Scientists develop self-driving car tech that can detect hidden objects

Stanford University researchers have developed a system that they say could transform driving safety

Researchers at Stanford University have developed a laser-based imaging technology that, they claim, can warn drivers about hidden objects.

The technology could also be built-in to autonomous vehicle systems

When driving a car, it is not always possible to identify potential hazards. But the scientists believe that their solution could radically improve car safety.

"A driverless car is making its way through a winding neighbourhood street, about to make a sharp turn onto a road where a child's ball has just rolled," wrote the university.

"Although no person in the car can see that ball, the car stops to avoid it. This is because the car is outfitted with extremely sensitive laser technology that reflects off nearby objects to see around corners."

The technology, which could one day be implemented into connected and driverless cars, users lasers to take images of hidden objects.

Even though this system is targeted at autonomous vehicles, the researchers explained that it could also be ground-breaking for aerial vehicles and rescue teams. For instance, it could help them locate national disaster victims.

Gordon Wetzstein, assistant professor of electrical engineering and author of the study, said: "It sounds like magic but the idea of non-line-of-sight imaging is actually feasible."

This is not the first time that technologists have developed software that uses lasers to photograph objects, but the scientists said they have developed an "extremely efficient and effective algorithm" to refine results.

Graduate student David Lindell, who helped with the project, added: "A substantial challenge in non-line-of-sight imaging is figuring out an efficient way to recover the 3D structure of the hidden object from the noisy measurements."

In the study, the researchers combined a laser with a photon detector to capture light particles.

"They shoot pulses of laser light at a wall and, invisible to the human eye, those pulses bounce off objects around the corner and bounce back to the wall and to the detector," explained the university.

It continued: "Currently, this scan can take from two minutes to an hour, depending on conditions such as lighting and the reflectivity of the hidden object."

After this stage, the algorithm unpacks the captured photons. The researchers said that their system can calculate results within less than a minute.

Matthew O'Toole, co-lead author of the paper, added: "We believe the computation algorithm is already ready for LIDAR systems. The key question is if the current hardware of LIDAR systems supports this type of imaging."