Pixel Calibration for Super Resolution Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital camera systems face limitations in resolution enhancement due to pixel-to-pixel sensitivity differences and the challenge of capturing clear images of moving objects, leading to propagating aberrations in move-based super resolution image processing.
Innovation Solution
A camera system that employs a move-based super resolution mode by shifting a detector array in increments relative to a focal plane, capturing multiple images which are then combined using an interleaving process or linear algebra to produce a single super resolution image, while also compensating for pixel sensitivity differences and motion within the scene.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If move-based super resolution mode is used to enhance image resolution, then image resolution is improved, but pixel-to-pixel sensitivity differences cause propagating aberrations
Solution Approach 1:
The system performs preliminary calibration of pixel sensitivity values before executing move-based super resolution imaging. By pre-characterizing each pixel's sensitivity and storing calibration data, the system eliminates propagating aberrations that would otherwise occur during the super resolution process, ensuring both high resolution and image accuracy.
Solution Approach 2:
The system uses calibration images to generate feedback about pixel sensitivity variations. This feedback is used to compute correction factors that are applied during super resolution processing, continuously compensating for pixel-to-pixel sensitivity differences and preventing aberration propagation.
2Measurement precision
If multiple images are captured to produce super resolution image, then image resolution is improved, but motion within the scene causes motion blur
Solution Approach 1:
The system dynamically adjusts the imaging process by capturing multiple frames at different positions and using motion detection algorithms to identify and compensate for scene movement. By making the super resolution process adaptive to motion conditions, the system maintains image clarity while achieving enhanced resolution.
Solution Approach 2:
The system extracts motion information from the sequence of captured images and separates the motion component from the resolution enhancement component. By taking out the motion analysis as a distinct processing step, the system can compensate for motion blur while preserving the resolution benefits of multi-frame capture.
3Measurement precision
If detector array is shifted in increments to capture multiple images, then image resolution is improved, but processing complexity increases
Solution Approach 1:
The super resolution processing is segmented into distinct modular stages: image capture with detector shifting, calibration application, motion compensation, and final image synthesis. By dividing the complex processing into manageable segments, the system reduces overall processing complexity while maintaining high resolution output.
Solution Approach 2:
Pixel calibration data is computed and stored in advance before super resolution imaging operations. This preliminary preparation of calibration lookup tables and sensitivity correction factors eliminates the need for complex real-time calculations during image processing, significantly reducing processing complexity.
Data Source
Figure 1~3
Figure 4~9
Figure 10~15
AI summary
A method of producing super resolution images includes a detector array capturing images of a scene in which detector array provides pixels for respective portions of the scene, and the detector array captures the images in increments in each of which a lens or the detector array is moved a length of an individual detector. The method also includes a processing unit determining light intensities of the pixels, and calibrating the pixels to a reference pixel based on the light intensities and thereby compensate for any differences in sensitivity of the individual detectors. This calibration is based on a comparison of light intensities of the reference pixel and an other pixel in respectively one and an other of the images in which the portion of the scene is the same for the reference pixel in the one of the images and the other pixel in the other of the images.