Pontifical Catholic University of Rio de Janeiro, Computer Vision Laboratory (LVC)

LVC logoPontifical Catholic University of Rio de Janeiro, Computer Vision Laboratory (LVC)

The Computer Vision Laboratory or Laboratório de Visão Computacional – LVC is part of the research infrastructure of the Department of Electrical Engineering of the Pontifical Catholic University of Rio de Janeiro (PUC-Rio). The primary research focus at LVC/PUC-Rio is the emulation in a computational environment of the humans’ ability of recognizing structures and objects in images. Although the research developed at LVC/PUC-Rio also encompasses medical image analysis and biometrics, the majority of the research currently under development at the institute is related to the field of automatic interpretation of remotely sensed (RS) images. LVC/PUC-Rio also takes part of the Brazilian Visualization Network (Rede Brasileira de Visualização – RBV), which is supported by the Brazilian Ministry of Science and Technology (MCT) and by FINEP (Brazilian Innovation Agency). RBV is an intelligence network composed by some of Brazil’s most important universities and research centers, and aims at developing the visualization technological macrosector in the country. LVC/PUC-Rio is the current leader of the GeoInformation sub area of RBV. LVC/PUC-Rio has a vast experience in interacting with Brazilian and international research institutes in the development of basic and applied research related to automated RS image interpretation. Some recent examples are:

  • PIMAR – Integrated Remote Monitoring of Forest Fragments and Urban Growth in Rio de Janeiro. The Project aims at developing a semi-automated methodology for the detection of vegetation suppression, and of horizontal and vertical growth of the settlements adjacent to forested and protected areas in Rio de Janeiro, using multidate, stereoscopic images from the IKONOS2 orbital sensor system. The methodology – currently under evaluation – is based on the InterIMAGE open-source software package. Financed by Government of the State of Rio de Janeiro, the Project’s scientific staff also includes members from the Brazilian Space Research Institute (INPE) and from Rio de Janeiro State University (UERJ).

  • M4 – Multi-Scale, Multi-Temporal and Multi-Dimensional Monitoring of Brazilian Settlements and Forests by Means of Remote Sensing Data. The project aimed at developing automated methods to aid monitoring of changes in the landscape, and for the inference of population growth in informal settlements, using high and medium resolution remotely sensed images. Partners: IPI and TNT Institutes of the Leibniz University of Hannover; Brazilian Space Research Institute (INPE).

  • ECOWATCH – Development and investigation of knowledge-based methods for the automatic interpretation of multi-sensor remote sensing data. Partners: IPI and TNT Institutes of the Leibniz University of Hannover; Rio de Janeiro State University (UERJ); Brazilian Agriculture Research Corporation (EMBRAPA).

  • ECOAIR – Development of models and automated routines to identify from optical (medium and low resolution) satellite images, environmental parameters associated to erosive processes and environmental degradation. Partners: Institut National de Recherche en Informatique et en Automatique (INRIA); Rio de Janeiro State University (UERJ); Brazilian Agriculture Research Corporation (EMBRAPA).

Last but not least, the LVC/PUC-Rio leads the development of the InterIMAGE software platform, an open-source, knowledge-based system for the interpretation of multi-source, multi-scale, multi-temporal remote sensing data. The system is based on the GeoAIDA system, originally developed at the TNT institute from University of Hannover, and on the TerraLib software library, maintained by INPE.