Digital Camera FPN Correction Using Preliminary Calibration Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for correcting Fixed Pattern Noise (FPN) in digital cameras suffer from estimation errors due to noise in calibration images, leading to insufficient noise reduction and residual vertical streak noise in corrected images, especially when the number of pixels used for estimation is limited to maintain shooting speed.
Innovation Solution
A signal processing method that attenuates the estimated noise shape using a calibration image to calculate a correction value, which is then applied to the image to be corrected, reducing the impact of estimation errors and improving noise correction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a sufficient number of pixels are used for FPN estimation to reduce noise, then FPN estimation accuracy is improved, but release time lag increases and continuous shooting speed decreases
Solution Approach 1:
The patent performs FPN correction using a small number of pixels during the shooting phase to maintain high continuous shooting speed. The FPN correction coefficients are preliminarily calculated during a separate calibration phase when time is not constrained. This separates the time-consuming calculation from the time-critical shooting operation, resolving the contradiction between estimation accuracy and shooting speed.
2Productivity
If a small number of pixels are used for FPN estimation to maintain shooting speed, then continuous shooting speed is maintained, but FPN estimation accuracy deteriorates and vertical streak noise remains
Solution Approach 1:
The patent introduces FPN correction coefficients as an intermediary element. These coefficients are calculated during calibration using sufficient pixels for high accuracy, then stored and applied during shooting using only a small number of pixels. This intermediary allows the system to achieve high shooting speed while maintaining accurate FPN correction, as the heavy calculation is performed once during calibration rather than repeatedly during shooting.
3Speed
If conventional FPN estimation is performed with limited pixels, then processing speed is maintained, but estimation error from noise in calibration image cannot be sufficiently reduced
Solution Approach 1:
The patent performs preliminary FPN correction coefficient calculation during a dedicated calibration phase where processing speed is not constrained. The coefficients are calculated once with high reliability using sufficient pixels, then reused during actual shooting where speed is critical. This preliminary action separates the reliability-critical calculation from the speed-critical application, resolving the contradiction between processing speed and correction reliability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An estimated noise shape estimated to be included in the signal to be corrected is calculated based on a calibration signal including a noise correlating with the noise of the signal to be corrected so as to correct a noise of a signal to be corrected. A correction value of a noise shape is generated by attenuating an amplitude of the estimated noise shape, and the noise of the signal to be corrected is corrected by using the correction value of the noise shape thus generated.