US 12,684,123 B2Granted

Method and system for energy-efficient approximate digital JPEG and MJPEG-compression

AssigneeQuasistatics Inc.

InventorsShreyas Sen, Archisman Ghosh

Life of this patent

Estimating the term…

Loading patent…

Abstract

A system and method for energy-efficient approximate digital JPEG and MJPEG-compression. The system includes a controller unit to control a processing loop for processing image blocks based on a comparison of a current image block to a previous image block. The system includes a quantization unit configured to quantize the frequency domain representation using an approximate quantization process and a quantization (Q) matrix. The quantization unit is configured to: identify, a nearest power of two value for each element of the quantization matrix; generate an updated Q matrix by assigning each element of the quantization matrix with the identified nearest power of two value; and shift each element of the updated Q matrix by a number of bits to generate a quantized frequency domain representation. The number of bits corresponds to the identified nearest power of two for the corresponding element of the updated Q matrix.

Claims20

Dependent claims

A system for energy-efficient approximate digital JPEG and MJPEG-compression samples, the system comprising:

a two-dimensional Discrete Cosine Transform (2D DCT) circuit configured to:
receive a plurality of input multimedia data from an image processing system;
segment the plurality of input multimedia data into a plurality of image blocks;
and
perform a two-dimensional discrete cosine transform (2D-DCT) on each image block to generate a corresponding frequency domain representation;
a controller unit configured to control a processing loop for processing the plurality of image blocks based on a comparison of a current image block to a previous image block, wherein processing of the current image block is skipped when the current image block satisfies a predetermined similarity criteria with respect to the previous image block;
a quantization unit configured to quantize the frequency domain representation using an approximate quantization process and a quantization (Q) matrix, wherein in the approximate quantization process, the quantization unit is configured to:
identify, a power of two value equal to 2s for each element of the quantization matrix, where s is a position of a most significant ‘1’ bit of the element;
generate an updated Q matrix by assigning each element of the quantization matrix with the identified power of two value;
and
shift each element of the updated Q matrix by a number of bits to generate a quantized frequency domain representation, wherein the number of bits corresponds to the identified power of two for the corresponding element of the updated Q matrix;
a zigzag traversing unit configured to serialize the quantized frequency domain representation to generate serialized samples;
a run length encoder unit configured to:
perform entropy coding on the serialized frequency domain representation;
and
generate compressed digitized samples for the coded frequency domain representation.

13 more elements

A method for compressing multimedia data into digital samples comprising:

receiving, by a processor, plurality of input multimedia data from an image processing system;
segmenting, by the processor, the plurality of input multimedia data into a plurality of image blocks;
performing, by the processor, a two-dimensional discrete cosine transform (2D-DCT) on each image block to generate a corresponding frequency domain representation;
controlling, by the processor, a processing loop for processing the plurality of image blocks based on a comparison of a current image block to a previous image block, wherein processing of the current image block is skipped when the current image block satisfies a predetermined similarity criteria with respect to the previous image block;
quantizing, by the processor, the frequency domain representation using an approximate quantization process and a quantization (Q) matrix, the approximate quantization process comprising:
identifying, by the processor, a power of two value equal to 2s for each element of the quantization matrix, where s is a position of a most significant ‘1’ bit of the element;
generating, by the processor, an updated Q matrix by assigning each element of the quantization matrix with the identified power of two value;
and
shifting, by the processor, each element of the updated Q matrix by a number of bits to generate a quantized frequency domain representation, wherein the number of bits corresponds to the identified power of two for the corresponding element of the updated Q matrix;
serializing, by the processor, the quantized frequency domain representation to generate serialized samples;
performing, by the processor, entropy coding on the serialized frequency domain representation;
and
generating, by the processor, compressed digitized samples for the coded frequency domain representation.

10 more elements

Dependencies are inferred from the claim text.

Facts

Application: 18/641,469Filed 22 Apr 2024 · us provisional application
Priority: 23 Apr 2023
Published: US 2025/0350725 A113 Nov 2025
Granted: 14 Jul 2026
Term ends: Not estimatedThe source records don’t support an estimate.

Classification7 codes

H04N 19/124
Characterised by the element, parameter or selection affected or controlled by the adaptive coding: quantisation
H04N 19/176
The unit being an image region, e.g. an object: the region being a block, e.g. a macroblock
H04N 19/18
Characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the…: the…
H04N 19/60
Methods or arrangements for coding, decoding, compressing or decompressing digital video signals: using transform coding