Phase Difference Pixel Reliability Adjustment Using Virtual Image Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional phase difference AF systems face challenges in achieving high precision focus adjustment, especially with subjects having high-frequency patterns, leading to erroneous detection and reduced accuracy due to ranging variations caused by separated phase difference pixel arrangements.
Innovation Solution
An imaging device with paired phase difference pixels and a processor that calculates virtual image pixel data and determines the degree of coincidence between pixel data, allowing for reliable phase difference detection and focus adjustment by excluding low-reliability ranging areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If phase difference pixels are separated to detect phase difference, then phase difference detection is enabled, but ranging variations occur for high-frequency patterns causing erroneous detection
Solution Approach 1:
The patent introduces virtual image pixels as an intermediary between the separated phase difference pixels. These virtual pixels are calculated through interpolation from image pixels and serve as a reference to mediate the comparison between phase difference pixel outputs. This intermediary allows the system to detect whether ranging variations have occurred without directly comparing the separated phase difference pixels alone, thereby resolving the reliability issue while maintaining detection capability.
Solution Approach 2:
The patent implements a feedback mechanism where the degree of coincidence between virtual image pixel output and phase difference pixel output is continuously monitored. When the degree of coincidence falls below a threshold (indicating ranging variation), the system adjusts its detection strategy by excluding unreliable phase difference pixel pairs from focus adjustment calculations. This feedback loop maintains detection reliability dynamically while preserving measurement precision for valid measurements.
2Ease of operation
If phase difference pixels are arranged on image plane, then focus adjustment can be performed, but erroneous detection results are adopted for high-frequency patterns
Solution Approach 1:
The patent performs preliminary calculations of virtual image pixel outputs through interpolation before conducting phase difference detection. This preliminary action creates a reference framework in advance that allows the system to identify and exclude erroneous detections caused by high-frequency patterns before they affect focus adjustment decisions, thereby maintaining both ease of operation and measurement precision.
Solution Approach 2:
The patent dynamically changes the parameter of detection reliability by calculating the degree of coincidence between virtual and phase difference pixel outputs. Based on this calculated parameter, the system adjusts which phase difference pixel pairs are used for focus adjustment, effectively changing the detection parameters in response to subject characteristics like high-frequency patterns, thus maintaining accuracy without compromising operational simplicity.
3Reliability
If virtual image pixel data is calculated through interpolation, then ranging variations are suppressed, but additional calculation steps are required
Solution Approach 1:
The patent creates virtual image pixels as copies of what the phase difference pixel outputs would be if no ranging variations occurred. These copied pixel values are obtained through interpolation from surrounding image pixels, providing a reference model that suppresses ranging variations. This copying approach maintains reliability while keeping the additional processing relatively simple by reusing existing image pixel data.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution enables high-precision focus adjustment even with high-frequency patterns by using virtual image pixel data to suppress ranging variations and exclude erroneous data, resulting in improved focus detection accuracy.
Implementation Method 1
image pixels which receive light of subject image through a photographing lens and perform photoelectric conversion, and paired phase difference pixels which respectively receive light flux corresponding to pupil regions that are paired with the photographing lens and perform photoelectric conversion
Data Source
AI summary
An imaging device, comprising an image sensor having image pixels and phase difference pixels, a phase difference detection section that detects a phase difference based on pixel data of the phase difference pixels, a pixel data calculation section that calculates pixel data of virtual imaging pixels at positions of the phase difference pixels, a degree of coincidence calculation section that calculates a degree of coincidence between each pixel data of the virtual imaging pixels that has been calculated, a reliability determination section that determines reliability of the phase difference detection result in accordance with the degree of coincidence, and a focus adjustment section that performs focus adjustment based on the phase difference detection result and the reliability.