Dual-Readout Pixel Architecture for Imaging and Dynamic Vision Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solid-state imaging devices face challenges in achieving high light-reception efficiency and effective isolation between imaging and event detection functions, leading to degraded resolution and dynamic range, particularly in high-speed applications like autonomous vehicles and robotics.
Innovation Solution
The implementation of a solid-state imaging device with a pixel array where each pixel includes a single photoelectric conversion region and separate readout circuits connected by transfer transistors, surrounded by an isolation structure, allowing for simultaneous imaging and dynamic vision sensing with improved light-reception efficiency and signal isolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single photodiode is shared between image sensor and DVS functions (DAVIS system), then device complexity is reduced, but interference between imaging and event detection functions occurs and dynamic range degrades
Solution Approach 1:
The pixel is divided into distinct functional regions: a photoelectric conversion region for light reception and separate first and second readout circuits for imaging and event detection functions. This segmentation allows independent signal processing paths that prevent interference between DVS and imaging operations while maintaining a single photodiode structure.
Solution Approach 2:
Transfer transistors are introduced as intermediary elements that selectively connect the photoelectric conversion region to either the first readout circuit or the second readout circuit. These transfer transistors act as mediators that route signals appropriately, enabling functional isolation between imaging and event detection pathways without requiring physical separation of the photodiodes.
2Reliability
If additional photodiodes are added per pixel for ATIS system, then imaging and DVS functions are better isolated, but resolution and image quality degrade
Solution Approach 1:
A single photoelectric conversion region is designed to serve multiple functions: it can convert light for imaging operations and generate events for DVS operations. By making the photodiode universal and using transfer transistors to direct its output to different readout circuits based on operational mode, the system achieves functional isolation without adding additional photodiodes, thereby preserving pixel density and image quality.
3Productivity
If readout of DVS and active image sensor signals is difficult, then dynamic range degrades
Solution Approach 1:
The readout system is segmented into separate first and second readout circuits, each optimized for specific signal types. The first readout circuit handles imaging signals while the second handles DVS events, allowing parallel independent readout operations that improve efficiency and prevent signal contention that would limit dynamic range.
Solution Approach 2:
Transfer transistors provide dynamic switching capability that allows the system to flexibly route signals from the photoelectric conversion region to the appropriate readout circuit based on operational requirements. This dynamic routing enables efficient signal capture for both imaging and event detection without loss of signal integrity, preserving dynamic range.
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 enables enhanced light-reception efficiency and improved isolation between pixels, enabling both imaging and event detection functions with improved resolution and dynamic range, suitable for high-speed applications.
Implementation Method 1
Each pixel includes a single photoelectric conversion region, a first readout circuit selectively connected to the photoelectric conversion region by a first transfer gate or transfer transistor
Data Source
AI summary
An imaging device includes a plurality of unit pixels or pixels, with each pixel separated from every other unit pixel by an isolation structure. Each unit pixel includes a photoelectric conversion unit, a pixel imaging signal readout circuit, and an address event detection readout circuit. A first transfer transistor selectively connects the photoelectric conversion unit to the pixel imaging signal readout circuit, and a second transfer transistor selectively connects the photoelectric conversion unit to the address event detection readout circuit. The photoelectric conversion unit, the pixel imaging signal readout circuit, the address event detection readout circuit, and the first and second transfer transistors for a given pixel are located within a pixel area defined by the isolation structure. The isolation structure may be in the form of a full thickness dielectric trench isolation structure.