Systems and methods for image data compression
View Patent ↗Systems and methods for image data compression in an array camera are disclosed. An array camera includes a processor, an array of cameras, and a compression module. The compression module is configured to receive image data from the array of pixels of a focal plane in a sequentially linear group of pixel data, analyze the received image data based upon truncation rules, compress the received data image based upon the analysis to generate compressed image data, and generate a bit mask identifying truncated image data in the compressed image data.
1. A system comprising:
compression circuitry including at least one input via which the compression circuitry is configured to:
receive image data captured by a plurality of cameras having different viewpoints, wherein the image data is in sequentially linear groups of pixel data, and wherein the pixel data are from corresponding pixel locations in a plurality of different images captured by the plurality of cameras;
wherein the compression circuitry is further configured to:
identify corresponding pixel locations in sequentially linear groups of pixel data received from different cameras in the plurality of cameras;
identify redundant pixel data in at least one sequentially linear group of pixel data by determining similarity between the pixel data, wherein identifying whether a given pixel data in the sequentially linear groups of pixel data is redundant comprises:
calculating a multiple of the given pixel data in the sequentially linear groups of pixel data,
calculating a difference between the multiple of the given pixel data and at least one pixel data neighboring the given pixel data, and
when the calculated difference is less than a threshold value, identifying the given pixel data as a redundant pixel data;
truncate pixel data identified as redundant from the at least one sequentially linear group of pixel data to form compressed image data;
multiplex compressed image data with pixel data from the identified corresponding pixel locations in sequentially linear groups of pixel data received from different cameras in the plurality of cameras; and
transmit the multiplexed compressed image data and pixel data from the compression circuitry to a processing system via interface circuitry.
2. A method for compressing image data captured by a plurality of cameras comprising:
receiving image data captured by a plurality of cameras having different viewpoints using a compression circuitry, wherein the image data is in sequentially linear groups of pixel data, and wherein the pixel data are from corresponding pixel locations in a plurality of different images captured by the plurality of cameras;
identifying corresponding pixel locations in sequentially linear group of pixel data received from different cameras in the plurality of cameras using the compression circuitry, wherein identifying whether a given pixel data in the sequentially linear groups of pixel data is redundant comprises:
calculating a multiple of the given pixel data in the sequentially linear groups of pixel data,
calculating a difference between the multiple of the given pixel data and at least one pixel data neighboring the given pixel data, and
when the calculated difference is less than a threshold value, identifying the given pixel data as a redundant pixel data;
identifying redundant pixel data in at least one sequentially linear group of pixel data by determining similarity between the pixel data using the compression circuitry;
truncating pixel data identified as redundant from the at least one sequentially linear group of pixel data to form compressed image data using the compression circuitry;
multiplexing compressed image data with pixel data from the identified corresponding pixel locations in sequentially linear groups of pixel data received from different cameras in the plurality of cameras using the compression circuitry; and
transmitting the multiplexed compressed image data and pixel data from the compression circuitry to a processing system via interface circuitry.
3. The method of claim 2 , wherein the threshold value is a noise floor below which pixel data is indistinguishable from noise.
4. The method of claim 2 , wherein:
the given pixel data is pixel data N −1 ,
the multiple is 2,
the at least one pixel data neighboring the given pixel data include pixel data N and pixel data N −2 , and
the difference K is calculated according to the following truncation rule: K=(2*N −1 )−N−N −2 .
5. The method of claim 2 , wherein the multiplexed compressed image data and pixel data is transmitted in a stream of packets.
6. The method of claim 5 , wherein the stream of packets comprises a header describing how image data from the plurality of cameras is multiplexed within the stream of packets.
7. The method of claim 2 further comprising storing the received image data in a shift register prior to identifying redundant pixel data in at least one sequentially linear group of pixel data.
8. The system of claim 1 , wherein the threshold value is a noise floor below which pixel data is indistinguishable from noise.
9. The system of claim 1 , wherein:
given pixel data is pixel data N −1 ,
the multiple is 2,
the at least one pixel data neighboring the given pixel data include pixel data N and pixel data N −2 , and
the difference K is calculated according to the following truncation rule: K=(2*N −1 )−N−N −2 .
10. The system of claim 1 , wherein the compression circuitry is further configured to transmit the multiplexed compressed image data and pixel data in a stream of packets.
11. The system of claim 10 , wherein the stream of packets comprises a header describing how image data from the plurality of cameras is multiplexed within the stream of packets.
12. The system of claim 1 , wherein the compression circuitry is further configured to store the received image data in a shift register prior to identifying redundant pixel data in at least one sequentially linear group of pixel data.