PDAF Sensor Remosaicing With Balanced Green Artifact Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Phase-detection autofocus (PDAF) in electronic cameras often results in defocused portions of images generating artifacts, which are difficult to remove without affecting the in-focus portions, leading to blurred images.
Innovation Solution
A remosaicing method using a specific kernel convolution process that balances the intensity contributions of neighboring pixels under a microlens to remove artifacts in defocused areas while preserving sharp images in focused areas, employing a kernel that satisfies certain symmetry and balance conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If low pass filtering is used to remove artifacts from defocused portions, then artifacts are removed, but in-focus portions become blurred
Solution Approach 1:
The patent applies different processing treatments to different regions of the image: defocused portions undergo artifact removal through specific filtering operations, while in-focus portions are preserved with their original sharpness. This is achieved by identifying defocused regions and applying artifact removal algorithms selectively, ensuring that only the problematic areas are processed while leaving the in-focus areas untouched.
Solution Approach 2:
The image is segmented into defocused portions and in-focus portions based on focus detection. This segmentation allows the system to apply different processing strategies to each region: artifact removal for defocused areas and preservation for in-focus areas, thereby resolving the contradiction between removing artifacts and maintaining sharpness.
2Object-affected harmful factors
If artifact removal is applied to defocused portions, then image quality in defocused areas improves, but overall image resolution may be compromised
Solution Approach 1:
The patent applies artifact removal algorithms selectively only to defocused portions of the image rather than the entire image. This partial action approach removes artifacts where needed while preserving the original resolution and detail in in-focus portions, thus avoiding the trade-off between artifact removal and overall resolution.
Data Source
AI summary
An imaging system comprising (1) a phase-detection image sensor comprising a plurality of phase-detection (PD) pixel units, each pixel unit comprises an upper left PD pixel, an upper right PD pixel, a lower left PD pixel, and a lower right PD pixel, the upper left PD pixel, the upper right PD pixel, the lower left PD pixel, and the lower right PD pixel covered by a microlens, and (2) a processor configured to: interpolate a green image to obtain a full resolution interpolated green image, and convolute the full resolution interpolated green image with an N×N kernel, where the N×N kernel satisfies conditions including the total contribution from all upper left pixels, the total contribution from all upper right pixels, the total contribution from all lower left pixels, and the total contribution from all lower right PD pixels in the convolution, being the same.