Image Sensor Exposure Convergence for Sudden Light Changes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image sensors in self-driving cars face challenges in quickly adjusting exposure levels during drastic illumination changes, leading to temporary overexposure or underexposure, which affects object detection accuracy and increases accident risk.
Innovation Solution
An image sensing device with a multi-frame rate array that generates pixel data with different exposure times, utilizing an automatic exposure convergence operation to adjust exposure parameters and perform image fusion, enabling fast auto-exposure convergence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If multiple independently captured images are merged to capture high dynamic range, then the dynamic range is improved, but the time consumption and processing complexity increase
Solution Approach 1:
The patent performs preliminary exposure convergence operations using only a subset of pixel data (first pixel data with shorter exposure time) before final image merging. This preliminary action establishes initial exposure parameters that guide the subsequent merging process, reducing the time needed to process all pixel data while still achieving high dynamic range results.
Solution Approach 2:
The patent segments the pixel data into different groups based on exposure time (first pixel data with shorter exposure time, second pixel data with longer exposure time). By processing these segments differently - using the first segment for rapid exposure convergence and the second for final high dynamic range merging - the system achieves both speed and quality.
2Illumination intensity
If multiple independently captured images are merged to capture high dynamic range, then the dynamic range is improved, but the processing complexity increases
Solution Approach 1:
The patent performs preliminary exposure convergence operations using only a subset of pixel data (first pixel data with shorter exposure time) before final image merging. This preliminary action establishes initial exposure parameters that guide the subsequent merging process, reducing the time needed to process all pixel data while still achieving high dynamic range results.
Solution Approach 2:
The patent segments the pixel data into different groups based on exposure time (first pixel data with shorter exposure time, second pixel data with longer exposure time). By processing these segments differently - using the first segment for rapid exposure convergence and the second for final high dynamic range merging - the system achieves both speed and quality.
3Measurement precision
If image sensors adjust exposure levels during drastic illumination changes, then the image quality is improved, but the response speed is insufficient
Solution Approach 1:
The patent performs preliminary exposure convergence operations using only a subset of pixel data (first pixel data with shorter exposure time) before final image merging. This preliminary action establishes initial exposure parameters that guide the subsequent merging process, reducing the time needed to process all pixel data while still achieving high dynamic range results.
Solution Approach 2:
The patent implements dynamic exposure adjustment by continuously updating exposure parameters based on real-time analysis of incoming pixel data. The system can rapidly adapt to changing illumination conditions by adjusting the second exposure time based on convergence results from the first pixel data, enabling fast response to drastic illumination changes.
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
The solution allows for rapid adjustment of exposure times and aperture sizes, resulting in clear and vivid image presentation under sudden light changes, improving image clarity and stability.
Implementation Method 1
a photodiode
Data Source
AI summary
The invention relates to an image sensing device and an image sensing method thereof. The image sensing device includes: an image sensing array and an image processing circuit. The image sensing array generates multiple pixel data, and the pixel data includes a first pixel data and a second pixel data. The first pixel data is generated by exposing for a first exposure time, the second pixel data is generated by exposing for a second exposure time, and the first exposure time is shorter than the second exposure time. The image processing circuit executes an automatic exposure convergence operation with the first pixel data to generate new exposure parameters for the automatic exposure convergence operation. The second exposure time is adjusted according to the new exposure parameters, so that the second pixel data is pixel data with changed exposure time.


