Image Sensor Pixel-Level Exposure Control for High-Speed HDR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CMOS image sensors face limitations in dynamic range, leading to motion artifacts and reduced resolution in high-speed photography applications, particularly in scenarios requiring over 100 dB dynamic range, such as security monitoring and autonomous driving.
Innovation Solution
An image sensor with a two-dimensional pixel array and a pixel-level exposure control method that includes an image segmentation module, prediction module, and regional exposure control module to adjust exposure time based on image features, allowing independent control of exposure time for each pixel.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If multiple exposures of different time lengths are performed on the same target scene, then dynamic range is extended, but motion artifacts occur when subjects move
Solution Approach 1:
The pixel array is divided into multiple regions, with each region independently controlling exposure time through separate transfer gate control units. This allows different exposure durations for different spatial regions within a single frame, extending dynamic range without requiring multiple frame captures, thereby avoiding motion artifacts.
Solution Approach 2:
The patent implements dynamic exposure time control where each pixel region can adjust its exposure duration independently based on local lighting conditions. The transfer gate control units dynamically regulate electron transfer timing, enabling adaptive exposure across the image sensor while capturing a single frame, thus preventing motion-related degradation.
2Illumination intensity
If long exposure and short exposure are performed simultaneously on different rows, then dynamic range is extended, but spatial resolution is sacrificed
Solution Approach 1:
Instead of dividing the sensor into rows with different exposure times, the patent segments the pixel array into regions that can be independently controlled. Each region maintains full spatial resolution while contributing to the overall dynamic range through localized exposure time variation, avoiding the resolution loss associated with row-based segmentation.
Solution Approach 2:
Different exposure times are applied to different spatial regions based on local lighting requirements rather than uniform row-based segmentation. This allows each region to optimize its exposure independently while maintaining the sensor's full spatial resolution, as no pixels are discarded or averaged across rows.
3Illumination intensity
If two different photodiodes with different photoresponsivity are used in a pixel unit, then dynamic range is extended, but pixel size increases
Solution Approach 1:
The patent segments the pixel array into multiple independently controllable regions, allowing dynamic range extension through spatial variation of exposure time rather than through multiple photodiodes per pixel. This maintains small pixel size while achieving HDR capability through regional exposure control.
Solution Approach 2:
Instead of changing the physical structure of pixels by adding multiple photodiodes, the patent changes the exposure time parameter dynamically across different regions. This parameter-based approach achieves extended dynamic range without increasing pixel area, as each pixel maintains its original size and structure.
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
Enables single-frame ultra-high dynamic range imaging with reduced motion artifacts, suitable for high-speed photography, and maintains sensor resolution, facilitating compatibility with small-size pixels and reducing chip costs.
Implementation Method 1
a photodiode (PD) 32′ used to accumulate electrons generated by the photoelectric effect
Data Source
AI summary
An image sensor includes a two-dimensional pixel array including a plurality of sensor pixels, wherein the sensor pixels are configured to collect image data of a target scene according to an adjustable exposure time; an image segmentation module configured to analyze image data collected by the two-dimensional pixel array, extract image features and establish a plurality of pixel partitions based on sensor pixels with same image features; a prediction module configured to obtain a predicted exposure time according to image features of pixel partitions; and a regional exposure control module configured to generate a control signal for an exposure time according to the predicted exposure time, and send the control signal to sensor pixels of a pixel partition corresponding to the predicted exposure time.