Image Capturing Device Blur Detection and Selective Multiplane Addition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital camera methods for correcting hand-blur through multiplane addition often result in unnecessary lowering of image quality when blur is minimal or absent, as they perform addition processing regardless of the actual need for correction, leading to misalignment and noise contamination.
Innovation Solution
The digital camera system captures multiple images with varying exposure times and uses projective transformation matrices to determine the optimal image for minimal blur, applying multiplane addition only when necessary, ensuring accurate alignment and quality preservation by selecting images based on blur detection and exposure conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiplane addition processing is performed for all captured images, then hand-blur correction is achieved, but image quality is unnecessarily lowered when blur is minimal or absent
Solution Approach 1:
The patent changes the parameter of exposure time by capturing images at multiple different exposure times. This allows the system to select an optimal reference image with minimal blur for the addition processing, thereby maintaining image quality while achieving hand-blur correction when needed
Solution Approach 2:
The patent implements feedback by evaluating the blur level of captured images and using this information to determine whether to perform multiplane addition processing. The system selectively applies addition processing only when blur is detected, avoiding unnecessary processing that would degrade image quality
2Reliability
If multiplane addition processing is performed, then signal-to-noise ratio is improved, but images may be misaligned or contaminated with noise
Solution Approach 1:
The patent varies the exposure time parameter across multiple images and selects a reference image with optimal exposure and minimal blur. This ensures that the reference image used for alignment is of high quality, improving alignment accuracy and reducing noise contamination in the final processed image
Solution Approach 2:
The patent performs preliminary selection of a reference image with minimal blur before performing the multiplane addition processing. This preliminary action ensures that the addition processing starts with a high-quality reference, preventing misalignment and noise contamination issues
3Productivity
If hand-blur correction decision is made before photography based on environment detection, then processing efficiency is improved, but correction accuracy decreases when conditions change
Solution Approach 1:
The patent performs preliminary capture of multiple images with varying exposure times before final processing. This preliminary action allows the system to evaluate actual blur conditions at the time of photography and select the optimal reference image, ensuring accurate blur detection even when environmental conditions change between decision and capture
Solution Approach 2:
The patent makes the hand-blur correction decision dynamic by evaluating the actual captured images rather than relying solely on pre-photography environment detection. The system adapts to actual conditions by selecting the best reference image from multiple captures with different exposure times
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
A digital camera carries out the following steps. The digital camera acquires one image photographed with a high sensitivity for subject-blur suppression, four images of a first group consecutive in a time series photographed with an ordinary exposure time, and four images of a second group consecutive in a time series photographed with an exposure time shorter than the ordinary exposure time for multiplane addition (Step A2). Then, the digital camera extracts, from among the images photographed with the ordinary exposure time, an image for which blur caused by hand movement is minimum, and if no image blur has occurred and if blur caused by hand movement is less than a predetermined value, selects the image (Step A11). The digital camera selects the image for subject-blur suppression if image blur has occurred for the extracted image (Step A9). Also, the digital camera generates a new image by carrying out multiplane addition of images for multiplane addition if blur caused by hand movement has occurred for the extracted image (Step A12).