High- and Low-Resolution Image Sensor Fusion for Photosensitivity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The resolution of merged images is significantly reduced due to pixel merging, failing to meet the high resolution requirements of certain industries, leading to poor performance in output images.
Innovation Solution
An image processing method that controls an image sensor to sequentially output images with different resolutions, including higher resolution first images and lower resolution second images, and fuses these images to achieve improved resolution and photosensitivity in the output images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If pixel merging technology is used to improve photosensitivity, then photosensitivity is improved, but resolution is greatly reduced
Solution Approach 1:
The patent segments the image capture process into multiple stages: first capturing images at high resolution (first images), then capturing images at lower resolution with better photosensitivity (second images). This segmentation allows the system to obtain both high resolution and good photosensitivity by processing different image types separately before fusion, rather than using a single merged image approach.
Solution Approach 2:
The patent merges the advantages of high resolution and good photosensitivity by combining first images (high resolution) and second images (good photosensitivity) through a fusion algorithm. The fusion module processes these images together to generate output images that achieve both high resolution and good photosensitivity, effectively merging the benefits of both image types.
2Illumination intensity
If pixel merging technology is used to improve photosensitivity, then photosensitivity is improved, but the resolution becomes 1/n of the original image
Solution Approach 1:
The patent divides the image acquisition process into separate captures: first images captured at high resolution and second images captured at lower resolution with enhanced photosensitivity. This segmentation prevents the resolution loss that would occur if all pixels were merged, as the high resolution first images preserve the original 1:1 pixel correspondence.
Solution Approach 2:
The patent changes the capture parameters dynamically: for first images, it uses high resolution settings (1:1 pixel correspondence); for second images, it uses lower resolution settings with better photosensitivity. This parameter change allows the system to optimize for different requirements and then combine the results, avoiding the fixed 1/n resolution limitation of traditional pixel merging.
3Illumination intensity
If traditional pixel merging is used, then photosensitivity is improved, but the output image quality does not meet high resolution requirements
Solution Approach 1:
The patent segments the image processing into distinct modules: a first image capture module for high resolution images, a second image capture module for photosensitive images, and a fusion module that combines them. This segmentation ensures that high resolution requirements are met by preserving the high resolution first images while incorporating the photosensitivity benefits of second images.
Solution Approach 2:
The fusion module merges the high resolution first images with the photosensitive second images to produce output images that meet both high resolution requirements and good photosensitivity. This combining process ensures output image quality reliability by integrating the strengths of both image types rather than relying on traditional pixel merging that sacrifices resolution.
Data Source
AI summary
The present disclosure describes an image processing method, apparatus and electronic device. The method includes controlling an image sensor to sequentially output first images and second images according to a predefined output rule, where the second images are images corresponding to the first images, and the first images and the second images have different resolutions; and fusing a first image with a second image corresponding to the first image to obtain a to-be-output image.