Hybrid Fixed-Rate Codec Switching Under Tight Bit Budgets

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing data compression technologies face challenges in achieving mathematically lossless performance while meeting bitrate constraints and minimizing memory usage, especially in applications like display panels where overdrive algorithms require efficient data handling.

Innovation Solution

A hybrid data compression codec system that switches between lossless and lossy codecs based on a bit budget, using differential pulse coded modulation (DPCM) codecs without quantization for lossless performance and with quantization for lossy compression, to encode data in a fixed-rate format without additional memory buffers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If a lossless codec is used to ensure mathematically lossless performance, then data accuracy is improved, but the bitrate increases beyond constraints

Engineering Contradiction:
Improvedata accuracyVSAvoidbitrate
Core Design Contradiction:
Loss of informationVSQuantity of substance

Solution Approach 1:

The system dynamically switches between lossless and lossy codecs based on real-time bit budget status. The rate controller monitors the bit budget and transitions from lossless DPCM to lossy DPCM with quantization when the bit budget is exhausted, enabling adaptive adjustment of compression aggressiveness to meet bitrate constraints while maximizing data accuracy when possible

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the quantization parameter of the DPCM codec to transition between lossless and lossy modes. When operating in lossy mode, quantization is applied to reduce the bitrate while maintaining acceptable data accuracy. This parameter change allows the same codec framework to operate across different compression regimes

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If a hybrid codec switching mechanism is implemented, then bitrate control is improved, but device complexity increases

Engineering Contradiction:
Improvebitrate controlVSAvoidcodec switching mechanism
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The encoding process is segmented into distinct phases: an initial lossless encoding phase followed by a lossy encoding phase if needed. This segmentation allows the system to guarantee lossless performance for representative data while providing a fallback mechanism for non-representative data that exceeds the bit budget, resolving the contradiction between bitrate control and complexity

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

A rate controller acts as an intermediary between the lossless and lossy codec paths. This mediator monitors the bit budget consumption and triggers the switch between encoding modes, simplifying the overall control logic compared to more complex adaptive systems while still achieving effective bitrate control

Inventive Principle:
Principle #24Intermediary (Mediator)

3Adaptability or versatility

If additional memory buffers are added to handle codec switching, then encoding flexibility is improved, but memory usage increases

Engineering Contradiction:
Improveencoding flexibilityVSAvoidmemory usage
Core Design Contradiction:
Adaptability or versatilityVSVolume of stationary object

Solution Approach 1:

The system processes data in a single pass through the encoding pipeline, with the rate controller making real-time decisions about codec selection based on bit budget status. This self-service approach eliminates the need for additional memory buffers to store intermediate results for later processing, maintaining encoding flexibility while minimizing memory usage

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3893504B1Systems and methods for low-complexity near lossless fixed-rate hybrid data compression codecs
Publication Date: 2024.03.13 SAMSUNG DISPLAY CO LTD
  • EP3893504B1 patent drawingFigure 1
  • EP3893504B1 patent drawingFigure 2A~2B
  • EP3893504B1 patent drawingFigure 3

AI summary

A method for encoding data includes: receiving a frame of data including subunits; initializing a remaining bit budget for the frame; setting a current codec to a lossless codec having a lossless codec bitrate; and sequentially encoding the subunits, including: encoding a first subunit using the lossless codec to compute a first encoded subunit; subtracting a length of the first encoded subunit from the remaining bit budget; determining whether the remaining bit budget exceeds a first lossy bits required to encode a plurality of remaining subunits using a first lossy codec having a first lossy codec bitrate lower than the lossless codec bitrate; in response to determining that the remaining bit budget is less than or equal to the first lossy bits required, setting the current codec to the first lossy codec; and encoding a second subunit using the first lossy codec to compute a second encoded subunit.