Mexico. The Center for Research in Mathematics -CIMAT- of Guanajuato, is developing a facial detection system with application to videoconferencing systems, which focuses on separating relevant information from the foreground, to transmit it with higher quality.The algorithm developed by CIMAT can work with any background, skin color, lighting levels or partially covered faces. Protection codes are also being developed that manage to recover damaged information during the communication process without the need to retransmit it, saving network resources. In this way, the images of faces are sent with excellent quality and protected against the loss of information, while the rest of the image is sent with medium and low quality.
Rogelio Hasimoto Beltrán, leader of the research, explained that this is the first stage of the investigation. The ultimate goal is to create automatic systems that can focus on the central part and protect it against transmission errors, for example in a football match, the main scene would be around the ball and who carries it, therefore the system would protect that region from errors.
Leave your comment