Global Shutter Camera Calibration for PRNU Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cameras with global shutters suffer from photo response nonuniformity (PRNU) due to variations in pixel values caused by parasitic light sensitivity and amplification gain, which can result in suboptimal image quality, especially at short exposure times, and conventional shielding methods are costly.
Innovation Solution
Applying correction factors to pixel values using a calibration process to account for dark signals, parasitic light sensitivity, and system gain, computed through multiple data frames captured with varying integration times, normalizing values based on early-read pixels to correct for PRNU.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physical shielding is used to minimize light reaching memory nodes, then PRNU is reduced, but manufacturing cost increases significantly
Solution Approach 1:
The patent replaces the mechanical/physical shielding system with a computational correction system. Instead of using physical structures (shielding) to prevent parasitic light from reaching memory nodes, the invention uses digital signal processing to measure and correct the effects of parasitic light through calibration procedures and correction factors applied during image processing.
Solution Approach 2:
The patent creates a computational model (correction matrix) that replicates and compensates for the unwanted effects of parasitic light. By capturing calibration data under controlled conditions and creating correction factors, the system copies the characteristics of the distortion and applies inverse corrections, avoiding the need for physical intervention.
2Productivity
If exposure time is decreased, then capture speed increases, but PRNU becomes worse
Solution Approach 1:
The patent implements a feedback mechanism where the system measures the actual effects of parasitic light through calibration procedures and uses this information to adjust and correct pixel values. The correction factors are derived from measured data and applied back to the image processing pipeline, creating a closed-loop system that compensates for PRNU regardless of exposure time.
Solution Approach 2:
The patent performs preliminary calibration measurements to characterize the parasitic light effects before actual image capture. By pre-computing correction factors based on calibration data collected under various conditions, the system prepares correction parameters in advance that can be applied during normal operation, enabling fast capture without suffering from PRNU.
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 effectively addresses PRNU, resulting in improved image uniformity and quality, suitable for applications requiring high-quality images without the high cost of physical shielding.
Implementation Method 1
a sensor node accumulates charge when 'on' and a memory node retains the charge until its value is readout from the memory node
Data Source
AI summary
Technologies are described herein that are configured to generate a corrected image by addressing photo response nonuniformity (PRNU) in a camera having a global shutter. A calibration procedure is described, where correction factors for each pixel in an image sensor are computed and subsequently employed to generate improved images.