Multimedia Duplicate Detection via Block Hashing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current service systems face challenges in reliably identifying and reducing duplicates of multimedia data items, which affects the reliability of multimedia content provided to users.

Innovation Solution

A method involving maintaining service system hash values, hashing received multimedia data items, and searching for matches within these values to determine duplicates, with approved items being stored and rejected items triggering further evaluation or rejection messages.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multimedia data items are uploaded to the service system without duplicate checking, then the upload process is fast and simple, but duplicate content accumulates and reduces system reliability

Engineering Contradiction:
Improvereliability of multimedia contentVSAvoidcomplexity of duplicate detection system
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The multimedia data is segmented into fixed-size blocks, and hash values are computed for each block individually. This segmentation allows the system to efficiently compare only relevant portions of data without processing entire large files, reducing computational complexity while maintaining duplicate detection accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent replaces direct comparison of entire multimedia files with a computational hashing mechanism. Instead of mechanically comparing byte-by-byte across large data items, the system uses hash function computations to generate compact representations that can be quickly compared, significantly reducing processing complexity.

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

2Measurement precision

If hash values for all multimedia data items are maintained and compared, then duplicate detection accuracy is high, but processing time and computational resources increase

Engineering Contradiction:
Improveaccuracy of duplicate identificationVSAvoidtime for hashing and searching
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

By dividing multimedia data into blocks and computing hashes for each block separately, the system achieves accurate duplicate detection through block-level comparison while reducing the overall computational burden compared to hashing entire large files. This segmentation enables parallel processing and faster search operations.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter of data representation from complete multimedia files to condensed hash values. This parameter transformation maintains the ability to identify duplicates accurately while dramatically reducing the time and resources required for comparison and search operations.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If block size for hashing is small, then duplicate detection precision is high, but processing overhead increases

Engineering Contradiction:
Improveprecision of duplicate detectionVSAvoidprocessing efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system optimizes the block size parameter to achieve an optimal balance between detection precision and processing efficiency. By carefully selecting the block size, the patent ensures sufficient granularity for accurate duplicate identification while maintaining reasonable processing throughput and minimizing overhead.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2829070B1Method and apparatus for reducing duplicates of multimedia data items in service system
Publication Date: 2017.07.19 P2S MEDIA GROUP OY
  • EP2829070B1 patent drawingFigure 1
  • EP2829070B1 patent drawingFigure 2~3
  • EP2829070B1 patent drawingFigure 4~5

AI summary

A method reducing duplicates of multimedia data items in a service system, wherein the method comprises maintaining service system hash values for the multimedia data items of the service system; receiving a first multimedia data item; and hashing the received multimedia data item to provide a first hash value. The method further comprises searching the first hash value from the service system hash values; and approving the received multimedia data item to the service system in response to the first hash value being not found when searching from the service system hash values.