Adaptive Compressive CMOS Image Sensor for High-Speed Acquisition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional CMOS image sensors face limitations in image acquisition rates and power consumption due to the lengthy process of acquiring and compressing entire digitized images, and existing compressive sensing methods do not offer optimal compression rates or image quality.
Innovation Solution
A CMOS image sensor that calculates statistical estimators in analog mode to adapt the compression rate block-by-block, using pseudo-random binary value generators with controllable output expectations to select pixels for reading and digitizing, thereby optimizing compression based on image sparseness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods acquire an entire digitized image before compression, then complete image data is captured, but image acquisition time increases and productivity decreases
Solution Approach 1:
The patent applies preliminary action by performing compressive sensing measurements during the image acquisition phase itself, rather than acquiring the complete image first and then compressing it. The sensor captures compressed measurements directly in analog domain before digitization, which preliminarily reduces the data volume and enables faster acquisition rates while still capturing sufficient information for image reconstruction.
2Measurement precision
If conventional methods acquire and store entire digitized images, then all image information is preserved, but electric power consumption increases
Solution Approach 1:
The patent extracts only the essential image information needed for reconstruction by performing compressive sensing measurements during acquisition. Instead of capturing and storing all pixel data, the system extracts a reduced set of measurements that contain sufficient information to reconstruct the image, thereby reducing the power consumption of readout and analog-to-digital conversion circuits while preserving necessary image information.
3Productivity
If pixel binning compressive sensing is used to decrease acquisition time, then image acquisition rate increases, but image quality may deteriorate due to loss of spatial resolution
Solution Approach 1:
The patent applies segmentation by dividing the image into multiple blocks and performing compressive sensing measurements on each block separately using different pseudo-random measurement matrices. This block-based approach allows the system to achieve fast acquisition rates through compressed sensing while maintaining image quality by preserving the spatial structure within each block and enabling independent optimization of measurement strategies for different regions.
4Device complexity
If fixed compression rate is applied to all image blocks, then device complexity is reduced, but adaptability to different image content decreases
Solution Approach 1:
The patent applies dynamics by making the compression rate adaptive based on the local characteristics of each image block. The system calculates statistical estimators (such as variance or gradient magnitude) for each block and dynamically adjusts the compression rate according to these estimators, allowing regions with important features to be captured at higher quality while allowing more aggressive compression in uniform regions, thus achieving both adaptability and reasonable complexity.
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 enhances image acquisition speed and reduces power consumption while maintaining or improving image quality by dynamically adjusting the compression rate according to statistical estimators, allowing for better compression efficiency and image reconstruction.
Implementation Method 1
Each pixel comprises a photodiode used in reverse mode, having its junction capacitor discharged by a photocurrent according to a received light intensity
Data Source
AI summary
A CMOS image sensor including: a plurality of pixels; a first analog circuit for calculating one or a plurality of statistical estimators based on the analog output values of sensor pixels; and a second circuit capable of implementing a compressive image sensing method, wherein the applied compression rate is a function of the statistical estimator(s) calculated by the first circuit.