Automated Content Recognition Database for Real-Time TV Ad Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current automated systems for detecting and identifying TV commercials in real-time across multiple channels and markets are inefficient and costly, as they rely on manual processes and struggle to scale to handle hundreds of video streams while maintaining computational and storage costs within reasonable limits.

Innovation Solution

An automated content recognition database method that ingests audiovisual streams, determines candidate segments by analyzing frames for matches in known advertisements, and removes identified commercial segments from the primary content database, using a path pursuit algorithm to refine matches and reduce false positives.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If automated content recognition is used to detect commercials in real-time across multiple channels, then detection accuracy and automation level improve, but computational cost and storage requirements increase

Engineering Contradiction:
Improveautomation levelVSAvoidcomputational cost
Core Design Contradiction:
Extent of automationVSLoss of energy

Solution Approach 1:

The system segments video content by creating separate databases for program content and advertisement content. The content recognition database is divided into distinct data structures (program content database and advertisement database) that are maintained separately. This segmentation allows the system to process and match content more efficiently without requiring all computational resources to handle the entire video stream simultaneously, thus reducing overall computational cost while maintaining high automation levels.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary action by pre-processing and storing advertisement content in a dedicated advertisement database before actual commercial detection occurs. Advertisements are ingested, processed, and organized in advance with their metadata and characteristics stored for rapid matching. This preliminary preparation significantly reduces the computational burden during real-time detection, as the system only needs to match against pre-organized advertisement data rather than analyzing all content from scratch.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If real-time monitoring of hundreds of video streams is implemented, then commercial detection coverage improves, but system complexity and processing load increase

Engineering Contradiction:
Improvedetection coverageVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system handles multiple video streams by segmenting the processing task into separate database operations. Each stream's content is matched against the pre-existing program content database and advertisement database independently. This segmentation allows parallel processing of multiple streams without requiring complex inter-stream coordination, thereby managing system complexity while maintaining high detection coverage across hundreds of channels.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system uses copying by maintaining duplicate databases (program content database and advertisement database) that can be independently queried. When monitoring multiple streams, the system copies relevant content characteristics from incoming streams and matches them against the stored databases rather than processing entire video streams in real-time. This approach reduces processing load while maintaining comprehensive detection coverage across numerous channels.

Inventive Principle:
Principle #26Copying

3Measurement precision

If candidate segments are tested against known advertisements in a comprehensive database, then match accuracy improves, but processing time increases

Engineering Contradiction:
Improvematch accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system applies preliminary action by pre-organizing advertisement content in a structured database with metadata, tags, and hierarchical categorization before matching occurs. Advertisements are processed in advance to extract key characteristics and store them in optimized data structures. This preliminary organization enables rapid retrieval and comparison during candidate segment testing, maintaining high match accuracy while significantly reducing the time required to search through the advertisement database.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The matching process is segmented into multiple stages: first testing candidate segments against the program content database, then against the advertisement database only if needed. The system segments the search space and applies different matching thresholds at different stages. This segmented approach ensures high match accuracy through thorough testing while reducing overall processing time by eliminating unnecessary full-database searches for obvious non-matches.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3534615B1Systems and methods for real-time television ad detection using an automated content recognition database
Publication Date: 2024.01.17 INSCAPE DATA INC
  • EP3534615B1 patent drawingFigure 1
  • EP3534615B1 patent drawingFigure 2
  • EP3534615B1 patent drawingFigure 3

AI summary

A system and method for the automated real-time detection and processing of commercial messages, public service announcements or similar short-duration, repeated TV programing segments occurring in one or more broadcast video steams is described. A process is utilized that identifies possible commercial segments by identifying discrete video segments that have specific short durations among other attributes. Video segments that appear to have these characteristics are considered likely to be television commercials, promotions, or public service announcements and are stored in a TV Ad database which is separate from the primary television content database. Incoming video from a plurality of television programming sources is process into fingerprints and placed in a master TV content database. At the same time, said incoming programming is tested for matches in a TV ad database. If a match is found, then the presumed TV ad is removed from the master content database.