Thermal based presentation attack detection for biometric systems
A presentation attack detection system includes RGB and thermal imaging cameras for acquiring images of an individual and a processor operable to determine a metric for whether the individual is a real human based on the thermal imaging and RGB image data. Related methods are also described.
1. A presentation attack detection system for detecting a presentation attack or spoof of an individual presenting a face, the system comprising:
an RGB camera;
a thermal imaging camera;
at least one processor operable to:
acquire an RGB image of the individual from the RGB camera and acquire a thermal image of the individual from the thermal imaging camera; and
determine a metric whether the individual is a real human based on data from the thermal image and RGB image
wherein the RGB and thermal imaging cameras are arranged with an optical substrate to provide a common aperture axis; and
wherein the optical substrate allows light having a wavelength in a visible spectrum range to pass through an entrance window of the RGB camera with thermal wavelengths reflecting to the thermal imaging camera.
2. The system of claim 1 , further comprising a calibration apparatus adapted to provide a thermal black body source for the at least one thermal imaging camera and a plurality of colors for the at least one RGB camera to improve an accuracy of skin detection.
3. The system of claim 1 , wherein the system is operable to simultaneously acquire the RGB and thermal images from the RGB and thermal imaging cameras.
4. The system of claim 1 , wherein the at least one processor is operable to correlate the RGB image to the thermal image.
5. The system of claim 4 , wherein the at least one processor includes a trained classifier to determine the metric based on the correlated images.
6. The system of claim 4 , wherein the at least one processor is operable to detect a presence of eye glasses based on the correlated images.
7. The system of claim 4 , wherein the at least one processor is operable to detect a use of makeup based on the correlated images.
8. The system of claim 4 , wherein the at least one processor is operable to detect a presence of perspiration based on the correlated images.
9. The system of claim 1 , wherein the at least one processor is operable to find the face based on the RGB image.
10. The system of claim 9 , wherein the at least one processor is operable to identify landmarks of the face in the RGB image.
11. The system of claim 10 , wherein the at least one processor is operable to correlate the landmarks of the RGB image to the thermal image.
12. The system of claim 11 , wherein the at least one processor is operable to evaluate the correlated landmarks for at least one of spatial varying IR patterns, temporal varying IR patterns, or calibrated IR patterns indicative of a real face.
13. The system of claim 12 , wherein the correlated landmarks comprise a medial or inner canthus of each eye.
14. The system of claim 1 , wherein the at least one processor is operable to determine emissivity values from the thermal image.
15. The system of claim 14 , wherein the at least one processor is operable to correlate the emissivity values to the RGB image.
16. The system of claim 15 , wherein the at least one processor includes a trained classifier to determine the metric based on the correlated emissivity values.
17. The system of claim 1 , wherein the optical substrate reflects light having a wavelength in a far infrared range.
18. The system of claim 17 , wherein the optical substrate is retractable.
19. The system of claim 1 , wherein the at least one processor is operable to recognize a real human from a presentation attack comprising at least one of paper, video, make-up, a 3D mask, a 3D model, or a prosthetic.
20. The system of claim 1 , wherein the at least one processor is further operable to provide an instruction for a slave device to perform an action based on the metric.
21. The system of claim 20 , wherein the slave device is at least one of a physical or logical access device, point of sale device, lock, e-lock, enrollment device, person validation device, or biometric identity device.
22. A system for detecting a presentation attack or spoof of an individual presenting a face, the system comprising:
a light source aimed at the face;
a polarimeter operable to generate polarization information of light reflected from the face; and
at least one processor operable to determine a metric whether the individual is a real human based on the polarization information from the polarimeter.
23. The system of claim 22 , wherein the at least one processor is programmed with a trained classifier to detect whether the face is real based on the polarization information from the polarimeter.
24. The system of claim 22 , wherein the polarization information comprises polarization intensity information.
25. A presentation attack detection system for detecting a presentation attack or spoof of an individual presenting a face, the system comprising:
at least one indirect time of flight (ITOF) sensor comprising a first illuminator and a second illuminator in the near IR for obtaining depth and imaging information from the individual, wherein the wavelength of light of the first illuminator is different than the wavelength of light of the second illuminator;
at least one processor operable to determine a metric whether the individual is a real human based on the depth and imaging information from the at least one ITOF sensor.
26. The system of claim 25 , wherein the at least one illuminator is a power calibrated illuminator.
27. The system of claim 26 , wherein the depth and image information is absolute skin reflectance information.
28. The system of claim 25 , further comprising a polarimeter operable to generate polarization information of the light reflected from the face; and wherein the at least one processor is further operable to determine whether the individual is a real human based on polarization information in combination with the depth and imaging information.