Researchers at Stanford University in the US have created a new type of artificially intelligent (AI) camera system that can classify images more quickly and could be used for autonomous vehicles.

The image recognition technology will enable autonomous vehicles to recognise objects such as animals, a pedestrian crossing the street or a stopped car.

The AI camera is expected to be built small enough to be embedded in the devices themselves in the future.

Stanford University Electrical Engineering assistant professor Gordon Wetzstein said: “That autonomous car you just passed has a relatively huge, relatively slow, energy intensive computer in its trunk. Future applications will need something much faster and smaller to process the stream of images.”

“The smaller, faster technology is expected to potentially replace the trunk-size computers that are used to help cars and drones recognise general surroundings.”

Researchers have developed the technology by combining two types of computers into one, making a hybrid of optical-electrical computer designed specifically for image analysis.

The first layer of the prototype AI camera is optical computer and does not require the power-intensive mathematics of digital computing, while the second layer is a traditional digital electronic computer.

Sanford University graduate student Julie Chang said: “We’ve outsourced some of the math of artificial intelligence into the optics.”

In terms of speed and accuracy, the prototype competes with existing electronic-only computing processors, but with substantial computational cost savings.

Researchers stated that the system could be miniaturised to fit in a handheld video camera or an aerial drone in the future.

During both simulations and real-world experiments, the system successfully identifies airplanes, automobiles, cats, dogs and more within natural image settings.

Stanford University graduate student Wetzstein said: “Some future version of our system would be especially useful in rapid decision-making applications, like autonomous vehicles.”

The smaller, faster technology is expected to potentially replace the trunk-size computers that are used to help cars and drones recognise general surroundings.