Homepage of Renaud PETERI

Dynamic Textures

Dynamic, or temporal, texture is a spatially repetitive, time-varying visual pattern that forms an image sequence with certain temporal stationarity. In dynamic texture (DT), the notion of self-similarity central to conventional image texture is extended to the spatiotemporal domain. DTs are typically videos of processes, such as waves, smoke,fire, a flag blowing in the wind, a moving escalator, or a walking crowd.
We are interested in detecting, segmenting and analyzing all kinds of DT occuring in videos sequences.

Collaborators: Michel Ménard, Sloven Dubois (University of La Rochelle, France), Sandor Fazekas, Dmitry Chetverikov (Sztaki, Hungarian Academy of Sciences), Mark Huiskes (University of Leiden, The Netherlands)

Decomposition and analysis

Our recent work were focused on dynamic texture analysis and characterization. Many dynamic textures can be modeled as large scale propagating wavefronts and local oscillating phenomena. After introducing a formal model for dynamic textures, the morphological component analysis (MCA) approach with a well-chosen dictionary is used to retrieve the components of dynamic textures.We define two new strategies for adaptive thresholding in the MCA framework, which greatly reduce the computation time when applied on videos.

Most relevant reference:
Decomposition of Dynamic Textures using Morphological Component Analysis.
Sloven Dubois, Renaud Péteri and Michel Ménard
IEEE Transactions on Circuits and Systems for Video Technology, vol.22, no.2, pp.188-201, Feb. 2012.
PDF
Demo page: here

Indexing

In image processing, the wavelet transform has been successfully used for characterizing static textures. To our best knowledge, only two works are using spatio-temporal multiscale decomposition based on tensor product for dynamic texture recognition. One contribution we made is to analyse and compare the ability of the 2D+T curvelet transform, a geometric multiscale decomposition, for characterizing dynamic textures in image sequences. Two approaches using the 2D+T curvelet transform have been compared using three new large databases.

Most relevant reference:
Characterization and Recognition of Dynamic Textures based on 2D+T Curvelet Transform.
Sloven Dubois, Renaud Péteri and Michel Ménard
Signal, Image and Video Processing, Springer-Verlag, in press (July 2013).
PDF

Spatio-temporal segmentation

We have developped a new approach for segmenting a video sequence containing dynamic textures. The proposed method is based on a 2D+T curvelet transform and an octree hierarchical representation. The curvelet transform enables to outline spatio-temporal structures of a given scale and orientation. The octree structure based on motion coherence enables a better spatio-temporal segmentation than a direct application of the 2D+T curvelet transform.

Most relevant reference:
A 3D discrete curvelet based method for segmenting dynamic textures.
Sloven Dubois, Renaud Péteri and Michel Ménard
ICIP 2009, the 2009 IEEE International Conference on Image Processing , Cairo, Egypt, November 7-11, 2009.
PDF

DynTex: A Comprehensive Database of Dynamic Textures

We present the DynTex database of high-quality dynamic texture videos. It consists of over 650 sequences of dynamic textures, mostly in everyday surroundings. Additionally, we propose a scheme for the manual annotation of the sequences based on a detailed analysis of the physical processes underlying the dynamic textures. Using this scheme we describe the texture sequences in terms of both visual structure and semantic content. The videos and annotations are made publicly available for scientific research.

Most relevant reference:
DynTex: A Comprehensive Database of Dynamic Textures.
Renaud Péteri, Sándor Fazekas and Mark J. Huiskes.
Pattern Recognition Letters, Volume 31, Issue 12, 1 September 2010, Pages 1627-1632.
PDF
DynTex website: here

Remote sensing

Collaborators: Thierry Ranchin, (Mines ParisTech, Sophia-Antipolis), Isabelle Couloigner (University of Calgary)

Road extraction from high spatial resolution images

There is a strong demand for accurate and up-to-date road network information. Road network knowledge is crucial for the creation and the update of maps, geographic information system (GIS) database, transportation or land planning. We have designed a new method for extracting road networks from high spatial resolution images is then described. It models roads as a surface and is built on cooperation between linear and surface representation of roads. In order to overcome local artifacts, the method makes use of advanced image processing tools, such as active contours and the wavelet transform. The method is applied on a high resolution images from the Quickbird or Ikonos satellites. Results have been quantitatively assessed compared to human interpretation.

Most relevant reference:
Road networks derived from high spatial resolution satellite remote sensing data.
Renaud Péteri and Thierry Ranchin.
Chapter 11 of Remote Sensing of Impervious Surfaces, pp. 215-236, editor: Q. Weng. CRC Press, Taylor & Francis Group, 2007. ISBN: 1420043749.
PDF

Computer Vision and Statistical learning

Collaborators: Nuno Vasconcelos (University of California, San Diego), Laurent Mascarilla, Cyrille Beaudry (MIA lab., University of La Rochelle)
More to come soon...

Other works

Tracking, bio-imaging,...
Collaborators: Eric Rosenfeld ,

More to come soon...