Image Sensor Focus Detection Noise Reduction via Signal Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional focus detection methods suffer from increased noise in image signals, leading to deteriorated image quality, especially in low contrast and low luminance environments, resulting in erroneous focus detection and hunting issues.
Innovation Solution
An image capturing apparatus with an image sensor having first and second photoelectric conversion portions per microlens, where an added signal is read out along with independent signals from each conversion portion, allowing for noise correction and accurate correlation calculation to enhance focus detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the A image and B image are independently read out to perform focus detection calculation, then the focus detection can be performed, but the random noise amount increases due to addition of A image and B image, deteriorating image quality
Solution Approach 1:
The patent extracts only the necessary signal components for focus detection by reading out the A image and A+B image independently, then calculating the B image through subtraction. This extraction approach allows focus detection to be performed while minimizing the propagation of random noise that would occur if both A and B images were independently read out and added together.
Solution Approach 2:
The patent performs preliminary action by reading out the A image and A+B image signals before performing the subtraction calculation to generate the B image. This sequencing allows the system to prepare the necessary signal components in advance, reducing the need for subsequent noise-prone operations and improving overall signal quality for focus detection.
2Quantity of substance
If the A+B image is generated by adding A image and B image, then the complete image signal is obtained, but the random noise amount increases, deteriorating image quality
Solution Approach 1:
Instead of generating the complete A+B image by adding A and B images (which would amplify noise), the patent extracts only the necessary signal components by reading out A and A+B independently, then deriving B through subtraction. This approach maintains image signal completeness for focus detection while avoiding the noise amplification that would occur with direct addition.
Solution Approach 2:
The patent uses the A+B image as an intermediary signal that contains both A and B components. By reading out this intermediary signal along with the A image, the system can derive the B image through subtraction without needing to directly add A and B images, thereby avoiding noise amplification while maintaining signal completeness.
3Difficulty of detecting and measuring
If pixels in the image sensor receive light from different pupil regions for phase difference detection, then focus detection is enabled, but the image quality deteriorates due to noise in low contrast and low luminance environments
Solution Approach 1:
The patent employs feedback mechanisms through correlation calculation that compares the A image with the derived B image to determine focus status. This feedback loop allows the system to detect focus conditions while the independent reading out of A and A+B signals provides noise-resistant signal paths, enabling focus detection without excessive noise in challenging lighting conditions.
Solution Approach 2:
The patent changes the signal processing parameters by reading out A and A+B images independently rather than A and B, then using subtraction to derive B. This parameter change in the signal acquisition and processing method reduces noise propagation while maintaining the phase difference detection capability necessary for focus detection in low contrast and low luminance environments.
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 reduces noise influence and enables highly accurate focus detection even in challenging conditions, such as low contrast and low luminance environments, by correcting noise through specific signal processing and correlation calculation.
Implementation Method 1
an image sensor in which a plurality of unit pixels, each of which has a first photoelectric conversion portion and a second photoelectric conversion portion arranged for one microlens
Data Source
AI summary
An image capturing apparatus comprises an image sensor; a readout unit that reads out an added signal and that independently reads out the signal of the first photoelectric conversion portion; a calculation unit that calculates a signal corresponding to a signal of the second photoelectric conversion portion by subtracting the signal of the first photoelectric conversion portion from the added signal; a correlation calculation unit that performs correlation calculation for the signal of the first photoelectric conversion portion and the signal corresponding to the signal of the second photoelectric conversion portion; and a subtraction unit that subtracts, from the result of the correlation calculation on an object image by the correlation calculation unit, a correction value for correcting noise caused by obtaining the signal corresponding to the signal of the second photoelectric conversion portion.