Camera Exposure Control via Region of Interest Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video cameras lack sufficient dynamic range, leading to underexposed or overexposed images of objects of interest, which can result in poor image quality.
Innovation Solution
A system and method for controlling automatic gain and exposure in cameras by determining a region of interest using sound source location, multi-person detection, and active speaker detection, and adjusting the image mean based on backlight weight regions to maintain optimal exposure and gain within a predetermined threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If a mask is used to determine the location of the object of interest and adjust exposure based on the mask area, then the object of interest can be properly exposed, but the overall image quality deteriorates when the scene has high dynamic range
Solution Approach 1:
The patent applies local quality by creating a mask that selectively weights different regions of the image. The mask assigns higher weights to regions containing objects of interest (detected via face detection, motion detection, or audio-visual correlation) and lower weights to background regions. This allows the exposure adjustment to be locally optimized for important subjects while being less influenced by high-contrast background elements, thereby maintaining both subject exposure quality and overall image consistency.
2Adaptability or versatility
If the exposure is adjusted to capture the entire scene, then the dynamic range is improved, but the object of interest becomes underexposed or overexposed
Solution Approach 1:
The patent implements feedback by continuously monitoring the image content through multiple detection mechanisms (face detection, motion detection, audio-visual correlation) and dynamically adjusting the mask and exposure settings based on detected objects of interest. The system calculates a weighted average luminance using the mask, compares it to target exposure levels, and iteratively adjusts exposure parameters to maintain optimal subject exposure while adapting to varying scene conditions, thus resolving the contradiction between dynamic range coverage and subject exposure quality.
3Productivity
If automatic exposure control is used without region of interest detection, then the processing speed is maintained, but the image quality deteriorates due to incorrect exposure of objects of interest
Solution Approach 1:
The patent applies preliminary action by performing object detection (face detection, motion detection, audio-visual correlation) and mask generation before the exposure calculation and adjustment process. This pre-processing step identifies regions of interest in advance, allowing the subsequent exposure control to focus computational resources on relevant areas. The mask is prepared beforehand and applied during exposure calculation, enabling faster processing compared to analyzing the entire image without pre-identification of important regions.
Data Source
AI summary
A region of interest may be determined using any or all of sound source location, multi-person detection, and active speaker detection. An weighted mean may be determined using the region of interest and a set of backlight weight regions, or, only the set of backlight weight regions if a region of interest could not be found. The image mean is compared to a target value to determine if the image mean is greater than or less than the target value within a predetermined threshold. If the image mean is greater than the predetermined target value and predetermined threshold value, the gain and exposure are decreased. If the image mean is lesser than the predetermined target value minus the predetermined threshold value, the gain and exposure are decreased.


