System and method for processing three dimensional images
A system and method for lightfield communication can be configured to receive one or more images, determine one or more features in the one or more images, and generate a lightfield image from the one or more images where at least one of the one or more features is in a focal plane of the lightfield image.
1. A lightfield communication system comprising:
a plurality of cameras configured to capture a plurality of images of a subject contemporaneously, a field of view of each camera of the plurality comprising a distinct perspective of the subject;
a feature module configured to identify a common feature in each of the plurality of images;
an alignment module configured to align each of the plurality of images based on the common feature; and
a display comprising:
a light source;
a lenticular lens optically coupled to the light source that, with the light source, generates a light output having viewing angle dependency; and
an optical volume optically coupled to the lenticular lens.
2. The lightfield communication system of claim 1 , wherein the common feature comprises a face region of the subject.
3. The lightfield communication system of claim 1 , wherein the alignment module aligns the plurality of images without a depth of an image of the plurality of images.
4. The lightfield communication system of claim 1 , wherein the alignment module is further configured to: determine a crop region for each image of the plurality of images associated with the common feature; and crop each image to the associated crop region.
5. The lightfield communication system of claim 1 , wherein the alignment module is configured to align each of the plurality of images using an inverse homography matrix transformation.
6. The lightfield communication module of claim 1 , further comprising a communication module configured to transmit the plurality of aligned images to a second display, wherein the second display is configured to display the plurality of aligned images as a lightfield image.
7. The lightfield communication system of claim 6 , wherein the second display is configured to display the lightfield image without generating a render using the plurality of aligned images.
8. The lightfield communication system of claim 6 , wherein the second display comprises a tracking sensor configured to determine a gaze of a viewer, wherein the common feature is further identified based on the gaze of the viewer.
9. The lightfield communication system of claim 6 , wherein the feature module and alignment module are collocated with the plurality of cameras.
10. The lightfield communication system of claim 1 , wherein the common feature is identified and the plurality of images are aligned within at most 17 ms.
11. The lightfield communication system of claim 1 , wherein the feature module is configured to identify the common feature using machine learning techniques.
12. The lightfield communication system of claim 1 , wherein the plurality of images comprise a second subject, wherein the feature module is configured to detect a second common feature associated with the second subject, wherein the alignment module is further configured to determine an alignment based on the common feature and the second common feature.
13. The lightfield communication system of claim 12 , further comprising a sensor configured to detect an action of at least one of the first or second subject, wherein the alignment is determined based on the detected action.
14. A method for generating a lightfield image comprising:
receiving a plurality of images, each image of the plurality of images capturing different perspectives of a shared scene;
determining a shared feature in each image of the plurality of images;
setting the shared feature to an approximately zero disparity position between each image of the plurality of images; and
displaying the plurality of images as a lightfield image in a display, wherein the lightfield image is perceived as three dimensional by one or more viewers, wherein the shared feature is in a focal plane of the display.
15. The method of claim 14 , wherein the shared feature comprises a face region of a subject in the shared scene.
16. The method of claim 14 , wherein the lightfield image is perceivable as three dimensional by the one or more viewers without using headgear.
17. The method of claim 14 , wherein the shared feature is set to approximately zero disparity position without determining a depth of the shared feature in any of the plurality of images.
18. The method of claim 17 , wherein setting the shared feature to the approximately zero disparity position comprises, for each image of the plurality of images:
determining a bounding box surrounding the shared featured, wherein the shared feature is centered within the bounding box; and
cropping the image, wherein the cropped image only includes information within the bounding box.
19. The method of claim 14 , wherein determining the shared feature comprises using machine learning to determine the shared feature.
20. The method of claim 19 , further comprising tracking a gaze of the one or more viewers, wherein the gaze is used to determine the shared feature.
21. The method of claim 14 , further comprising tracking a subject in the shared scene, wherein determining the shared feature comprises using a tracking information to determine the shared feature.
22. The method of claim 14 , wherein determining the shared feature and setting the shared feature to the approximately zero disparity position is performed within 17 ms.