Emulating Human Light Sensitivity in Body Camera Video
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advances in image sensors in police body cameras capture events in low light conditions with greater sensitivity than the human visual system, leading to discrepancies between recorded video and actual perceptions, which complicates evaluations based on these recordings.
Innovation Solution
A system comprising a configuration system, recording device, and media playback system that encodes and adjusts video recordings to emulate human light sensitivity by generating and applying camera metadata, including exposure settings and gain values, to approximate the visual experience of a typical human during playback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If image sensors with higher light sensitivity are used in police body cameras, then video recording quality in low light conditions is improved, but the discrepancy between recorded video and human visual perception increases
Solution Approach 1:
The system pre-calculates and stores metadata including exposure settings and gain values during video recording. This preliminary action enables later adjustment of the recorded video to match human visual perception by applying the stored metadata information during playback processing.
Solution Approach 2:
The system changes the parameters of the recorded video by adjusting exposure settings and gain values based on the stored metadata. This parameter adjustment transforms the high-sensitivity camera footage into a representation that matches human visual perception, resolving the contradiction between recording quality and perceptual accuracy.
2Measurement precision
If video recordings are adjusted to match human visual perception, then evaluation accuracy is improved, but processing complexity increases
Solution Approach 1:
The system performs the complex processing work in advance by storing exposure settings and gain values as metadata during recording. This preliminary action reduces playback processing complexity because the adjustment parameters are already calculated and stored, requiring only application during playback rather than complex real-time processing.
Solution Approach 2:
The system introduces metadata as an intermediary layer between the raw video recording and the final adjusted output. This metadata contains pre-calculated exposure and gain information that mediates the transformation process, simplifying the complexity by providing ready-made adjustment parameters rather than requiring complex processing algorithms during playback.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In an embodiment, a method of emulating light sensitivity of a target includes, for each of at least some frames of a video recording, receiving an image. The method also includes accessing image metadata associated with the image. The method also includes discovering, from the image metadata, information related to an exposure setting and a gain value used by a camera in capture of the image. The method also includes deriving, from stored camera metadata for the camera, information related to a luminous flux associated with the exposure setting and the gain value used by the camera in the capture of the image. The method also includes determining an adjusted gain value corresponding to a target light sensitivity using the derived information related to the luminous flux. The method also includes generating an adjusted image using the adjusted gain value.