Video Encoding Bit Rate Control via Preliminary Statistical Multiplexing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In video encoding systems, steep changes in image feature values make it difficult for bit rate control to follow real-time encoding timing, leading to potential image quality deterioration and the need for delay circuits.

Innovation Solution

A video encoding system that includes a program bank device for accumulating and decoding video/voice data, extracting image feature values, and notifying a statistical multiplexing controller in advance to calculate and output video encoding bit rates to delivery encoders, allowing bit rates to follow changes without increasing delay time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If statistical multiplexing control is performed using a two-pass real-time encoder with a delay circuit, then the bit rate control can follow steep changes in image feature values, but the encoding delay increases

Engineering Contradiction:
Improvebit rate control accuracyVSAvoidencoding delay
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs statistical multiplexing control in advance by analyzing image feature values before encoding. The control results are stored and reused during actual encoding, eliminating the need for delay circuits and two-pass encoding while maintaining accurate bit rate control that can follow steep changes in image content.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If feedback control is used to calculate video encoding bit rates based on image feature values, then image quality can be maintained, but the system cannot respond to steep changes in real-time encoding

Engineering Contradiction:
Improveimage qualityVSAvoidresponse speed
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

The system calculates optimal video encoding bit rates in advance by analyzing image feature values such as activity and complexity. These pre-calculated bit rate control values are stored and applied during real-time encoding, enabling the system to maintain high image quality while responding instantly to changes in image content without feedback delay.

Inventive Principle:
Principle #10Preliminary action

3Productivity

If delay circuits are used to enable statistical multiplexing control, then bit rate allocation can be optimized, but the real-time encoding capability is reduced

Engineering Contradiction:
Improvebit rate allocation efficiencyVSAvoidencoding delay time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system performs statistical multiplexing analysis and bit rate allocation in advance by processing image feature values before encoding. The results are stored in a database and retrieved during encoding, achieving optimal bit rate allocation for multiple programs without requiring delay circuits, thereby maintaining real-time encoding capability.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS9392289B2Video encoding system, program bank device, and video encoding method
Publication Date: 2016.07.12 NEC CORP
  • US9392289B2 patent drawing
  • US9392289B2 patent drawing
  • US9392289B2 patent drawing

AI summary

Encoders of a program bank device perform compression coding on video/voice data of a program from program VTRs, and extract an image feature value of video data of the program to be accumulated in a program database. When a time earlier by a predetermined time than a program delivery time indicated by program operation information from a program operation management server is reached, the program database notifies a statistical multiplexing controller of an image feature value and preliminarily calculates a video encoding bit rate by statistical multiplexing analysis processing. When the program delivery time is reached, the video/voice data of the program from the program database is decoded by decoders and output to delivery encoders at a timing when the data is output to the delivery encoders.