CMOS Image Sensor Pixel with Pupil-Divided Photoelectric Units for HDR Autofocus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging devices, such as CMOS image sensors, face challenges in achieving high dynamic range moving images while maintaining effective autofocus functionality.
Innovation Solution
The proposed imaging device employs a configuration with multiple photoelectric conversion units and holding units to generate and output signals based on charges accumulated during different exposure times, allowing for the combination of high dynamic range imaging and phase detection autofocus by adjusting the timing of signal readouts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If pixels with different sensitivities are used to generate high dynamic range images and phase difference information, then high dynamic range imaging is achieved, but autofocus function is insufficient
Solution Approach 1:
The pixel array is divided into multiple regions with different photoelectric conversion units (first, second, third, and fourth photoelectric conversion units) that capture light from different pupil areas. This segmentation allows simultaneous acquisition of images for high dynamic range (combining first and third units) and phase difference autofocus (comparing second and fourth units), resolving the contradiction between HDR and autofocus performance
Solution Approach 2:
The imaging device integrates multiple functions within a single pixel structure by incorporating multiple photoelectric conversion units that can serve different purposes. The same pixel array generates both high dynamic range images (using first and third photoelectric conversion units) and phase difference information for autofocus (using second and fourth photoelectric conversion units), making the system universally capable of both functions simultaneously
2Illumination intensity
If multiple signals with different exposure times are output, then wide dynamic range moving images are captured, but device complexity increases
Solution Approach 1:
The device performs preliminary actions by capturing multiple exposure signals (first signal with first exposure time, second signal with second exposure time) simultaneously across the pixel array before processing. The holding units store charges from different exposure times, and the readout circuitry is pre-configured to output these signals in different sequences for different frames, simplifying the overall processing complexity while achieving wide dynamic range
Solution Approach 2:
The imaging device uses periodic action by alternating the output sequence of signals between different frames. In odd frames, the first and second signals are output before the third and fourth signals, while in even frames, the reverse sequence is used. This periodic alternation simplifies the processing of wide dynamic range moving images by maintaining a regular pattern that can be efficiently handled by the readout circuitry
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 enables the capture of wide dynamic range moving images while achieving high-performance autofocus, improving image quality and focus accuracy.
Implementation Method 1
a first photoelectric conversion unit on which a first part of pupil-divided incident light is incident, a second photoelectric conversion unit on which a second part of the pupil-divided incident light is incident
Data Source
AI summary
An imaging device includes pixels each including first and second photoelectric conversion units on which pupil-divided parts of incident light are incident and a holding unit that holds charges transferred from the first and second photoelectric conversion units, and outputting signals based on amounts of charges held by the holding unit. Each pixel outputs a first signal and a second signal based on amounts of charges generated by the first photoelectric conversion unit and by the first and second photoelectric conversion units, respectively, during a first exposure time, and a third signal and a fourth signal based on amounts of charges generated by the first photoelectric conversion unit and by the first and second photoelectric conversion units, respectively, during a second exposure time. The first and second signals are output before the third and fourth signals in one frame and after the third and fourth signals in another frame.