Adaptive HDR Image Capture for Motion and Dynamic Range
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image capturing techniques struggle to generate high-quality images with a wide dynamic range when there is a large difference in brightness and subject movement, leading to issues like blocked up shadows, blown out highlights, and unnatural composite images.
Innovation Solution
An image capturing apparatus that adaptively switches between multiple shooting modes: one mode involves multiple exposures to capture images with different dynamic ranges, and the other mode applies different gains to the image signal within the image sensor to achieve the same, allowing for high-quality image synthesis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If multiple shots are performed with different exposure values to expand dynamic range, then the dynamic range is improved, but the image quality deteriorates when the subject is moving due to unnatural composite images
Solution Approach 1:
The patent applies dynamics by making the shooting method adaptable based on subject movement detection. The system dynamically switches between multi-shot HDR (when subject is stationary) and single-shot multi-gain (when subject is moving), allowing the image capturing approach to change based on real-time conditions. This resolves the contradiction by enabling dynamic range expansion only when necessary while maintaining image quality when the subject moves.
Solution Approach 2:
The patent changes the shooting parameters (number of shots, gain values, exposure settings) based on the detected subject movement. When movement is detected, the system changes from multi-shot mode to single-shot mode with varied gains, preventing unnatural compositing while still achieving extended dynamic range through the multi-gain approach within a single exposure.
2Illumination intensity
If the difference in amplification gain is increased to expand dynamic range, then the dynamic range is improved, but the tone of pixels in intermediate output levels is lost
Solution Approach 1:
The patent segments the pixel output levels into multiple ranges, each handled by a different gain setting. Instead of using a single large gain difference that would lose intermediate tones, the system divides the dynamic range into segments and applies appropriate gains to each segment, preserving tone information across the full range while still achieving extended dynamic range coverage.
Solution Approach 2:
The patent applies different gain values to different pixel signals based on their output levels. Pixels in different brightness ranges receive different gain treatments, with intermediate tone pixels receiving gains that preserve their tonal information. This local differentiation of gain application resolves the contradiction by maintaining tone quality in intermediate levels while still expanding the overall dynamic range through strategic gain selection.
3Illumination intensity
If multiple shots are performed to capture different exposure levels, then the dynamic range is improved, but the shooting time increases
Solution Approach 1:
The patent dynamically adjusts the number of shots based on subject movement detection. When movement is detected, the system reduces the number of shots from multiple to one, significantly reducing shooting time while compensating for dynamic range requirements through multi-gain processing of the single shot. This resolves the time-dynamic range contradiction by adapting the shooting duration to actual scene requirements.
Data Source
AI summary
An image capturing apparatus comprises an image sensor; and a control unit that, in a case where a first shooting mode is set, controls to perform shooting a plurality of times with the image sensor while changing exposure to obtain a plurality of images, and in a case where a second shooting mode different from the first shooting mode is set, controls to amplify an image signal obtained by performing shooting once with the image sensor by a plurality of different gains within the image sensor to obtain a plurality of images.