The project is being developed by the Institute for Systems and Computer Engineering, Technology and Science (INESC TEC), represented by its Centre for Information Systems and Computer Graphics (CSIG). A total of five researchers are working on this project, with the collaboration of the University of Trás-os-Montes and the University of Warnick (UK). The project has a duration of 24 months.

According to Maximino Bessa, researcher at CSIG, monitoring these objects by analysing images using computer vision techniques is key to some of the areas of expertise of the Office of Naval Research of the United States (ONRG). “Obtaining accurate and reliable information through image analysis is very important, particularly in the area of ​​autonomous and unmanned systems”, the researcher added.

These systems need to be operated “in a variety of lighting conditions” and there may be objects in a “very dark” (shadow) area, while others may be in a “very bright” location (under the sun).

According to the researcher, currently, in this type of scenarios, even the most recent algorithms are unable to perform object tracking because they cannot extract information from the collected data. This happens because the images may be under or overexposed.

With this project, the goal is to use High Dynamic Range (HDR) techniques in order to capture and represent the whole scenario. For this, a system will be created that will make it possible to track multiple objects in real time.

According to Maximino Bessa, this is a pioneer project and there are currently no studies that systematically evaluate the new HDR video formats in this dimension.

The project is funded by the Office of Naval Research of the United States of America (ONR), an integral part of the US Navy Department which coordinates, implements and promotes the science and technology programs of the US Navy and Marine Corps.

Notícias ao Minuto, 22 December 2016