and techniques for image processing and image analysis

Cited by 81 publications

References 17 publications

Environment Learning for Indoor Mobile Robots

Environment Learning for Indoor Mobile Robots

Robot Vision for Autonomous Object Learning and Tracking

The Use of Graph Techniques for Identifying Objects and Scenes in Indoor Building Environments for Mobile Robots

Contact Info

Product

Resources

About