Pulse and Thermal Person Detection for Privacy-Safe Audience Counting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audience measurement systems inaccurately count people in media exposure environments due to false positives from non-human objects being mistaken for humans, leading to inflated or deflated viewership data, and raise privacy concerns with image data processing.
Innovation Solution
Collect light information from horizontally arranged segments of an environment using an array of light sensors to detect pulse-induced brightness patterns, verifying with thermal imaging to ensure accurate person detection without identifying individuals, thus maintaining privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image data processing is used to detect people, then person detection capability is achieved, but false positives occur from non-human objects being mistaken for humans
Solution Approach 1:
The patent segments the detection process into multiple independent analysis stages: initial object detection, human-specific feature verification, and cross-modal validation. By dividing the detection pipeline into discrete segments with specialized functions, the system can identify and eliminate false positives at each stage rather than relying on a single imperfect detection mechanism.
Solution Approach 2:
The patent introduces thermal imaging as an intermediary verification layer between the initial image detection and the final person count. Thermal data serves as a mediator that confirms whether detected objects are truly human, since thermal signatures provide additional biological verification that distinguishes humans from non-human objects in the visual data.
2Measurement precision
If image data processing is used to detect people, then person detection is performed, but privacy concerns are raised
Solution Approach 1:
The patent extracts only the essential detection information (presence, count, and basic location of people) from the full image data, discarding all identifying visual details. By separating the detection function from the visual identification function, the system achieves person detection without capturing or storing privacy-sensitive information such as faces, clothing details, or body postures that could identify individuals.
Solution Approach 2:
The patent replaces the mechanical/visual inspection method with a thermal sensing approach for verification. Thermal imaging substitutes the need for detailed visual analysis, allowing the system to confirm human presence through thermal signatures alone, thereby eliminating the requirement to process and store identifying visual data.
3Measurement precision
If traditional image processing methods are used, then people detection is attempted, but computational costs are high
Solution Approach 1:
The patent applies partial action by using thermal imaging selectively only for verification of detected objects, rather than processing complete high-resolution images for all analysis. This partial approach processes only the minimal necessary data (thermal signatures of already-detected objects) to confirm humanity, significantly reducing computational load compared to full image analysis while maintaining detection accuracy.
Solution Approach 2:
The patent changes the detection parameter from detailed visual image data to thermal radiation data. This parameter transformation allows verification of human presence using a different physical modality that requires less computational processing for the same verification task, as thermal pattern recognition is computationally simpler than detailed visual feature analysis.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Accurately counts people in media exposure environments efficiently and privately, reducing computational costs and improving the accuracy of media exposure data generation.
Implementation Method 1
Collect light information from horizontally arranged segments of an environment using an array of light sensors to detect pulse-induced brightness patterns
Implementation Method 2
verifying with thermal imaging to ensure accurate person detection without identifying individuals, thus maintaining privacy
Data Source
AI summary
Example apparatus disclosed herein includes a grouper to generate a first presence indication based on a threshold number of segments of image data including patterns of light values corresponding to a first human pulse. Disclosed example apparatus also include a face identifier to identify a first location associated with a human face in an environment associated with the image data based on thermal imaging data from a thermal imaging device, and output the first location associated with the human face in the environment associated with the image data.