Intra Prediction Mode Encoding for Video Compression

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video encoding technologies, such as H.265/HEVC and the next-generation Versatile Video Codec (VVC), face inefficiencies in reducing the code amount for intra prediction, particularly when estimating the occurrence probability of binary signals, leading to insufficient compression in encoded streams.

Innovation Solution

The encoding apparatus converts intra prediction mode numbers from 87 to 67 and further reduces them to 64 bits, generating a binary signal by binarizing the remaining value and calculating a predicted value for the most significant bit, which is then used to increase the probability of the most significant bit in arithmetic encoding, thereby reducing the code amount in the encoded stream.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of substance

If intra prediction mode numbers are reduced from 87 to 67 and then to 64 bits, then the code amount in the encoded stream is reduced, but the precision in representing reference directions is degraded

Engineering Contradiction:
Improvecode amountVSAvoidreference direction precision
Core Design Contradiction:
Loss of substanceVSMeasurement precision

Solution Approach 1:

The patent segments the 6-bit intra prediction mode number into two parts: a 2-bit most significant bit (MSB) and a 4-bit least significant bit (LSB). The MSB is used to determine the reference direction with higher probability, while the LSB provides additional precision. This segmentation allows the system to achieve good compression by focusing probability modeling on the MSB while maintaining adequate directional precision through the combined use of both parts.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter representation by converting the 6-bit intra prediction mode number into a binary signal where the MSB is explicitly modeled with enhanced probability. The system transforms the original uniform 6-bit representation into a structured format where the MSB can be predicted with higher probability based on neighboring block modes, thereby improving compression efficiency without losing essential directional information.

Inventive Principle:
Principle #35Parameter changes

2Productivity

If arithmetic encoding is used with enhanced probability for the most significant bit, then compression efficiency is improved, but the complexity of the encoding process increases

Engineering Contradiction:
Improvecompression efficiencyVSAvoidencoding process complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-determining the probability of the MSB based on the intra prediction modes of neighboring blocks before performing arithmetic encoding. The encoder calculates the probability distribution of the MSB in advance using the neighboring block information, which allows the arithmetic encoder to operate more efficiently with known probability estimates, thereby improving compression without requiring overly complex real-time probability calculations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary probability model that mediates between the 6-bit intra prediction mode number and the final binary arithmetic encoding. This probability model serves as an intermediary layer that enhances the MSB probability based on neighboring block patterns, allowing the arithmetic encoder to achieve better compression ratios without directly processing the full 6-bit value, thus managing complexity effectively.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS11381810B2Encoding apparatus, encoding method, and decoding apparatus
Publication Date: 2022.07.05 FUJITSU LTD
  • US11381810B2 patent drawing
  • US11381810B2 patent drawing
  • US11381810B2 patent drawing

AI summary

An encoding apparatus includes: a memory; and a processor coupled to the memory and configured to: generate a value corresponding to a reference direction selected in intra prediction for a process target block by converting a value indicating the reference direction in accordance with a shape of the process target block; calculate a predicted value in the process target block from the values corresponding to the reference directions generated in intra prediction for neighboring blocks adjacent to the process target block; and generate a binary signal for arithmetic encoding based on the predicted value and the value corresponding to the reference direction generated in the intra prediction for the process target block.