Delphi develops automotive facial recognition system
Electronics and technology supplier, Delphi, has created a facial recognition system for cars which reduces strain on the computer’s processing power and that the company says improves vehicle safety, Automotive Engineer reports.
Delphi’s system uses a system of lights and a camera to project a line pattern onto the driver’s face. The system then records a two-dimensional image of the driver and compares that image to stored images already saved on the car’s computer.
As opposed to three-dimensional processing systems, two-dimensional systems like Delphi’s facial recognition system interpret the line pattern of the subject’s face as a facial fingerprint, whereas in a three-dimensional system, visualizations are created and are compared geometrically and spatially. The ForensicaGPS made by Animetrics is an example of a three-dimensional recognition system.
“This system is considerably different from other facial recognition systems,” the company said. “Most systems are designed to create a geometric map of the subject’s face using points such as the nose, mouth and eyes. This method requires large processing capability and additional time to determine the location of those facial features contained in the image.”
Once a positive match is established, Delphi’s system allows for cabin personalization such as seat, radio, climate and unlocking the vehicle.
Car manufacturers are increasingly looking to include biometrics into their most recent offerings. As reported previously in BiometricUpdate.com, engineers at the Ford Research and Innovation Lab are designing cars with biometric sensors which will examine pulse and respiratory patterns among other measurements to gauge a driver’s current level of stress.