Image Sensor Dual-Mode Pixel Segmentation for Phase Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image sensors face challenges in achieving both high-resolution image capture and effective phase difference focus detection, particularly in large f-number conditions where vignetting occurs, leading to unstable focus detection and limited focus-detectable defocus range.
Innovation Solution
The image sensor employs a dual-mode design with first and second imaging pixels, each comprising segmented photoelectric conversion units with varying base-line lengths for phase difference detection, allowing for adaptive focus detection and 3D image acquisition by selectively using photoelectric conversion units with different base-line lengths based on the focus detection conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the photoelectric conversion unit area is increased to receive light beams from wider pupil regions for high-resolution image capture, then image capturing ability is improved, but focus detection performance deteriorates due to vignetting in large f-number conditions
Solution Approach 1:
Each pixel is divided into multiple photoelectric conversion units (first and second units) with different positions and areas. The first unit is positioned to receive light from the central region for image capture, while the second unit is positioned to receive light from the peripheral region for focus detection, allowing both functions to operate simultaneously without interference
Solution Approach 2:
Different regions of the pixel are assigned different functional characteristics: the first photoelectric conversion unit is optimized for image capturing with larger area, while the second photoelectric conversion unit is optimized for focus detection with specific positioning to receive light from peripheral pupil regions, creating local functional differentiation
2Measurement precision
If the base-line length of photoelectric conversion units is increased to improve focus detection accuracy, then detection accuracy is improved, but the focus-detectable defocus range is limited
Solution Approach 1:
The system dynamically selects which photoelectric conversion units to use for focus detection based on the defocus amount. When the defocus amount is within a predetermined range, the first units are used for high-accuracy detection. When the defocus amount exceeds this range, the second units with different base-line lengths are used to extend the detectable range, creating an adaptive detection system
3Device complexity
If pupil division is limited in one direction to simplify pixel structure, then device complexity is reduced, but focus detection capability for objects with luminance distribution only in orthogonal direction is lost
Solution Approach 1:
Instead of limiting pupil division to one direction, the invention divides the photoelectric conversion units in both horizontal and vertical directions, creating a two-dimensional array. This allows the system to detect focus in any direction by selecting appropriate units based on the luminance distribution orientation of the object
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 approach widens the focus-detectable defocus range, improves detection accuracy in near-in-focus states, and optimizes both focus detection and 3D image acquisition abilities, ensuring stable focus detection even in large defocus states and varying luminance distributions.
Implementation Method 1
each pixel comprises a plurality of photoelectric conversion units segmented in a first direction, the plurality of photoelectric conversion units having an ability of photoelectrically converting a plurality of images formed by split light beams
Data Source
AI summary
An image sensor comprises a first imaging pixel and a second imaging pixel each of which detects an object image formed by a photographing optical system and generates a recording image. Each of the first imaging pixel and the second imaging pixel comprises a plurality of photoelectric conversion units segmented in a first direction, the plurality of photoelectric conversion units have an ability of photoelectrically converting images formed by split light beams out of a light beam from the photographing optical system and outputting focus detection signals to be used to detect a phase difference. A base-line length of photoelectric conversion units to be used to detect the phase difference included in the first imaging pixel is longer than that of photoelectric conversion units to be used to detect the phase difference included in the second imaging pixel.