Multi-focus image data compression
Systems and methods for converting multi-focus image data into a compressed video format. Multi-focus image data of a scene may be sampled to generate a stack of images spanning multiple focal depths. Each image focal distance is mapped to a time stamp and a video encoder is employed to generate a compressed video stream of scene information. In further embodiments, the video encoded scene information is decoded and frames from the stream selected for display based on the focal distance-time stamp mapping. In further embodiments, multi-focus video data is encoded with consecutive frames spanning different focal distances defining a Group of Pictures (GOP), and/or spanning different hierarchical layers.
1. A multi-focus image data encoder, comprising:
a processor to:
render a plurality of image frames, each having a unique image focal distance, from light field image data collected for a single scene;
order the plurality of image frames into a series of frames having monotonically varying focal distances;
map each image frame in the focal-distance ordered series to a monotonically varying time stamp indicative of a time-sequential image frame series;
encode the plurality of image frames as a time sequential series of an encoded video based on the image frame time stamps, wherein the processor is to encode a first image frame in the ordered series as an intra-coded frame, and encode a second frame in the ordered series as an inter-predicted frame based, at least in part, on the intra-coded frame; and
an electronic memory to store the encoded video.
2. The encoder of claim 1 , wherein the processor is to implement a codec compliant with one or more of: H.262, H.263, H.264, or high efficiency video coding (HEVC) specifications.
3. The encoder of claim 1 , wherein the processor is to encode the first image frame associated with a first focal distance to a base layer of a multi-layered media bitstream, and to encode the second image frame associated with a second focal distance to an enhancement layer of the multi-layered media bitstream.
4. The encoder of claim 1 , wherein the processor is to:
segment each of the image frames into bounding boxes;
determine an edge detection score for each segment; and
associate each segment with a focal distance based on the edge detection score.
5. A computer-implemented method of encoding multi-focus image data, the method comprising:
rendering a plurality of image frames, each having a unique image focal distance, from light field image data collected for a single scene;
ordering the plurality of image frames into a series of frames having monotonically varying focal distances;
mapping each image frame in the focal-distance ordered series to a monotonically varying time stamp indicative of a time-sequential image frame series;
encoding the plurality of image frames as a time sequential series of an encoded video based on the image frame time stamps, wherein the encoding comprises encoding a first image frame in the ordered series as an intra-coded frame, and encoding one or more second frame in the ordered series as an inter-predicted frame based, at least in part, on the intra-coded frame; and
storing the encoded video to an electronic memory.
6. The method of claim 5 , wherein the encoding video further comprises a codec compliant with one or more of: H.262, H.263, H.264, or high efficiency video coding (HEVC) specifications.
7. The method of claim 5 , wherein the encoding video further comprises:
encoding the first image frame associated with a first focal distance to a base layer of a multi-layered media bitstrearn; and
encoding the second image frame associated with a second focal distance to an enhancement layer of the multi-layered media bitstream.
8. The method of claim 1 , further comprising:
segmenting each of the image frames into bounding boxes;
determining an edge detection score for each segment; and
associating each segment with a focal distance based on the edge detection score.