3D imaging recognition by stereo matching of RGB and infrared images

A device for improved anti-spoofing in 3D image recognition for biometric authentication, comprising

• a first imaging sensor, configured to collect a first image of a first wavelength range of light from an object;
• a second imaging sensor, configured to collect a second image of a second wavelength range of light from the object, wherein the first imaging sensor and a second imaging sensor are placed spatially apart and triggered substantially concurrently;
• and
• a processor, configured to:
• identify a quantity of landmark areas of the first image, collected from the first imaging sensor, the quantity consisting of a lower threshold of 5 and an upper threshold of 15,
• identify a quantity of matching landmark areas in the second image, collected from the second imaging sensor, each of the matching landmark areas matching a landmark area of the quantity of landmark areas, and
• create a sparse depth map based on stereo vision algorithms comparing each of the landmark areas and matching landmark areas,
• compare the sparse depth map to one or more geometric templates, and
• determine whether the sparse depth map matches the one or more geometric templates.