About 90% of vehicle accidents can be attributed to human error such as carelessness or an error in judgment. Fujitsu has developed an in-vehicle 3D image synthesis technology that displays minor collision risks in various colors on in-vehicle monitor so the driver can intuitively identify dangers that may result in an accident. This technology supports driver safety when there is a high risk of a minor collision, for example when parking or when passing a car at night.
Display risk of minor collisions with 3D structures using a combination of an in-vehicle camera and wide-angle laser radar
Automotive technology that supports safer driving in society is advancing rapidly. Features like collision safety brakes, monitors, and lane departure warnings are already in practical use. By around 2020, motor vehicles that can run automatically by simply selecting the destination are expected to appear on the market.
Through the recently developed, world first, in-vehicle 3D image synthesis technology, Fujitsu is pushing this rapid evolution of automotive ICT. The technology displays images of 3D objects including people and objects without distortion and indicates the risk of minor collisions using different colors. It combines four cameras installed at the front and back of the vehicle to show all areas around the car, with a 3D wide-angle laser, which measures the distance to surrounding objects, and corrects the clarity of images displayed on the monitor. With this accurate representation drivers can easily see objects around the car without distortion, enabling drivers to quickly identify the risk of collision.
In the future this technology will be used to improve driving safety when the risk of a minor collision is higher.
Commercialization of driving support systems to help identify risks and reduce vehicle accidents
The ability to reproduce images from four cameras without distortion is the power behind this groundbreaking technology. Previous technologies used four onboard cameras to reproduce the 3D images, however these had several problems. For example, distortions generated when images were synthesized made it difficult for drivers to fully understand the situation around them well enough to determine the distance to the objects. This new technology combines images from onboard cameras with information on their distances measured using a laser radar, to enable projection planes to be created that meet the shapes of the 3D objects, thus successfully eliminating image distortions. In addition, by using the laser radar, distance analysis ensures minor collision risks are measured at an accuracy level of up to two centimeters. For this reason, the new technology enables drivers to avoid minor collisions while confirming distortion-free images even in complicated situations in which vehicles, pedestrians, and other 3D objects are involved. Fujitsu expect the system to reduce minor collisions as drivers will not only be able to fully understand their surroundings but will also be able to determine the distances to 3D objects.
Currently approximately 90% of accidents can be attributed to human error, such as carelessness and an error in judgment. Going forward, Fujitsu will contribute to reducing the amount of accidents by developing driver support systems that leverage this new technology.