Axis network cameras contribute to Sensing Web Project, a research project on the next generation of sensor systems

Pioneer project aims to increase privacy protection in image recognition and processing of surveillance society

“Axis network cameras have detailed settings, and the API is open so program development was easy. Axis products are highly compatible with products from other companies, and were my preferred choice as a research developer.“ Ikuhisa Mitsugami, Researcher, Kyoto University.

Mission

Recently, security cameras and other sensors have been installed in various locations, and information obtained from these devices have begun to be used in many ways. For example, just like being able to see traffic congestion by viewing footage from cameras on the Internet, there are ways to utilize information from cameras and other sensors for uses beyond mere security. On the other hand, there is the problem of how to keep the privacy of those individuals filmed in consideration, since people’s faces and bodies can been seen from visual data coming straight from the camera. For this reason, use of information from sensors and security cameras is typically limited to the building/property management companies that installed the cameras. The Sensing Web Project was started for a more multifaceted utilization of information from cameras and other various sensors. The theme of this project is to realize the publication and use of data from sensors in a way that does not violate any individual’s privacy, by removing information regarding privacy in real-time as it is recorded. This was a large scale research project that was implemented with assistance from the Ministry of Education, Culture, Sports, Science and Technology of Japan over a 3 year period from April 2006 to March 2010, with the participation of Kyoto University, Wakayama University, Osaka University, the University of Tsukuba, Kyushu University, Toyohashi University of Technology, Nagoya University, and Kyoto Sangyo University.

Solution

Among the various sensors, visual data from cameras is most prone to privacy issues because it contains information about individuals’ figures. Technologies used for the removal of private information in visual data became an important topic of research in the Sensing Web Project. Making full use of technologies for image recognition/processing, private information contained in the visual data will be converted to numerical or symbolic data (such as whether or not people are present and the number if any are present) when recorded.

Result

Even if all the sensors in the world were made into “safe sensors”, this would be meaningless if there was no benefit to viewing/using the data obtained from sensors. In this project, various universities and research institutions propose and implement applications so that people will better understand these benefits. Some examples of applications are the “Festivity Map” that shows congestion in each area on a 2-dimensional map, the “Digital Diorama” that shows people, their positions, and their movements in a building in 3D computer graphics, and “See-through Vision” that shows supplementary visual data taken by a separate camera of what is beyond/beneath obstacles.

Related links

AXIS 243SA Video Server
AXIS 211 Network Camera
Education