Noise parameter estimation for poisson corrupted images using variance stabilization transforms.

Area of Science:

• Image processing
• Computational imaging
• Statistical modeling

Background:

• Real-world image sensors inherently capture noise.
• Photon arrival processes are often modeled using Poisson distribution, matching pixel value distributions.
• Accurate noise parameter estimation is crucial for image quality.

Purpose of the Study:

• To propose a novel method for estimating unknown noise parameters in Poisson corrupted images.
• To leverage properties of variance stabilization for noise parameter estimation.
• To achieve accurate and stable noise parameter estimation with lower computational complexity.

Main Methods:

• Utilizing variance stabilization properties to develop a noise parameter estimation technique.
• Applying the method to Poisson corrupted images.
• Comparing the proposed method's accuracy and stability against state-of-the-art techniques.

Main Results:

• The proposed method accurately estimates unknown noise parameters from Poisson corrupted images.
• The technique demonstrates significantly lower computational complexity compared to existing methods.
• Improved stability in estimation was observed, yielding comparable accuracy to state-of-the-art approaches.

Conclusions:

• The developed method provides an efficient and stable approach for estimating noise parameters in images with Poisson noise.
• This technique offers a valuable alternative for applications requiring accurate noise characterization in digital imaging.