The annotation of images data is the main tool for associating semantics to an image. In this article we are interested in the semi-automatic annotation of images data. Indeed, with the great mass of data managed throughout the world and especially with the Web, the manual annotation of these images is almost impossible. This article presents a preliminary work on an neighbourhood graphs based ...

In recent years, image-processing solutions have been to automatically detect incidents and make measurements on video images from CCTV cameras, relieving the staff in control rooms of much of the difficulty in finding out where events that require attention are occurring. In this paper we illustrate some of the results obtained with on-site experiments carried out as part of projects funded by...

Multimedia databases are witnessing an exponential growth. Information retrieval in these data collections, especially in images, becomes a hard task. Furthermore, there is a difficulty in indexing these data since the present techniques don’t allow an efficient description of the content of images. In this paper, we focus on knowledge discovery in images database and we propose a new semantic ...

This paper presents an approach to grouping natural scenes into (semantically) meaningful categories. The proposed approach exploits the statistics of natural scenes to define relevant image categories. A ridgelet-based signature is used to represent images. This signature is used by a support vector classifier that is well designed to support high dimensional features, resulting in an effectiv...

Geolocalized databases are becoming necessary in a wide variety of application domains. The manual creation of such databases is an expensive operation which stimulated the interest for the automation of their construction, by mining geographic information from the Web. In this article, we present and evaluate a new automated approach for creating a geographical database. Our technique uses Wik...

We propose in this paper a document image classification method. In contrary to most of existing systems, the proposed approach allow locating the document and recognizing its type simultaneously. First, a knowledge base of document models is created from reference images. Training images are not indispensable and though only one reference image is enough to create a document model. Then, key-p...

Tables are complex elements that can disturb the automatic analysis of the structure of an image of a document. In this article, we present a method based on the alternation of the color of lines to extract color tables that are not materialized by physical rulings. Experimental results, obtained on a dataset of document images with various layouts, enable to validate the interest of this appro...

In this paper, we present a method for the annotation of faces in a large real corpus. We use a face detector and a named entities detector together: pictures containing faces can be annotated by the most frequent named entities found in the text surrounding the pictures. Although this method is quite simple, it yields good results. This is a first step towards more intelligent image annotation...

This paper presents a novel approach, the first to our knowledge, that exploits a complete extension of the language modeling approach from information retrieval to the problem of graph-based image retrieval and categorization. Since photographic images are 2D data, we first use image regions and local interest points (mapped to automatically induced concepts) and then relationships between the...

This paper presents a generalization of La argue's algorithm for simulating rational expectation models with several leads and lags and its implementation in a Gauss program, DYNARE. Mots clés: anticipations rationnelles, modèles non linéaires, simulations.