Low power data generation for iris-related detection and authentication

1. A vision sensor system for sensing scene-based occurrences, the vision sensor system comprising:

dedicated computer vision (CV) computation hardware configured to receive sensor data from at least one sensor array comprising a plurality of sensor pixels and capable of computing one or more CV features using a local binary pattern (LBP) calculation, wherein the LBP calculation comprises scoring readings from multiple neighboring sensor pixels within the at least one sensor array based on at least their respective difference in intensity; and

a first processing unit communicatively coupled with the dedicated CV computation hardware, wherein the vision sensor system is configured to, in response to processing of the one or more computed CV features indicating a presence of one or more irises in a scene captured by the at least one sensor array, ( 1 ) generate location data for a plurality of facial landmarks, including location data indicating location within an image for a first landmark and location data indicating location within the image for a second landmark corresponding to a left corner and a right corner, respectively, of a first eye and location data indicating location within the image for a third landmark and location data indicating location within the image for a fourth landmark corresponding to a left corner and a right corner, respectively, of a second eye, in support of iris-related operations to be performed in a second processing unit and ( 2 ) send the generated location data for the plurality of facial landmarks, including location data indicating location within the image for the first landmark and location data indicating location within the image for the second landmark corresponding to the left corner and the right corner, respectively, of the first eye and location data indicating location within the image for the third landmark and location data indicating location within the image for the fourth landmark corresponding to the left corner and the right corner, respectively, of the second eye, to the second processing unit.

2. The system of claim 1 , wherein the processing of the one or more computed CV features indicating the presence of the one or more irises in the scene comprises the one or more computed CV features indicating a presence of at least one face in the scene.

3. The system of claim 1 , wherein the generated location data is indicative of location of one or more eyes.

4. The system of claim 1 , wherein the generated location data is indicative of a size for each of one or more eyes.

5. The system of claim 1 , wherein the vision sensor system is configured to generate the location data for the plurality of facial landmarks using a trained classifier.

6. The system of claim 5 , wherein the trained classifier comprises a plurality of cascaded stages.

7. The system of claim 1 ,

wherein the at least one sensor array comprises two sensor arrays, a visual sensor array configured to capture visible light and an infrared (IR) sensor array configured to capture IR light,

wherein the vision sensor system is further configured to send an indication for an IR light source to be lit, in response to the processing of the one or more computed CV features indicating the presence of one or more irises in the scene, and

wherein the vision sensor system is configured to generate the location data in support of iris-related operations using readings from the IR sensor array captured while the IR light source is lit.

8. The system of claim 7 , wherein the visual sensor array is configured to capture images at a first resolution, and the IR sensor array is configured to capture images at a second resolution higher than the first resolution.

9. The system of claim 1 ,

wherein the at least one sensor array comprises an IR-capable visual sensor array capable of capturing both visual light and IR light,

wherein the vision sensor system is further configured to send an indication for an IR light source to be lit, in response to the processing of the one or more computed CV features indicating the presence of one or more irises in the scene, and

wherein the vision sensor system is configured to generate the location data in support of iris-related operations using readings from the IR-capable visual sensor array captured while the IR light source is lit.

10. The system of claim 1 , wherein the vision sensor system is further configured to:

generate an event in response to the processing of the one or more computed CV features indicating the presence of the one or more irises in the scene captured by the at least one sensor array, and

send the event to the second processing unit; and

wherein the second processing unit is configured to, in response to receiving the event, transition from a first power state to a second power state, the second power state being associated with a higher power level than the first power state.

11. A method of sensing scene-based occurrences comprising:

receiving, by dedicated computer vision (CV) computation hardware in a vision sensor system comprising the dedicated CV computation hardware and a first processing unit communicatively coupled with the dedicated CV computation hardware, sensor data from at least one sensor array comprising a plurality of sensor pixels;

computing, by the dedicated CV computation hardware, one or more CV features using an LBP calculation comprising scoring readings from multiple neighboring sensor pixels within the at least one sensor array based on at least their respective difference in intensity;

generating, by the vision sensor system, in response to processing of the one or more computed CV features indicating a presence of one or more irises in a scene captured by the at least one sensor array, location data for a plurality of facial landmarks, including location data indicating location within an image for a first landmark and location data indicating location within the image for a second landmark corresponding to a left corner and a right corner, respectively, of a first eye and location data indicating location within the image for a third landmark and location data indicating location within the image for a fourth landmark corresponding to a left corner and a right corner, respectively, of a second eye, in support of iris-related operations to be performed in a second processing unit; and

sending the generated location data, including location data indicating location within the image for the first landmark and location data indicating location within the image for the second landmark corresponding to the left corner and the right corner, respectively, of the first eye and location data indicating location within the image for the third landmark and location data indicating location within the image for the fourth landmark corresponding to the left corner and the right corner, respectively, of the second eye, to the second processing unit.

12. The method of claim 11 , wherein the processing of the one or more computed CV features indicating the presence of the one or more irises in the scene captured by the at least one sensor array comprises the processing of the one or more computed CV features indicating a presence of at least one face in the scene.

13. The method of claim 11 , wherein the generated location data is indicative of location of one or more eyes.

14. The method of claim 11 , wherein the generated location data is indicative of a size for each of one or more eyes.

15. The method of claim 11 , wherein the vision sensor system is configured to generate the location data for the plurality of facial landmarks using a trained classifier.

16. The method of claim 15 , wherein the trained classifier comprises a plurality of cascaded stages.

17. The method of claim 11 , further comprising:

sending, by the vision sensor system, an indication for an infrared (IR) light source to be lit, in response to the processing of the one or more computed CV features indicating the presence of the one or more irises in the scene captured by the at least one sensor array,

wherein the generating, by the vision sensor system, the location data in support of iris-related operations comprises generating the location data using readings from an IR sensor array captured while the IR light source is lit, and

wherein the at least one sensor array comprises two sensor arrays, a visual sensor array configured to capture visible light and the IR sensor array configured to capture IR light.

18. The method of claim 17 , wherein the visual sensor array is configured to capture images at a first resolution, and the IR sensor array is configured to capture images at a second resolution higher than the first resolution.

19. The method of claim 11 , further comprising:

sending, by the vision sensor system, an indication for an IR light source to be lit, in response to the processing of the one or more computed CV features indicating the presence of the one or more irises in the scene captured by the at least one sensor array,

wherein the generating, by the vision sensor system, the location data in support of iris-related operations comprises generating the location data using readings from an IR-capable visual sensor array captured while the IR light source is lit, and

wherein the at least one sensor array comprises the IR-capable visual sensor array, the IR-capable visual sensor capable of capturing both visual light and IR light.

20. The method of claim 11 , further comprising:

generating, by the vision sensor system, an event in response to the processing of the one or more computed CV features indicating the presence of the one or more irises in the scene captured by the at least one sensor array; and

sending, by the vision sensor system, the event to the second processing unit, wherein the second processing unit is configured to, in response to receiving the event, transition from a first power state to a second power state, the second power state being associated with a higher power level than the first power state.

21. A system of sensing scene-based occurrences comprising:

means for receiving sensor data from at least one sensor array comprising a plurality of sensor pixels;

hardware means for computing one or more CV features using an LBP calculation, wherein the LBP calculation comprises scoring readings from multiple neighboring sensor pixels within the at least one sensor array based on at least their respective difference in intensity;

means for processing the one or more CV features to indicate a presence of one or more irises in a scene captured by the at least one sensor array;

means for generating, in response to an indication of the presence of the one or more irises in the scene captured by the at least one sensor array, location data for a plurality of facial landmarks, including location data indicating location within an image for a first landmark and location data indicating location within the image for a second landmark corresponding to a left corner and a right corner, respectively, of a first eye and location data indicating location within the image for a third landmark and location data indicating location within the image for a fourth landmark corresponding to a left corner and a right corner, respectively, of a second eye, in support of iris-related operations to be performed in a second processing unit; and

means for sending the generated location data, including location data indicating location within the image for the first landmark and location data indicating location within the image for the second landmark corresponding to the left corner and the right corner, respectively, of the first eye and location data indicating location within the image for the third landmark and location data indicating location within the image for the fourth landmark corresponding to the left corner and the right corner, respectively, of the second eye, to the second processing unit.

22. The system of claim 21 , wherein the means for generating corresponding to the plurality of facial landmarks defined with respect to one or more eyes comprises a trained classifier.

23. A non-transitory computer-readable medium storing instructions therein for execution by one or more processing units, comprising instructions to:

control receipt, by dedicated computer vision (CV) computation hardware in a vision sensor system comprising the dedicated CV computation hardware and a first processing unit communicatively coupled to the dedicated CV computation hardware, sensor data from at least one sensor array comprising a plurality of sensor pixels;

control computation, by the dedicated CV computation hardware, of one or more CV features using an LBP calculation, wherein the LBP calculation comprises scoring readings from multiple neighboring sensor pixels within the at least one sensor array based on at least their respective difference in intensity;

control generation, by the vision sensor system, in response to processing of the one or more computed CV features indicating a presence of one or more irises in a scene captured by the at least one sensor array, location data for a plurality of facial landmarks, including location data indicating location within an image for a first landmark and location data indicating location within the image for a second landmark corresponding to a left corner and a right corner, respectively, of a first eye and location data indicating location within the image for a third landmark and location data indicating location within the image for a fourth landmark corresponding to a left corner and a right corner, respectively, of a second eye, in support of iris-related operations to be performed in a second processing unit; and

control sending of the generated location data, including location data indicating location within the image for the first landmark and location data indicating location within the image for the second landmark corresponding to the left corner and the right corner, respectively, of the first eye and location data indicating location within the image for the third landmark and location data indicating location within the image for the fourth landmark corresponding to the left corner and the right corner, respectively, of the second eye, to the second processing unit.

24. The non-transitory computer-readable medium of claim 23 , wherein the location data for the plurality of facial landmarks is generated using a trained classifier.