Frame-Accurate Media Segment Extraction Using Fingerprint Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for identifying and extracting segments of interest in media streams, such as repeated video content, are inefficient and require significant processing and storage resources, especially when dealing with long or infinitely long streams.

Innovation Solution

A method using hashed fingerprint techniques to generate unique fingerprints for each frame in a media stream, representing a window of time, which reduces processing and storage needs by comparing and matching these fingerprints to identify and extract repeating segments accurately.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional methods are used to identify and extract segments of interest in media streams, then segmentation accuracy can be achieved, but processing and storage resource requirements become excessively high

Engineering Contradiction:
Improvesegmentation accuracyVSAvoidprocessing and storage resources
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent divides the media stream into discrete frames and generates individual fingerprints for each frame. This segmentation allows the system to process and compare only essential frame-level data rather than entire media streams, dramatically reducing storage and processing requirements while maintaining the ability to accurately identify repeating segments through frame-level fingerprint matching

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates fingerprint representations (copies) of media frames that capture essential identifying features. These fingerprints serve as simplified substitutes for the actual media content, enabling accurate segment identification through fingerprint comparison without requiring storage or processing of the full-resolution media streams

Inventive Principle:
Principle #26Copying

2Reliability

If traditional methods are used to analyze long or infinitely long media streams, then complete segment identification is possible, but computational requirements become unsustainable

Engineering Contradiction:
Improvecomplete segment identificationVSAvoidcomputational requirements
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent pre-generates fingerprints for all frames in the media stream and stores them in a database before actual segment identification is needed. This preliminary action allows the system to quickly retrieve and compare pre-processed fingerprint data when analyzing long or infinite streams, avoiding the need for computationally intensive real-time processing of entire media streams

Inventive Principle:
Principle #10Preliminary action

3Manufacturing precision

If frame-accurate segment extraction is implemented, then precise identification of repeating content is achieved, but processing complexity increases significantly

Engineering Contradiction:
Improveframe accuracyVSAvoidprocessing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex mechanical or algorithmic media analysis systems with a fingerprint-based matching system. By substituting direct media content comparison with fingerprint hash comparison, the system achieves frame-accurate segment identification with significantly reduced processing complexity, as fingerprint matching is computationally simpler than full media stream analysis

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3844967B1Frame-accurate automated cutting of media content by using multiple airings
Publication Date: 2023.01.25 GOOGLE LLC
  • EP3844967B1 patent drawingFigure 1
  • EP3844967B1 patent drawingFigure 2
  • EP3844967B1 patent drawingFigure 3

AI summary

At least one aspect of the present disclosure is directed to systems and methods of extracting media segments based on fingerprint matching. The method can include receiving a media stream comprising a plurality of frames and generating a plurality of fingerprints corresponding to each frames. The method can receive a target timestamp and determine a target fingerprint from the plurality that corresponds to the target timestamp. The method can retrieve candidate fingerprints, each of the candidate fingerprints corresponding to a frame in a candidate media stream. The method can compare the target fingerprint to the candidate fingerprints to determine a matching candidate fingerprint. The method can match fingerprints that correspond to media frames before and after the target fingerprint to determine the upper and lower bounds of a segment of interest. The method can extract the segment of interest based on the bounds and provide it to a respective party.