To prevent the spread of COVID-19 infection, NTT has developed a system that can visualize locations touched by people. This technology uses a thermal camera to capture images of the heat traces left after a human hand touches any surface (like a wall, door, or desk in a public space). Using the developed technology, the locations that were touched can be detected with high speed and accuracy, and the detection results can be projected on the actual surface. Using this technology, it is possible to intuitively identify the locations touched by a person, and as a result, avoid the location, and disinfect it to reduce the risk of infection.
Various measures have been adopted to prevent the spread of COVID-19. In response to the risks of virus particles that attach to the surface of objects that are touched by people on a daily basis, such as shelves, doors, and desks, regular disinfection is effective. However, regular disinfection does not eliminate the anxiety associated with another person touching the surface after disinfection or with the effectiveness of disinfection of the location. The conventional method of visualizing traces using chemicals and black light has limitations because of the use of chemicals. In addition, methods for identifying the touched locations using a general camera (like those used in virtual keyboards) or using a depth sensor are conceivable. However, it is difficult to reliably determine the location touched by a person.
NTT has developed a method to visualize the locations touched by an individual by detecting heat traces. Generally, the temperature of human hands is higher than that of the surroundings, so when human hands touch any surface in the surrounding environment, the heat from their touch remains on the surface. The heat remaining on the surface is called a heat trace, and this trace can be captured in the form of an image using a thermal camera (that employs far-infrared rays). Using this, the technology detects heat traces to identify locations that have been touched by people. This heat trace does not alter other wavelengths such as visible light and near-infrared light and cannot be captured using ordinary cameras.
The developed technology uses a thermal camera, along with a visible-light or near-infrared camera. With the background subtraction algorithm, only the thermographic image remains changed, and the other cameras detect the portion of the heat trace that does not change when a location is touched by a person. Using a projector-aided projection display, the location touched by the detected person is superimposed on the surface of the actual object and presented. Since the projection is performed in the visible-light region, identification is possible using a lightweight algorithm by photographing the movement of a person in the near-infrared region and separating the wavelength of the light.
In summary, the researchers succeeded in reducing the weight of the algorithm by using two types of cameras with different wavelengths that could capture photographs in combination. This algorithm can be implemented in a small sensor node that can be installed in various locations. The installation of numerous sensor nodes equipped with a thermal camera will enable wide application for recording the location touched by a person and for tracking the location touched by a feverish person. In the future, NTT plans to continue the study of this technology to expand its practical applicability by considering the differences among various material surfaces and demonstrating them in the actual environment.
This article has been translated by JST with permission from The Science News Ltd.(https://sci-news.co.jp/).
Unauthorized reproduction of the article and photographs is prohibited.