Image Processing Apparatus Blur Correction via Multi-Exposure Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing coded exposure methods struggle to determine the relative merits of normal exposure and coded exposure images, leading to difficulties in capturing high-quality images with minimal blur, especially in varying illuminance and imaging environments, and require significant computational resources and information detection units.
Innovation Solution
An image processing apparatus that acquires and compares multiple images, including normal exposure, coded exposure, and corrected images, to evaluate and select or combine them based on quality, ensuring the highest quality output image is generated, thereby simplifying the acquisition of high-quality images with minimal blur.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coded exposure method is used to correct blur, then blur correction capability is improved, but image quality varies depending on illuminance and imaging environment making it difficult to determine relative merits in advance
Solution Approach 1:
The patent performs both normal exposure and coded exposure imaging in advance before determining which image quality is superior. This preliminary action captures multiple images simultaneously, eliminating the need for complex pre-determination of relative merits and allowing selection based on actual imaging results.
Solution Approach 2:
The patent introduces an image quality evaluation mechanism that compares normal exposure and coded exposure images, using the comparison results to select the superior image. This feedback loop enables adaptive selection based on actual imaging conditions rather than predetermined rules.
2Measurement precision
If an attempt is made to determine the relative merits of normal exposure and coded exposure images, then image quality selection accuracy is improved, but many information detection units and enormous computing volume are required
Solution Approach 1:
Both normal exposure and coded exposure images are captured in advance, performing the imaging action before quality determination. This eliminates the need for complex real-time analysis and reduces computing requirements by pre-acquiring all necessary image data.
Solution Approach 2:
The patent uses a simple image quality evaluation metric that can be computed efficiently without requiring numerous information detection units or enormous computing resources, making the selection process practical and scalable.
3Manufacturing precision
If an attempt is made to determine the relative merits before imaging, then image quality optimization is improved, but a chance of capturing an object that should be imaged may be missed
Solution Approach 1:
The patent performs both normal exposure and coded exposure imaging actions before quality determination, ensuring that potential capture opportunities are not missed while still enabling quality optimization through subsequent comparison and selection.
Solution Approach 2:
By implementing a feedback mechanism that evaluates actual image quality after capture, the system can reliably select the superior image without risking missed capture opportunities, as both imaging modes are executed in advance.
4Measurement precision
If normal exposure and coded exposure are performed alternately, then blur correction precision is improved, but the image quality of moving images does not necessarily improve
Solution Approach 1:
The patent introduces an image quality evaluation and selection mechanism that compares normal exposure and coded exposure results, using feedback from this comparison to determine which image type provides superior quality for the specific imaging conditions, thereby improving moving image quality.
Solution Approach 2:
The system dynamically selects between normal exposure and coded exposure based on actual imaging results and quality evaluation, rather than rigidly alternating between modes. This dynamic adaptation enables optimization of moving image quality according to specific conditions.
Data Source
AI summary
An image processing apparatus has an acquisition unit that acquires a first image captured by coded exposure in which a transfer unit is driven for n times, and acquires a second image captured by driving the transfer unit for m times (m<n), an image correction processing unit that generates a third image by performing blur correction processing on the first image, and an image comparison unit that evaluates a plurality of images including at least two of the first image, the second image, and the third image, and selects an image of which evaluation is the highest among the plurality of images as an output image, or increases a weight for an image of which evaluation is highest among the plurality of images, relative to those of the other images, when an output image is generated by combining the plurality of images.