Camera Exposure Prediction Using Reference Images in Projector Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera systems lack the intelligence to set precise exposure values for capturing images of display screens, as they rely on ambient light measurement rather than the brightness of the displayed content, leading to over- or under-exposure issues during calibration and fault detection.
Innovation Solution
A method and system that predict ambient light and adjust camera exposure by using a reference image to determine optimal exposure settings, considering both the displayed image and ambient light, allowing for robust calibration and fault detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If auto exposure mode is used to detect ambient light, then the camera can automatically adjust exposure settings, but the exposure prediction is inaccurate because it measures ambient light instead of the brightness of the displayed content
Solution Approach 1:
The patent introduces an intermediary computational model that mediates between the camera's auto-exposure mechanism and the display content. This model uses a reference image displayed on the screen and captured by the camera to predict the optimal exposure settings, serving as an intermediary that translates display brightness characteristics into appropriate camera exposure parameters, thereby resolving the mismatch between ambient light detection and content brightness measurement
Solution Approach 2:
The system performs preliminary action by displaying a reference image and capturing it at multiple exposures before actual content capture. This preliminary capture phase allows the system to establish exposure predictions based on the reference image's luminance characteristics, which are then applied to subsequent content capture, ensuring accurate exposure settings are established in advance
2Productivity
If multiple reference images are displayed and captured at one exposure, then the calibration process is faster, but determining the optimal exposure requires applying a minimal saturation criterion to multiple images
Solution Approach 1:
The patent implements feedback by capturing multiple reference images at a single exposure setting and evaluating them against a minimal saturation criterion. The system receives feedback from analyzing the saturation levels across multiple captured images, using this feedback to determine whether the exposure setting is optimal or needs adjustment, thereby systematically resolving the exposure determination complexity through iterative evaluation
3Reliability
If a predetermined exposure is set for dark environments, then calibration can be performed under controlled lighting conditions, but the exposure settings are not adaptable when ambient light conditions change
Solution Approach 1:
The patent applies dynamics by transitioning from static predetermined exposure settings to dynamic exposure adjustment. The system continuously monitors the reference image captured under current ambient light conditions and adjusts the exposure prediction accordingly. This dynamic approach allows the exposure settings to adapt automatically to changing ambient light conditions while maintaining calibration consistency through the reference image-based prediction mechanism
Data Source
AI summary
The present invention is directed to predicting and optimizing the exposure value of cameras in order to properly capture images, achieve robust and efficient calibration, and detect faults. The present invention features a method comprising displaying one or more reference images and using the camera to capture the reference image at one or more camera exposures. This allows an optimal camera exposure and an optimal reference image to be determined. The method may further comprise estimating ambient light. The method may further comprise the display device displaying an input image and identifying relative properties of the input image in comparison to the reference image. The method may further comprise predicting an optimal exposure based on the relative properties of the input image, ambient light, and the optimal reference exposure, adjusting the exposure value of the camera based on the optimal exposure, and capturing the input image at the proper exposure.