System and method to improve object detection accuracy by focus bracketing

1. A method comprising:

receiving, by a processor, a first image from an image sensor, wherein the first image is obtained by the image sensor using a first focal length;

receiving, by the processor, a second image from the image sensor, wherein the second image is obtained by the image sensor using a second focal length that is different than the first focal length;

detecting, by the processor, one or more objects in the first image;

generating, by the processor, a third image by applying one or more object masks to the second image, wherein the one or more object masks identify one or more regions of the first image that correspond to the one or more first objects detected in the first image;

detecting, by the processor, one or more second objects in the third image; and

combining, by the processor, the one or more objects detected in the first image with the one or more objects detected in the third image.

2. The method of claim 1 , wherein the first image comprises a near field focus image and the second image comprises a far field focus image.

3. The method of claim 2 , further comprising, prior to obtaining the first image and the second image, configuring, by the processor, the image sensor to selectively transition between the near field focus image and the far field focus image.

4. The method of claim 1 , wherein the image sensor is positioned to obtain a top view of an image scene.

5. The method of claim 4 , wherein each of the one or more objects detected in the first image and each of the one or more second objects detected in the third image is a person.

6. The method of claim 5 , wherein the one or more regions correspond to head and shoulders of the person.

7. The method of claim 1 , further comprising generating, by the processor, a dictionary based on detection results in the first image and the third image.

8. The method of claim 1 , further comprising saving, by the processor, information about the one or more objects detected in the first region and the one or more objects detected in the third region as metadata.

9. A system comprising:

a hardware processor configured to:

receive a first image from an image sensor, wherein the first image is obtained by the image sensor using a first focal length;

receive a second image from the image sensor, wherein the second image is obtained by the image sensor using a second focal length that is different than the first focal length;

detect one or more objects in the first image;

generate a third image by applying one or more object masks to the second image, wherein the one or more object masks identify one or more regions of the first image that corresponds to the one or more first objects detected in the first image;

detect one or more objects in the third image; and

combine the one or more objects detected in the first image with the one or more second objects detected in the third image.

10. The system of claim 9 , wherein the first image comprises a near field focus image and the second image comprises a far field focus image.

11. The system of claim 10 , wherein the hardware processor is further configured to configure, prior to obtaining the first image and the second image, the image sensor to selectively transition between the near field focus image and the far field focus image.

12. The system of claim 9 , wherein the image sensor is positioned to obtain a top view of an image scene.

13. The system of claim 12 , wherein each of the one or more first objects detected in the first image and each of the one or more objects detected in the third image is a person.

14. The system of claim 13 , wherein the one or more regions correspond to head and shoulders of the person.

15. The system of claim 9 , wherein the hardware processor is further configured to generate a dictionary based on detection results in the first image and the third image.

16. The system of claim 9 , wherein the hardware processor is further configured to save information about the one or more first objects detected in the first image and the one or more second objects detected in the third image as metadata.

17. The system of claim 9 , wherein the one or more object masks identify the one or more regions of the first image that correspond to the one or more first objects that are in-focus at the first focal length in the first image.

18. The system of claim 17 , wherein to generate the third image, the hardware processor is configured to apply the one or more object masks to mask out the one or more first objects as detected at the second focal length in the second image.

19. The method of claim 1 , wherein the one or more object masks identify the one or more regions of the first image that correspond to the one or more first objects that are in-focus at the first focal length in the first image.

20. The method of claim 19 , wherein generating the third image comprises applying the one or more object masks to mask out the one or more first objects as detected at the second focal length in the second image.