Dynamic Decoded Picture Buffer Allocation for Video Resolution Transitions

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Advanced video coding standards like AVC/H.264 pose challenges in managing resources such as the decoded picture buffer due to changes in picture resolution, requiring efficient memory allocation and deallocation to maintain seamless video processing.

Innovation Solution

The method involves dynamically determining the maximum number of frame buffers in the Decoded Picture Buffer (DPB) based on the picture resolution, allowing for efficient memory allocation and reuse without the need for deallocation, by using auxiliary information to adjust the frame buffer sizes according to the maximum coding block size, enabling smooth transitions between different picture resolutions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the decoded picture buffer is allocated based on maximum picture resolution, then the buffer can store pictures at any resolution, but memory is wasted when processing lower resolution pictures

Engineering Contradiction:
Improvebuffer compatibilityVSAvoidmemory usage
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent applies dynamics by making the decoded picture buffer size adaptive rather than fixed. The buffer dynamically adjusts its allocation based on the actual picture resolution being processed, allowing it to scale between minimum and maximum sizes according to the coding video sequence requirements, thus eliminating memory waste while maintaining versatility

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the parameter of buffer size from a static maximum value to a dynamic value that varies with picture resolution. By deriving the number of frame buffers based on actual picture dimensions rather than maximum supported dimensions, the system optimizes memory usage while maintaining the ability to handle various resolutions

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If the decoded picture buffer is deallocated and reallocated during picture resolution changes, then memory is optimized for current resolution, but processing continuity is interrupted

Engineering Contradiction:
Improvememory efficiencyVSAvoidprocessing continuity
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent applies preliminary action by pre-allocating the decoded picture buffer based on the maximum expected picture resolution before processing begins. This initial allocation ensures that sufficient memory is available for any resolution changes that may occur during processing, eliminating the need for subsequent reallocation and maintaining processing continuity

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent makes the buffer universally applicable to multiple picture resolutions by allocating it based on maximum resolution requirements from the outset. This single buffer allocation serves multiple functions across different resolution scenarios, eliminating the need for separate allocations and deallocations when switching between resolutions

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Quantity of substance

If auxiliary information is used to derive frame buffer size, then memory allocation is optimized for actual picture resolution, but additional processing steps are required

Engineering Contradiction:
Improvememory allocationVSAvoidprocessing complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent applies self-service by having the decoding system automatically derive frame buffer size requirements from auxiliary information contained in the bitstream itself. The system uses the picture resolution data already present in the coded video sequence parameters to self-determine the appropriate buffer allocation without external intervention or complex manual configuration

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS9602817B2Management of decoded picture buffer at picture format transitions
Publication Date: 2017.03.21 CISCO TECHNOLOGY INC
  • US9602817B2 patent drawing
  • US9602817B2 patent drawing
  • US9602817B2 patent drawing

AI summary

In one method embodiment a method of processing of a bitstream is disclosed. The method may include receiving by a processing device a bitstream comprising a first portion of compressed pictures encoded at a first encoding level, the first portion of compressed pictures corresponding to a first picture resolution; determining a ratio of a picture size of at least one picture of the first portion of the compressed picture and a picture size of the largest picture encoded at the first encoding level; determining a first number of picture buffers for processing of the first portion of the compressed pictures, wherein the first number of picture buffers are determined based on the determined ratio; and mapping the first number of picture buffers in a memory allocated for decoding of bitstream.