Depth Value Correction Using Pixel Contribution Ratios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for calculating depth values from stereo images or image plane phase difference pixels often result in inaccurate depth values due to the assumption that all pixels are at the same distance, leading to pseudo depth errors when pixels at different distances are included in the calculation.
Innovation Solution
An image processing device that calculates a contribution ratio for each pixel or region in the input image and uses this ratio to correct the depth value, distributing the depth value based on the contribution of each pixel to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If depth value is calculated using conventional methods assuming all pixels are at the same distance, then calculation process is simple, but depth accuracy deteriorates when pixels at different distances are included
Solution Approach 1:
The patent segments the calculation process into two distinct stages: first calculating initial depth values using conventional methods, then performing correction by calculating contribution ratios for each pixel. This segmentation allows the system to maintain simple initial calculation while adding targeted correction only where needed, resolving the contradiction between simplicity and accuracy.
Solution Approach 2:
The patent applies local quality by calculating contribution ratios specifically for pixels that may be at different distances, rather than uniformly processing all pixels. The correction is applied locally to affected regions based on their specific contribution ratios, maintaining accuracy where needed while preserving simplicity in regions where conventional methods suffice.
2Measurement precision
If contribution ratio calculation and depth correction are performed for all pixels, then depth accuracy is improved, but calculation time increases
Solution Approach 1:
The patent implements partial action by performing contribution ratio calculation and depth correction only for pixels that are likely to be affected by distance variations, rather than uniformly processing all pixels. This selective approach achieves sufficient depth accuracy improvement while minimizing the additional calculation time required.
3Quantity of substance
If pixels at different distances are included in depth calculation, then more pixels contribute to depth value, but pseudo depth errors occur
Solution Approach 1:
The patent changes the parameter of pixel contribution by introducing contribution ratios that vary for each pixel based on its distance characteristics. Pixels at different distances are assigned different contribution ratios, allowing more pixels to contribute to the depth calculation while maintaining reliability through weighted contribution rather than uniform treatment.
Data Source
AI summary
An image processing device includes: a contribution ratio calculation unit that calculates a contribution ratio of a predetermined pixel or a predetermined region in depth calculation in each of a plurality of pixels or a plurality of regions included in an input image; and a correction unit that corrects a depth value of the predetermined pixel or the predetermined region based on the contribution ratio.