Dynamic Exposure Time Selection for HDR Video Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High Dynamic Range Imaging (HDRI) and Video applications face challenges in preserving texture information due to limited sets of predefined exposure times, leading to potential loss of image and texture details, especially under time constraints.
Innovation Solution
A device and method that select a small set of optimal exposure times from a larger set by computing score values based on intensity-dependent values of calibration images, using pixel blocks and entropy or variance calculations to identify the highest or lowest disorder, thereby preserving image and texture details.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a small set of predefined exposure times is used for HDR imaging, then time constraints are satisfied and processing speed is improved, but texture information and image details are lost
Solution Approach 1:
The patent changes the parameter of exposure time selection from a static predefined set to a dynamic optimized set. By computing entropy values for different exposure times and selecting those that maximize texture information preservation, the system adapts the exposure time parameters to specific imaging conditions, resolving the contradiction between speed and information preservation
Solution Approach 2:
The patent replaces the mechanical approach of using a fixed small set of exposure times with an information-theoretic approach using entropy calculations. This substitution allows the system to intelligently select exposure times based on the actual content and texture information in the images, rather than relying on a predetermined mechanical set
2Loss of information
If a large set of different exposure times is captured, then texture information and image details are preserved, but time constraints are violated and processing becomes inefficient
Solution Approach 1:
The patent extracts only the essential exposure times from a larger set by calculating entropy values and selecting those that provide maximum information content. This extraction process identifies and removes redundant exposure times that do not contribute significantly to texture information preservation, thereby reducing capture time while maintaining image quality
Solution Approach 2:
The patent performs preliminary entropy calculations on a set of candidate exposure times before actual HDR capture. This preliminary action allows the system to pre-determine the optimal subset of exposure times for a given scene, avoiding the need to capture unnecessary images and thus reducing overall capture time while preserving image details
3Object-generated harmful factors
If only two different exposure times are used in video HDR, then ghosting effects are minimized and processing is simplified, but texture information is lost
Solution Approach 1:
The patent introduces dynamics into the exposure time selection by allowing the number of exposure times to be adjusted based on scene requirements. While video HDR typically uses two exposure times to minimize ghosting, the system can dynamically increase this number when texture information preservation becomes the priority, adapting to different video content and lighting conditions
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The invention proposes a method for controlling a capture of a plurality of images for obtaining a High Dynamic Range, HDR, image. The method comprises storing a main set of selectable exposure times and a plurality of candidate reduced sets being respective subsets of the main set, controlling a capture of calibration images at respectively each exposure time of the main set, for each candidate reduced set, selecting the calibration images captured with the exposure times of the candidate reduced set and computing a score value depending on intensities of the selected calibration images, selecting a candidate reduced set from among the plurality of the candidate reduced sets on the basis of the computed score values of all candidate reduced sets, and controlling the capture of the plurality of images at respectively each exposure time of the selected candidate reduced set for obtaining the HDR image.