Image Sensor Coprime Pixel Division for Focus Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image sensor technologies using a two-dimensional image sensor with microlenses for focus detection suffer from low light-receiving sensitivity due to separation bands between photoelectric conversion portions, leading to unnatural blurring and deteriorated image quality, especially when the f-number of the taking lens is narrowed.
Innovation Solution
The image sensor is designed with multiple image sensing pixel groups, each having photoelectric conversion portions divided in different patterns, ensuring that adjacent pixels are not adjacent, and the division patterns are coprime, which helps in averaging the influence of low sensitivity bands and maintaining image quality across various f-number settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If photoelectric conversion portions are divided to enable pupil division and focus detection, then focus detection capability is improved, but light-receiving sensitivity decreases due to separation bands
Solution Approach 1:
The photoelectric conversion portions are divided into multiple segments (first and second pixels with different division patterns) to enable pupil division for focus detection. This segmentation allows the system to capture light from different regions of the exit pupil, providing the necessary information for phase difference focus detection while maintaining adequate light reception through strategic arrangement of the segments.
Solution Approach 2:
Different pixels are assigned different division patterns (coprime numbers of divisions in first and second directions) to create local variations in how light is received and processed. This local quality differentiation ensures that while each pixel has reduced sensitivity in specific directions due to separation bands, the overall system maintains uniform sensitivity by averaging out the directional biases across pixels with different patterns.
2Measurement precision
If separation bands are formed to separate photoelectric conversion portions, then pupil division is achieved, but unnatural blurring occurs in captured images
Solution Approach 1:
The patent employs asymmetric division patterns where the number of divisions in the first direction and second direction are coprime natural numbers. This asymmetry ensures that separation bands in different pixels are oriented at different angles and positions, preventing systematic blurring artifacts while maintaining the necessary pupil division for accurate focus detection.
Solution Approach 2:
The patent addresses the blurring problem by introducing a second division direction perpendicular to the first direction. By dividing photoelectric conversion portions in both directions with coprime numbers, the separation bands are distributed across multiple dimensions, preventing concentration of sensitivity loss in a single direction and thereby reducing unnatural blurring in the captured images.
3Area of stationary object
If f-number of the taking lens is narrowed to improve depth of field, then image coverage is improved, but pixel output decreases and image quality deteriorates
Solution Approach 1:
The pixel structure is segmented into multiple photoelectric conversion portions with different division patterns, allowing each segment to capture light from different regions of the exit pupil. This segmentation enables effective use of available light even at smaller apertures, maintaining pixel output when the f-number is narrowed to increase depth of field.
Solution Approach 2:
The patent changes the structural parameters of the photoelectric conversion portions by using coprime numbers of divisions in two perpendicular directions. This parameter change optimizes the light-receiving characteristics to maintain adequate pixel output across a range of f-numbers, including narrower apertures where depth of field is increased.
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
This configuration suppresses the impact of low sensitivity bands, stabilizes image quality, and allows for effective focus detection and image capture even at lower f-numbers, preventing the deterioration of image quality associated with separation bands.
Implementation Method 1
a plurality of photoelectric conversion portions arrayed in a first direction for a first number of divisions M1 and in a second direction perpendicular to the first direction for a second number of divisions N1... have a function that photoelectrically converts a plurality of images formed by divided light fluxes
Data Source
AI summary
An image sensor comprising a plurality of image sensing pixel groups is provided. Each of the image sensing pixel groups has a plurality of first pixels each having photoelectric conversion portions arrayed in first and second directions for first and second numbers of divisions, respectively, and a plurality of second pixels each having photoelectric conversion portions arrayed in the first and second directions for third and fourth numbers of divisions, respectively. The photoelectric conversion portions comprising the first pixel and the second pixel have a function of photoelectrically converting a plurality of images formed by divided light fluxes of a light flux from an imaging optical system and outputting a focus detection signal for phase difference detection. The first and third numbers of division are coprime natural numbers, and the second and fourth numbers of divisions are coprime natural numbers.