Closed Captioning Signal Capture for TV Viewership Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional audience measurement systems for television viewership rely on self-reporting and subaudible audio watermarks, which are inaccurate and costly due to the need for encoders at every broadcast source, and are inefficient for short programs and data compression.
Innovation Solution
Utilizing closed captioning data encoded in EIA-608 and EIA-708 formats within the video signal, which is content-specific and can identify programs being watched, eliminating the need for additional encoders and improving accuracy by decoding and matching captioning text with a database.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If subaudible audio watermarks are used for audience measurement, then program identification can be achieved, but error rates are high and data collection efficiency is low
Solution Approach 1:
The patent replaces the acoustic watermark detection method with optical signal processing. Instead of listening for subaudible codes in the audio track, the system captures and analyzes the visual display of program information (such as closed captions, on-screen graphics, or other visual identifiers) on the television screen. This substitution of detection modality from acoustic to optical eliminates the fundamental limitations of audio watermarking while maintaining program identification capability.
Solution Approach 2:
The system creates a visual copy of the program identification information by capturing the display output. Rather than relying on embedded audio codes that must be inaudible and therefore difficult to detect accurately, the system copies the visual representation of program data that is already being displayed to the viewer. This copying approach provides higher fidelity data collection without requiring modifications to the broadcast signal.
2Reliability
If audio watermarks are embedded in broadcast signals, then program tracking is enabled, but encoders are required at every broadcast source increasing cost and complexity
Solution Approach 1:
The patent extracts the program identification function from the broadcast signal embedding process. Instead of requiring encoders to insert watermarks into the audio signal at each broadcast source, the system extracts identification information from the visual display that is already present in the broadcast content. This extraction approach eliminates the need for complex encoder infrastructure while maintaining reliable program tracking capability.
Solution Approach 2:
The broadcast signal itself provides the identification data through its visual components (closed captions, on-screen graphics, etc.) without requiring additional encoding equipment. The system uses resources already present in the broadcast - the visual information that is naturally displayed - to enable program tracking. This self-service approach eliminates the need for separate encoder hardware at broadcast facilities.
3Quantity of substance
If subaudible codes are used for measurement, then viewership data can be collected, but the device must listen for substantial duration reducing effectiveness for short programs
Solution Approach 1:
The system performs preliminary capture of visual identification information that is already displayed on screen during program playback. By capturing the visual output in real-time as it appears, the system eliminates the need for prolonged listening periods required by audio watermark detection. The identification data is available immediately in visual form, allowing rapid detection and measurement even for short programs.
Data Source
Figure 1A
Figure 1B
Figure 2A
AI summary
To provide audience measurement services without requiring broadcasters to install additional watermarking equipment, a small video capture device may receive a video signal viewed by an audience member, the signal including embedded closed captioning data, and may decode and transmit the closed captioning data to an audience measurement server. The measurement server may receive the data and compare the received data to a database of closed captioning text to identify matches. The database may be populated via one or more additional capture devices receiving local or national broadcast signals, or with data separately provided by broadcasters. When a match is found, the measurement server may determine that the audience member is watching a program having the corresponding closed captioning. Dynamically generated confidence scores may allow for reliable identification of audience members of content, even with very short closed captioning data lengths.