Optimized light field representation for compression

1 . A computer-implemented method, comprising:

accessing image data, wherein the image data comprises a plurality of two-dimensional (2D) perspective images that exhibit parallax for presentation on a three-dimensional (3D) display;

encoding the image data, wherein the encoding comprises:

generating a group of pictures (GOP) that comprises the plurality of 2D perspective images; and

ordering the plurality of 2D perspective images within the GOP in a particular order based on a set of evaluated metrics derived from content of the plurality of 2D perspective images by determining a prediction efficiency for each 2D perspective image of the plurality of 2D perspective images, wherein the prediction efficiency corresponds to a degree to which each 2D perspective image is usable to predict characteristics of other 2D perspective images of the plurality of 2D perspective images; and

transmitting the encoded image data for display.

2 . The method of claim 1 , wherein the encoded image data is transmitted to a client device configured to decode the encoded image data and generate the decoded image data for display, and

the particular order of the plurality of 2D perspective images within the GOP enables the image data to be transmitted to the client device in a manner that maximizes visual quality of the displayed decoded image data for a given data rate of transmitting the image data, or the particular order of the plurality of 2D perspective images within the GOP enables the image data image data for a given visual quality of the displayed decoded image data.

3 . The method of claim 1 , wherein the plurality of 2D perspective images are ordered within the GOP in the particular order based on the set of evaluated metrics derived from content of the plurality of 2D perspective images, and the ordering further comprises:

identifying a first 2D perspective image of the plurality of 2D perspective images, wherein a first data structure comprises the plurality of 2D perspective images;

modifying the first data structure by removing the first 2D perspective image from the first data structure;

determining a second 2D perspective image, of the 2D perspective images included in the modified first data structure, based on which the first 2D perspective image can be encoded using the least amount of data; and

generating a second data structure comprising the first 2D perspective image and the second 2D perspective image;

wherein ordering the plurality of 2D perspective images within the GOP is based on the second data structure.

4 . The method of claim 3 , wherein the ordering further comprises:

determining a third 2D perspective image, of the 2D perspective images included in the modified first data structure, that can be encoded based on the first 2D perspective image using the least amount of data; and

generating a second data structure comprising the first 2D perspective image and at least one of the second 2D perspective image or the third 2D perspective image;

wherein ordering the plurality of 2D perspective images within the GOP is based on the second data structure.

5 . The method of claim 1 , wherein ordering the plurality of 2D perspective images within the GOP in the particular order is further based on equipment characteristics associated with equipment used to capture the plurality of 2D perspective images, and the method further comprises:

ordering the plurality of 2D perspective images within the GOP in the particular order further based at least in part on a first weight and a second weight, wherein the first weight is associated with the prediction efficiencies for each 2D perspective image of the plurality of 2D perspective images and is assigned based at least in part on comparing the prediction efficiencies to a threshold, and wherein the second weight is associated with the equipment characteristics and is assigned based on how closely the equipment characteristics match stored equipment characteristics associated with historical GOP ordering data.

6 . The method of claim 5 , wherein the equipment characteristics comprise at least one of a spatial arrangement of the equipment used to capture the plurality of 2D perspective images or an angular orientation of the equipment used to capture the plurality of 2D perspective images with respect to a captured scene corresponding to the image data.

7 . The method of claim 5 , wherein the equipment characteristics comprise a type of the equipment used to capture the plurality of 2D perspective images, and the type of the equipment used to capture the plurality of 2D perspective images comprises a multi-camera array, a camera gantry or a lenslet camera.

8 . The method of claim 5 , wherein the equipment characteristics comprise a stored depth range corresponding to a depth range from which a scene corresponding to the image data is captured by the equipment used to capture the plurality of 2D perspective images.

9 . The method of claim 5 , wherein the equipment characteristics comprise a stored distance corresponding to a distance between the equipment used to capture the plurality of 2D perspective images and a captured scene corresponding to the image data.

10 . A computer-implemented system, comprising:

input/output (I/O) circuitry; and

control circuitry configured to:

access image data, wherein the image data comprises a plurality of two-dimensional (2D) perspective images that exhibit parallax for presentation on a three-dimensional (3D) display, the image data being received via the I/O circuitry;

encode the image data by:

generating a group of pictures (GOP) that comprises the plurality of 2D perspective images; and

ordering the plurality of 2D perspective images within the GOP in a particular order based on a set of evaluated metrics derived from content of the plurality of 2D perspective images by determining a prediction efficiency for each 2D perspective image of the plurality of 2D perspective images, wherein the prediction efficiency corresponds to a degree to which each 2D perspective image is usable to predict characteristics of other 2D perspective images of the plurality of 2D perspective images; and

transmit the encoded image data for display.

11 . The system of claim 10 , wherein the control circuitry is configured to transmit the encoded image data to a client device that is configured to decode the encoded image data and generate the decoded image data for display, and

the particular order of the plurality of 2D perspective images within the GOP enables the image data to be transmitted to the client device in a manner that maximizes visual quality of the displayed decoded image data for a given data rate of transmitting the image data, or the particular order of the plurality of 2D perspective images within the GOP enables the image data to be transmitted to the client device in a manner that minimizes a data rate of transmitting the image data for a given visual quality of the displayed decoded image data.

12 . The system of claim 10 , wherein the plurality of 2D perspective images are ordered within the GOP in the particular order based on the set of evaluated metrics derived from content of the plurality of 2D perspective images, and the control circuitry is further configured to perform the ordering by:

identifying a first 2D perspective image of the plurality of 2D perspective images, wherein a first data structure comprises the plurality of 2D perspective images;

modifying the first data structure by removing the first 2D perspective image from the first data structure;

determining a second 2D perspective image, of the 2D perspective images included in the modified first data structure, based on which the first 2D perspective image can be encoded using the least amount of data; and

generating a second data structure comprising the first 2D perspective image and the second 2D perspective image;

wherein ordering the plurality of 2D perspective images within the GOP is based on the second data structure.

13 . The system of claim 12 , wherein the control circuitry is further configured to perform the ordering by:

determining a third 2D perspective image, of the 2D perspective images included in the modified first data structure, that can be encoded based on the first 2D perspective image using the least amount of data; and

generating a second data structure comprising the first 2D perspective image and at least one of the second 2D perspective image or the third 2D perspective image;

wherein ordering the plurality of 2D perspective images within the GOP is based on the second data structure.

14 . The system of claim 10 , wherein:

ordering the plurality of 2D perspective images within the GOP in the particular order is further based on equipment characteristics associated with equipment used to capture the plurality of 2D perspective images; and

the plurality of 2D perspective images are ordered within the GOP in the particular order based on the equipment characteristics associated with the equipment used to capture the plurality of 2D perspective images, the equipment characteristics comprising at least one of a spatial arrangement of the equipment used to capture the plurality of 2D perspective images or an angular orientation of the equipment used to capture the plurality of 2D perspective images with respect to a captured scene corresponding to the image data.

15 . The system of claim 10 , wherein:

ordering the plurality of 2D perspective images within the GOP in the particular order is further based on equipment characteristics associated with equipment used to capture the plurality of 2D perspective images; and

the equipment characteristics comprise a type of the equipment used to capture the plurality of 2D perspective images, and the type of the equipment used to capture the plurality of 2D perspective images comprises a multi-camera array, a camera gantry or a lenslet camera.

16 . The system of claim 10 , wherein:

ordering the plurality of 2D perspective images within the GOP in the particular order is further based on equipment characteristics associated with equipment used to capture the plurality of 2D perspective images; and

the equipment characteristics comprise a depth range from which a scene corresponding to the image data is captured by the equipment used to capture the plurality of 2D perspective images.

17 . The system of claim 10 , wherein:

ordering the plurality of 2D perspective images within the GOP in the particular order is further based on equipment characteristics associated with equipment used to capture the plurality of 2D perspective images; and

the equipment characteristics comprise a distance between the equipment used to capture the plurality of 2D perspective images and a captured scene corresponding to the image data.

18 . The system of claim 10 , wherein ordering the plurality of 2D perspective images within the GOP in the particular order is further based on equipment characteristics associated with equipment used to capture the plurality of 2D perspective images, and the control circuitry is further configured to store a plurality of ordering arrangements and corresponding equipment characteristics associated with the equipment used to capture the plurality of 2D perspective images; and

wherein the plurality of 2D perspective images are ordered within the GOP in the particular order based on the characteristics associated with the equipment used to capture the plurality of 2D perspective images, and the control circuitry is further configured to perform the ordering by:

determining the equipment characteristic;

matching the determined equipment characteristic to a stored equipment characteristic; and

ordering the plurality of 2D perspective images based on an ordering arrangement corresponding to the stored equipment characteristic, wherein the ordering arrangement corresponds to the particular order.

19 . The system of claim 10 , wherein the control circuitry is further configured to:

order the plurality of 2D perspective images within the GOP in the particular order further based on equipment characteristics associated with equipment used to capture the plurality of 2D perspective images; and

order the plurality of 2D perspective images within the GOP in the particular order further based at least in part on a first weight and a second weight, wherein the first weight is associated with the prediction efficiencies for each 2D perspective image of the plurality of 2D perspective images and is assigned based at least in part on comparing the prediction efficiencies to a threshold, and wherein the second weight is associated with the equipment characteristics and is assigned based on how closely the equipment characteristics match stored equipment characteristics associated with historical GOP ordering data.