Camera auto-calibration system

1. A computer-implemented method comprising:

calibrating a seed camera disposed at a first location;

detecting, with a second camera disposed at a second location, a physical marker based on predefined characteristics of the physical marker, the physical marker being located within an overlapping field of view between the seed camera and the second camera;

computing a physical location of the physical marker with the seed camera; and

calibrating the second camera based on a combination of the physical location of the physical marker, the first location of the seed camera, the second location of the second camera, a first image of the physical marker generated with the seed camera, and a second image of the physical marker generated with the second camera.

2. The computer-implemented method of claim 1 , further comprising:

moving the physical marker to a fourth location, the fourth location being located within an overlapping field of view between the second camera and a third camera;

detecting, with the third camera disposed at a third location, the physical marker based on a predefined visual pattern of the physical marker;

computing an updated physical location of the physical marker with the second camera; and

calibrating the third camera based on a combination of the updated physical location of the physical marker, the third location of the third camera, a third image of the physical marker generated by the second camera, and a fourth image of the physical marker generated by the third camera.

3. The computer-implemented method of claim 1 , wherein calibrating the seed camera further comprises:

receiving parameters related to the seed camera based on a combination of lens characteristics of the seed camera, three-dimensional coordinates corresponding to the first location of the seed camera, and an orientation of the seed camera.

4. The computer-implemented method of claim 1 , wherein the physical marker comprises a lamp, wherein the predefined characteristics comprise a predefined height of the lamp, a predefined size of the lamp, and a visual pattern produced by a predefined toggle switch state pattern of the lamp.

5. The computer-implemented method of claim 4 , wherein detecting the physical marker comprises:

detecting the lamp based on the predefined toggle switch state pattern of the lamp.

6. The computer-implemented method of claim 5 , further comprising:

computing real-world coordinates of the lamp based on an image of the lamp obtained with the seed camera and the first location of the seed camera.

7. The computer-implemented method of claim 5 , further comprising:

detecting foreground pixels in an image generated with the seed camera;

detecting an object based on the foreground pixels;

validating the object based on at least one of a predefined static height, a shape of the lamp, and a size of the lamp; and

tracking the object within images from the seed camera and images from the second camera.

8. The computer-implemented method of claim 1 , wherein the seed camera and the second camera are part of a network of cameras disposed at different locations within a facility.

9. The computer-implemented method of claim 8 , further comprising:

moving the physical marker along a path throughout the facility, the path covering different overlapping fields of view between calibrated cameras and uncalibrated cameras from the network of cameras.

10. The computer-implemented method of claim 8 , wherein the seed camera computes the physical location of the physical marker with the seed camera, wherein the second camera calibrates itself based on the combination of the physical location of the physical marker, the first location of the seed camera, the second location of the second camera, the first image of the physical marker generated with the seed camera, and the second image of the physical marker generated with the second camera.

11. A computing apparatus, the computing apparatus comprising:

a processor; and

a memory storing instructions that, when executed by the processor, configure the apparatus to:

calibrate a seed camera disposed at a first location;

detect, with a second camera disposed at a second location, a physical marker based on predefined characteristics of the physical marker, the physical marker being located within an overlapping field of view between the seed camera and the second camera;

compute a physical location of the physical marker with the seed camera; and

calibrate the second camera based on a combination of the physical location of the physical marker, the first location of the seed camera, the second location of the second camera, a first image of the physical marker generated with the seed camera, and a second image of the physical marker generated with the second camera.

12. The computing apparatus of claim 11 , wherein the instructions further configure the apparatus to:

move the physical marker to a fourth location, the fourth location being located within an overlapping field of view between the second camera and a third camera;

detect, with the third camera disposed at a third location, the physical marker based on a predefined visual pattern of the physical marker;

compute an updated physical location of the physical marker with the second camera; and

calibrate the third camera based on a combination of the updated physical location of the physical marker, the third location of the third camera, a third image of the physical marker generated by the second camera, and a fourth image of the physical marker generated by the third camera.

13. The computing apparatus of claim 11 , wherein calibrating the seed camera further comprises:

receive parameters related to the seed camera based on a combination of lens characteristics of the seed camera, three-dimensional coordinates corresponding to the first location of the seed camera, and an orientation of the seed camera.

14. The computing apparatus of claim 11 , wherein the physical marker comprises a lamp, wherein the predefined characteristics comprise a predefined height of the lamp, a predefined size of the lamp, and a visual pattern produced by a predefined toggle switch state pattern of the lamp.

15. The computing apparatus of claim 14 , wherein detecting the physical marker comprises:

detect the lamp based on the predefined toggle switch state pattern of the lamp.

16. The computing apparatus of claim 15 , wherein instructions further configure the apparatus to:

compute real-world coordinates of the lamp based on an image of the lamp obtained with the seed camera and the first location of the seed camera.

17. The computing apparatus of claim 15 , wherein the instructions further configure the apparatus to:

detect foreground pixels in an image generated with the seed camera;

detect an object based on the foreground pixels;

validate the object based on at least one of a predefined static height, a shape of the lamp, and a size of the lamp; and

track the object within images from the seed camera and images from the second camera.

18. The computing apparatus of claim 11 , wherein the seed camera and the second camera are part of a network of cameras disposed at different locations within a facility.

19. The computing apparatus of claim 18 , wherein the instructions further configure the apparatus to:

move the physical marker along a path throughout the facility, the path covering different overlapping fields of view between calibrated cameras and uncalibrated cameras from the network of cameras.

20. A non-transitory computer-readable storage medium, the computer-readable storage medium including instructions that when executed by a computer, cause the computer to:

calibrate a seed camera disposed at a first location;

detect, with a second camera disposed at a second location, a physical marker based on predefined characteristics of the physical marker, the physical marker being located within an overlapping field of view between the seed camera and the second camera;

compute a physical location of the physical marker with the seed camera; and

calibrate the second camera based on a combination of the physical location of the physical marker, the first location of the seed camera, the second location of the second camera, a first image of the physical marker generated with the seed camera, and a second image of the physical marker generated with the second camera.