System and method for transmitting color and depth information

1. A system for transmitting color and depth information, comprising:

a camera configured to capture a color image frame of a scene;

a depth camera configured to capture a grayscale depth image frame of the scene;

at least one computing device comprising a processor and memory, the at least one computing device being programmed to perform steps including:

identifying a target in the scene with a target detector using computer vision algorithms;

generating a bounding box encapsulating the target, the bounding box having a depth dimension;

applying the bounding box to the grayscale depth image to create a bound grayscale depth image and to the color image frame to create a bound color image;

converting the bound grayscale depth image into a bound color depth image by linking and normalizing the depth dimension of the bounding box to a transferred color gamut by implementing a transfer function, wherein the transfer function varies a compression ratio of the target within the bounding box, reducing the compression ratio in key points of the target and increasing the compression ratio for low interest areas of the target;

merging the bound color image and the bound color depth image together into a bound merged color image, the bound merged color image containing information that is focused on the target; and

encoding the bound merged color image into an image bitstring.

2. A system as in claim 1 , wherein the at least one computing device is further programmed to generate a plurality of bounding boxes and to create a collage of images by merging the bounding boxes.

3. A system as in claim 1 , wherein the at least one computing device is further programmed to generate the color gamut and to generate the transfer function that is used to generate, from the bound grayscale depth images, the bound color depth images.

4. A system as in claim 2 wherein the at least one computing device is further programmed to use target key points of the target when generating the transfer function.

5. A system as in claim 2 wherein the bounding box adapts dynamically according to the size, shape and orientation of the target, wherein the transferred color gamut accordingly adapts by altering in length to stay normalized along the depth dimension, and wherein the transfer function is adjustable and also adjusts to provide for the change in depth dimension so that the depth dimension of the target is substantially covered by the transfer function and transferred color gamut.

6. A system as in claim 5 wherein the transfer function is adjustable by varying the gradient of the function.

7. A system as in claim 1 , wherein the transfer function is a continuous bijective or invertible function.

8. A system as in claim 1 wherein the transfer function is implemented by a generalized logistic function.

9. A system as in claim 1 wherein the target is a person comprising a face and a torso or the full body of a person.

10. A system as in claim 1 wherein at least one computing device is further programmed to crop the bound grayscale depth image and the bound color images.

11. A method for transmitting color and depth information comprising:

capturing a color image frame of a scene;

capturing a grayscale depth image frame of the scene;

identifying a target in the scene with a target detector using computer vision algorithms;

generating a bounding box encapsulating the target, the bounding box having a depth dimension;

applying the bounding box to the grayscale depth image to create a bound grayscale depth image and to the color image frame to create a bound color image;

converting the bound grayscale depth image into a bound color depth image by linking and normalizing the depth dimension of the bounding box to a transferred color gamut;

implementing a transfer function, wherein the transfer function varies a compression ratio of the target within the bounding box, reducing the compression ratio in key points of the target and increasing the compression ratio for low interest areas of the target;

merging the bound color image and the bound color depth image together into a bound merged color image, the bound merged color image containing information that is focused on the target; and

encoding the bound merged color image into an image bitstring.

12. A method as in claim 11 further comprising generating a plurality of bounding boxes and creating a collage of images composed of merged bounding boxes.

13. A method as in claim 11 further comprising;

generating the color gamut; and

generating the transfer function that is used to generate, from the bound grayscale depth image, the bound color depth image.

14. A method as in claim 13 wherein the generating of the transfer function uses target key points of the target.

15. A method as in claim 13 further comprising:

adapting the bounding box dynamically according to the size, shape and orientation of the target;

adapting the transferred color gamut by altering its length to stay normalized along the depth dimension; and

adjusting the transfer function to provide for the change in depth dimension so that the depth dimension of the target is substantially covered by the transfer function and transferred color gamut.

16. A method as in claim 15 further comprising: adjusting the transfer function by varying the gradient of the function.