SR-Wiki Normalization

The SR-Wiki Normalization User Interface

---

📖 Overview

SR-Wiki Normalization is an ImageJ/Fiji plugin designed for robust preprocessing of biological microscopy images, with a particular emphasis on low signal-to-noise ratio (low-SNR) fluorescence data.

Microscopy images often suffer from baseline drift, uneven background, hot pixels, and weak structural signals that are easily overwhelmed by noise. Conventional global normalization strategies such as Min-Max or Max normalization implicitly assume stable background and well-behaved intensity distributions, which are rarely satisfied in real experimental conditions.

To address this, SR-Wiki integrates a set of seven normalization and standardization methods within a unified interface, centered around the Adaptive Percentage Normalization (APN) algorithm. The plugin enables both display-oriented enhancement and analysis-oriented standardization, making it suitable for downstream tasks such as unsupervised deep learning denoising, quantitative measurement, and visualization.

---

📥 Installation

1. Download the latest srwiki_normalization-3.0.0.jar
2. Copy it into:ImageJ/Fiji/plugins/
3. Restart ImageJ/Fiji
4. Launch via:Plugins > SRWiki > SRWiki Normalization

---

⚙️ Usage Workflow

1. Open a grayscale image or stack in ImageJ/Fiji
2. Launch the plugin
3. Select a normalization or standardization method
4. If using Percentile mode, set lower and upper thresholds
5. Optionally enable histogram visualization
6. Click Run SRWiki
7. A processed image or stack will be generated

---

📊 Normalization and Standardization Modes

Mode	Mathematical Principle	Best Used For
APN (Adaptive)	$x' = \dfrac{x - X_{\min}^{\text{(bg)}}}{X_{\max}^{\text{(sig)}} - X_{\min}^{\text{(bg)}}} \times 255$	Low-SNR microscopy, real biological data
Percentile	$x' = \mathrm{clip}\left(\dfrac{x - P_{\text{low}}}{P_{\text{high}} - P_{\text{low}}},0,1\right)\times 255$	Controlled removal of outliers
Min-Max	$x' = \dfrac{x - x_{\min}}{x_{\max} - x_{\min}} \times 255$	Clean images with stable intensity range
Max Only	$x' = \dfrac{x}{x_{\max}} \times 255$	Pre-calibrated data with zero background
Z-Score	$z = \dfrac{x - \mu}{\sigma}$	Deep learning preprocessing
Mean	$x' = \dfrac{x - \mu}{x_{\max} - x_{\min}}$	Centered normalization for analysis
Vector	$x' = \dfrac{x}{|x|_2}$	Energy normalization / feature comparison

---

⚙️ Parameters

Percentile Parameters (active only in Percentile mode)

• Lower Percentile (%)

• Default: 1.0

• Intensities below this value are mapped to 0

• Upper Percentile (%)

• Default: 99.8

• Intensities above this value are mapped to 255

These parameters allow manual control over intensity clipping when automatic methods are not preferred.

---