Adaptive HDR Image Generation via Non-Destructive Sensor Readings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image capture devices face limitations in generating high dynamic range (HDR) images due to narrow dynamic range, leading to overexposed and underexposed areas, requiring costly and time-consuming multi-exposure methods, and complex processing to correct artifacts from moving elements, while also being resource-intensive in terms of computing power.
Innovation Solution
A method for generating HDR images using non-destructive reading (NDRO) of image sensors, where multiple NDRO images are taken during a single exposure time, with adaptive selection of pixels based on signal thresholds to optimize the number of readings and exposure times, allowing real-time processing and storage of a single image plane.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If multiple LDR images with different exposure times are captured and processed to generate HDR image, then the dynamic range of the image is improved, but the time required for image acquisition and processing increases significantly
Solution Approach 1:
The patent applies periodic action by performing multiple non-destructive readings at different time points during a single continuous exposure. Instead of capturing separate LDR images with different exposure times, the sensor performs periodic readings (first reading at time t1, second reading at time t2) during one exposure interval, thereby obtaining multiple signal values that represent different exposure levels without requiring multiple separate captures.
Solution Approach 2:
The patent applies preliminary action by performing the first non-destructive reading during the exposure period before the exposure is complete. This preliminary reading captures an intermediate signal value that represents a partial exposure, which can then be combined with the final signal value from the second reading to reconstruct HDR information without waiting for multiple separate exposures to complete.
2Illumination intensity
If multiple LDR images are captured and processed to generate HDR image, then the dynamic range is improved, but the computational resources and processing complexity increase
Solution Approach 1:
The patent applies taking out by extracting only the necessary signal values from the multiple readings performed during exposure. Instead of processing complete LDR images, the method extracts specific pixel signal values from the first and second NDRO images and combines only those relevant values to generate the HDR image, thereby reducing computational complexity while maintaining HDR quality.
Solution Approach 2:
The patent applies partial action by performing exactly two non-destructive readings during the exposure period - no more, no less. This partial approach (rather than continuous or excessive readings) provides sufficient information for HDR reconstruction while minimizing processing complexity and computational resources required.
3Productivity
If multiple NDRO images are acquired during a single exposure, then the HDR image generation time is reduced, but the computing power required increases due to multiple readings
Solution Approach 1:
The patent applies local quality by processing different regions of the image with different numbers of readings based on their local characteristics. For pixels where the first reading already provides sufficient signal quality, only one reading value is used. For pixels where the second reading provides improved signal-to-noise ratio or additional dynamic range information, both reading values are combined. This local adaptation reduces overall computing power requirements while maintaining HDR quality where needed.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces the time and computational resources required for HDR image generation, adapts to varying brightness levels, and improves image quality by dynamically optimizing the number of NDRO readings, enabling real-time HDR image creation without the need for extensive storage or complex processing.
Implementation Method 1
each associated with a photoelectric conversion element making it possible to convert a light received into electric charges and to accumulate the electric charges during a time of exposure to light
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to the generation of a high dynamic range image of a scene, termed an HDR image, on the basis of a plurality of images obtained by non-destructive reading of an image sensor, called NDRO images. Such an HDR image generation method comprises: the determination of a criterion of desired quality for the HDR image; at least two non-destructive readings of the sensor delivering at least two successive NDRO images; the selection, as a function of the criterion of desired quality, of the first and of the last NDRO image to be used to generate the HDR image; the generation of the HDR image on the basis of information extracted from a series of successive NDRO images starting with the first and terminating with the last NDRO image to be used; the storage of a single image at one and the same time throughout the entire HDR generation phase.