Alpage est un programme de recherche, initié en 2006, grâce au soutien de l'agence nationale de la recherche.
ANR
Ce programme est fondé sur l’active collaboration entre 4 équipes de recherches (3 SHS et 1 STIC), qui ont signé un accord de consortium en 2009 :


  • Performance Evaluation : Snapshots

  • A Prototype Based Reduction Scheme for Structural Data Classification.
  • [PDF Presentation]

    This work was carried out in association with the LITIS lab at the university of Rouen. especially with Sebastien Adam and Pierre Héroux.

  • A graph matching method based on probes assignments
  • This software aims at matching two attributed graphs. It runs fast O(n^3) and the univalent mapping can be expresssed as the minimum-weight probe matching between G1 and G2 Download and More information

    Image and Ontology

    Here are material I gave to Master degree students. This work is dealing with Image and Ontology. More precisely, how to classify regions of an image using an ontology reasoner.

  • Vectorial image segmentation and a graph based merging process.
  • This software performs two colour image segmentations. In addition, as a final stage an image vectorization algorithm is carried out. Concepts: Color gradient, statistical merging, neighborhood graph merging, raster image vectorization Download and More information

  • A Content Based Image Retrieval Method Using A Graph Representation.
  • Here, we propose an automatic system to annotating and retrieving images. We assume that regions in an image can be described using a small vocabulary of blobs. Blobs are generated from image features using clustering. Each image sees its blobs structured into a graph, a blob adjacency graph. This representation is used to perform a similarity search into an image set. Hence, the user can express his need by giving a query image, and thereafter receiving as a result all similar images. Download and More information

  • Best Colour Space Finder:
  • This software computes on a RGB image different feature selection methods in order to find the best colour space in term of data separability. Download and More information