Adaptive Video Encoding via Confidence Values

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video processing systems face challenges in efficiently reducing invalid data storage, leading to high costs and reduced system adaptability due to misjudgment and missing judgment in motion detection and intelligent analysis technologies.

Innovation Solution

A method and device that generate a confidence value for video frames to determine a target encoding frame rate, allowing for adaptive encoding and recording based on a mapping relationship curve, thereby minimizing the impact of misjudgment and improving system adaptability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If motion detection or intelligent analysis technologies are used to determine whether to encode and record video, then storage space is reduced, but misjudgment and missing judgment occur leading to loss of key data and reduced system reliability

Engineering Contradiction:
Improvestorage spaceVSAvoidsystem reliability
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent changes the parameter from binary judgment (yes/no) to continuous confidence value judgment. By calculating confidence values and comparing them against dynamically adjusted thresholds, the system achieves more nuanced decision-making that reduces both misjudgment and missing judgment while optimizing storage space utilization.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces dynamic threshold adjustment based on confidence value distributions. The threshold is not fixed but adapts according to the calculated confidence values, allowing the system to dynamically balance between reducing storage and maintaining reliability under different scene conditions.

Inventive Principle:
Principle #15Dynamics

2Device complexity

If binary judgment (yes/no) is used to determine encoding, then processing is simple, but adaptive capability is lowered due to misjudgment and missing judgment

Engineering Contradiction:
Improveprocessing complexityVSAvoidadaptive capability
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The system transforms the simple binary parameter into a continuous confidence value parameter, enabling fine-grained control over encoding decisions. This parameter change allows the system to adapt to varying scene conditions while maintaining reasonable processing complexity through efficient confidence calculation and threshold comparison.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent implements feedback mechanisms where confidence values are continuously calculated and used to adjust encoding decisions. The system learns from the confidence distribution and adapts thresholds accordingly, creating a feedback loop that enhances adaptive capability without significantly increasing processing complexity.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2950529B1Video image storage method and apparatus
Publication Date: 2020.05.27 HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO LTD
  • EP2950529B1 patent drawingFigure 1~2
  • EP2950529B1 patent drawingFigure 3

AI summary

The present disclosure provides a method and device for storing a video image. The method includes the follows. For each frame of image collected, a confidence value is generated after passing M frames of image each time, in which M is a positive integer. A target encoding frame rate may be determined based on the confidence value latest generated, after passing N frames of image each time. The N frames of image may be encoded and recorded, based on the target encoding frame rate determined, in which N is a positive integer. By adopting the technical solutions of the present disclosure, adaptive capability and availability of system may be improved.