Thierry Géraud

Ruminations on Tarjan’s Union-Find algorithm and connected operators

By Thierry Géraud

2005-01-05

In Proceedings of the 7th international symposium on mathematical morphology (ISMM’05)

Abstract

This papers presents a comprehensive and general form of the Tarjan’s union-find algorithm dedicated to connected operators. An interesting feature of this form is to introduce the notion of separated domains. The properties of this form and its flexibility are discussed and highlighted with examples. In particular, we give clues to handle correctly the constraint of domain-disjointness preservation and, as a consequence, we show how we can rely on “union-find” to obtain algorithms for self-dual filters approaches and levelings with a marker function.

Continue reading

Fast road network extraction in satellite images using mathematical morphology and Markov random fields

By Thierry Géraud, Jean-Baptiste Mouret

2004-09-05

In EURASIP Journal on Applied Signal Processing

Abstract

This paper presents a fast method for road network extraction in satellite images. It can be seen as a transposition of the segmentation scheme “watershed transform + region adjacency graph + Markov random fields” to the extraction of curvilinear objects. Many road extractors can be found in the literature which are composed of two stages. The first one acts like a filter that can decide from a local analysis, at every image point, if there is a road or not. The second stage aims at obtaining the road network structure. In the method we propose, we rely on a “potential” image, that is, unstructured image data that can be derived from any road extractor filter. In such a potential image, the value assigned to a point is a measure of its likelihood to be located in the middle of a road. A filtering step applied on the potential image relies on the area closing operator followed by the watershed transform to obtain a connected line which encloses the road network. Then a graph describing adjacency relationships between watershed lines is built. Defining Markov random fields upon this graph, associated with an energetic model of road networks, leads to the expression of road network extraction as a global energy minimization problem. This method can easily be adapted to other image processing fields where the recognition of curvilinear structures is involved.

Continue reading

Fast color image segmentation based on levellings in feature space

By Thierry Géraud, Giovanni Palma, Niels Van Vliet

2004-08-11

In Computer vision and graphics—international conference on computer vision and graphics (ICCVG), warsaw, poland, september 2004

Abstract

This paper presents a morphological classifier with application to color image segmentation. The basic idea of a morphological classifier is to consider that a color histogram is a 3D gray-level image and that morphological operators can be applied to modify this image. The final objective is to extract clusters in color space, that is, identify regions in the 3D image. In this paper, we particularly focus on a powerful class of morphology-based filters called levellings to transform the 3D histogram-image to identify clusters. We also show that our method gives better results than the ones of state-of-the-art methods.

Continue reading

Fusion of spatial relationships for guiding recognition, example of brain structure recognition in 3D MRI

By Isabelle Bloch, Olivier Colliot, Oscar Camara, Thierry Géraud

2004-07-09

In Pattern Recognition Letters

Abstract

Spatial relations play an important role in recognition of structures embedded in a complex environment and for reasoning under imprecision. Several types of relationships can be modeled in a unified way using fuzzy mathematical morphology. Their combination benefits from the powerful framework of fuzzy set theory for fusion tasks and decision making. This paper presents several methods of fusion of information about spatial relationships and illustrates them on the example of model-based recognition of brain structures in 3D magnetic resonance imaging.

Continue reading

Generic algorithmic blocks dedicated to image processing

By Jérôme Darbon, Thierry Géraud, Patrick Bellot

2004-03-10

In Proceedings of the ECOOP workshop for PhD students

Abstract

This paper deals with the implementation of algorithms in the specific domain of image processing. Although many image processing libraries are available, they generally lack genericity and flexibility. Many image processing algorithms can be expressed as compositions of elementary algorithmic operations referred to as blocks. Implementing these compositions is achieved using generic programming. Our solution is compared to previous ones and we demonstrate it on a class image processing algorithms.

Continue reading

A static C++ object-oriented programming (SCOOP) paradigmc mixing benefits of traditional OOP and generic programming

By Nicolas Burrus, Alexandre Duret-Lutz, Thierry Géraud, David Lesage, Raphaël Poss

2003-10-29

In Proceedings of the workshop on multiple paradigm with object-oriented languages (MPOOL)

Abstract

Object-oriented and generic programming are both supported in C++. OOP provides high expressiveness whereas GP leads to more efficient programs by avoiding dynamic typing. This paper presents SCOOP, a new paradigm which enables both classical OO design and high performance in C++ by mixing OOP and GP. We show how classical and advanced OO features such as virtual methods, multiple inheritance, argument covariance, virtual types and multimethods can be implemented in a fully statically typed model, hence without run-time overhead.

Continue reading

Representation and fusion of heterogeneous fuzzy information in the 3D space for model-based structural recognition—application to 3D brain imaging

By Isabelle Bloch, Thierry Géraud, Henri Maître

2003-08-01

In Artificial Intelligence

Abstract

We present a novel approach of model-based pattern recognition where structural information and spatial relationships have a most important role. It is illustrated in the domain of 3D brain structure recognition using an anatomical atlas. Our approach performs simultaneously segmentation and recognition of the scene and the solution of the recognition task is progressive, processing successively different objects, using different of knowledge about the object and about relationships between objects. Therefore the core of the approach is the representation part, and constitutes the main contribution of this paper. We make use of a spatial representation of each piece of information, as a spatial set representing a constraint to be satisfied by the searched object, thanks in particular to fuzzy mathematical operations. Fusion of these constraints allows to, segment and recognize the desired object.

Continue reading

Segmentation d’objets curvilignes à l’aide des champs de markov sur un graphe d’adjacence de courbes issu de l’algorithme de la ligne de partage des eaux

By Thierry Géraud

2003-06-01

In Proceedings of the international conference on image and signal processing (ICISP)

Abstract

This paper presents a general framework to segment curvilinear objects in 2D images. A pre-processing step relies on mathematical morphology to obtain a connected line which encloses curvilinear objects. Then, a graph is constructed from this line and a Markovian Random Field is defined to perform objects segmentation. Applications of our framework are numerous: they go from simple surve segmentation to complex road network extraction in satellite images.

Continue reading

Document type recognition using evidence theory

By Thierry Géraud, Geoffroy Fouquier, Quoc Peyrot, Nicolas Lucas, Franck Signorile

2003-04-29

In Proceedings of the 5th IAPR international workshop on graphics recognition (GREC)

Abstract

This paper presents a method to recognize the type of a document when a database of models (document types) is given. For instance, when every documents are forms and when we know every different types of forms, we want to be able to assign to an input document its type of form. To that aim, we define each model by a set of characteristics whose nature can vary from one to another. For instance, a characteristic can be having a flower-shaped logo on top-left as well as having about 12pt fonts. This paper does not intent to explain how to extract such knowledge from documents but it describes how to use such information to decide what the type of a given document is when different document types are described by characteristics.

Continue reading

Multi-band segmentation using morphological clustering and fusion: Application to color image segmentation

By Heru Xue, Thierry Géraud, Alexandre Duret-Lutz

2003-04-10

In Proceedings of the IEEE international conference on image processing (ICIP)

Abstract

In this paper we propose a novel approach for color image segmentation. Our approach is based on segmentation of subsets of bands using mathematical morphology followed by the fusion of the resulting segmentation channels. For color images the band subsets are chosen as RG, RB and GB pairs, whose 2D histograms are processed as projections of a 3D histogram. The segmentations in 2D color spaces are obtained using the watershed algorithm. These 2D segmentations are then combined to obtain a final result using a region split-and-merge process. The CIE L a b color space is used to measure the color distance. Our approach results in improved performance and can be generalized for multi-band segmentation of images such as multi-spectral satellite images information.

Continue reading

Search

Tags