IP Library Granted Patent US 12,101,492
Granted Patent B2
US 12,101,492 · App. 17/355,842 · Granted Sep 24, 2024

Unprocessed image coding and decoding

Inventors: Bin Chen (San Jose, CA); Donghui Wu (San Mateo, CA)
H04N19/186H04N19/119H04N19/124H04N19/176H04N19/18H04N19/625H04N19/63H04N19/64H04N19/91
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,101,492
App. No.
17/355,842
Granted
Sep 24, 2024
Kind
B2
Abstract

A method of image compression, including, receiving at least one unprocessed image frame, transforming a domain of the at least one unprocessed image frame to output a transformed domain dataset, block processing the transformed domain dataset to yield a blocked dataset, quantizing the blocked dataset to produce a quantized dataset and entropy encoding the quantized dataset to construct at least one compressed image frame.

Claims (28)

1. A method of image compression, comprising:

receiving at least one unprocessed image frame, wherein the unprocessed image frame is at least one of unprocessed and minimally processed;

transforming a domain of the at least one unprocessed image frame to output a transformed domain dataset wherein the unprocessed image frame is sampled in a Bayer color filter array format and the transformed domain dataset is in a domain having a luminance channel, a blue-green difference channel, a red-green difference channel, and a first green to second green difference channel:

block processing the transformed domain dataset to yield a blocked dataset, wherein the block processing utilizes at least one of discrete cosine transform and wavelet transform;

quantizing the blocked dataset to produce a quantized dataset; and

entropy encoding the quantized dataset to construct at least one compressed image frame.

2. The method of image compression of claim 1 , wherein the block processing divides the transformed domain dataset into a set of regular blocks.

3. The method of image compression of claim 1 , further comprising reorganizing a set of transform coefficients of the block processing.

4. The method of image compression of claim 1 , further comprising retransforming a set of transform coefficients of the block processing.

5. The method of image compression of claim 1 , wherein a quantization factor applied for quantization is based on a set of coefficient frequencies.

6. The method of image compression of claim 1 , wherein a quantization factor applied for quantization is based on a coefficient criticality.

7. The method of image compression of claim 1 , wherein the entropy encoding is processed by at least one of Huffman coding and Arithmetic coding.

8. The method of image compression of claim 1 , wherein the transforming of the at least one unprocessed image frame is a lossless transformation.

9. The method of image compression of claim 1 , wherein the at least one unprocessed image frame is a video frame.

10. A method of image decompression, comprising:

receiving at least one compressed image frame;

entropy decoding the at least one compressed image frame to construct an entropy decoded dataset;

de-quantizing the entropy decoded dataset to produce a de-quantized dataset;

inverse block processing the de-quantized dataset to yield produce an inverse blocked dataset, wherein the inverse block processing utilizes at least one of inverse discrete cosine transform and inverse wavelet transform; and

inverse transforming the inverse blocked dataset to output at least one unprocessed image frame, wherein the uncompressed image frame is in a Bayer color filter array format and the inverse blocked dataset is in a domain having a luminance channel, a blue-green difference channel, a red-green difference channel, and a first green to second green difference channel.

11. The method of image decompression of claim 10 , wherein the entropy decoding is processed by at least one of Huffman decoding and Arithmetic decoding.

12. The method of image decompression of claim 10 , wherein a de-quantization factor applied to de-quantization based on a coefficient criticality.

13. The method of image decompression of claim 10 , wherein a de-quantization factor applied to de-quantization based on a set of coefficient frequencies.

14. The method of image decompression of claim 10 , further comprising retransforming a set of inverse transform coefficients of the inverse block processing.

15. The method of image decompression of claim 10 , further comprising reorganizing a set of inverse transform coefficients of the inverse block processing.

16. The method of image decompression of claim 10 , wherein the inverse transformation includes at least three color channels.

17. The method of image decompression of claim 10 , wherein the inverse transforming of the at least one compressed image frame is a lossless inverse transformation.

18. The method of image decompression of claim 10 , wherein the at least one compressed image frame is a video frame.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Dec 3, 2021
From: BLACK SESAME INTERNATIONAL HOLDING LIMITED
To: BLACK SESAME TECHNOLOGIES INC.
Reel/Frame 058302/0860 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 26, 2021
From: CHEN, BIN; WU, DONGHUI
To: BLACK SESAME INTERNATIONAL HOLDING LIMITED
Reel/Frame 058641/0364 →
Continuity (1)
Related Publication 20220417537A1 · Dec 29, 2022