First Internet

Optical filters near infrared help robot to ‘see’ humans in disasters

Optical filters near infrared are among the components helping a new robotic system to ‘see’ humans caught in disaster zones, and crucially to be able to tell whether they are humans, or are simply debris.

The hugely complex project has been undertaken in Mexico at the University of Guadalajara, led by researcher Nancy Guadalupe Arana Daniel, and centres on a mobile robot equipped with several imaging systems.

Among them are a stereoscopic camera and flashlight, a laser, motion sensors, and an infrared system, all of which work together to allow the robot to build up an image of its environment.

With onboard optical filters near infrared, the robot can be self-contained, or can transmit its data back to a laptop for processing, if the environment is too hazardous for the more fragile self-contained robot to be sent in.

It looks for structures among the debris, virtually reconstructs them, and then determines whether or not it is likely that they are the silhouette of a human figure.

“Pattern recognition allows the descriptors to automatically distinguish objects containing information about the features that represent a human figure,” the developer explains.

“This involves developing algorithms to solve problems of the descriptor and assign features to an object.”

Ultimately, the robot is to be equipped with learning capabilities – meaning from each successful identification of a human, or each previous failure, its future decisions should become more accurate.