APN: Adaptive Percentage Normalization

The IPIC-APN User Interface

Introduction

IPIC-APN is an ImageJ/Fiji plugin specifically designed for unsupervised learning denoising.

It utilizes the Adaptive Percentage Normalization (APN) algorithm to prepare raw image data for downstream analysis. By automatically calculating optimal intensity thresholds, this plugin effectively mitigates the influence of baseline drift and hot pixels, providing a cleaner and more consistent input for deep learning models or advanced analysis.

📥 Installation

1. Download: Locate the file named Adaptive_Percentage-Normalization-1.0.0.jar in the file list above (the main directory of this repository) and download it.
2. Install: Copy the downloaded .jar file into the plugins/ folder of your ImageJ or Fiji directory.
3. Restart: Restart ImageJ/Fiji to complete the installation.

🚀 Usage

1. Open Image: Load your target image or stack.
2. Run Plugin: Navigate to the menu bar: Plugins > APN Tool > Run APN.
3. Process:
   • Click Start Processing.
4. Output:
   • The plugin generates a normalized version of the image/stack.
   • It also displays the histogram with the calculated percentile thresholds used for the normalization.

📊 Comparison: Raw vs. APN

Below is a comparison showing the removal of baseline noise and hot pixel artifacts.

Raw Image	APN Processed
Original image with baseline drift/hot pixels	Normalized output ready for training

🛠 Building from Source

This project is managed by Maven. To compile the plugin yourself:

1. Clone the repository:

   git clone [https://github.com/YourUsername/APN_Adaptive-Percentage-Normalization.git](https://github.com/YourUsername/APN_Adaptive-Percentage-Normalization.git)

2. Navigate to the project directory containing the pom.xml.
3. Build using Maven:

   mvn clean package

4. The compiled .jar file will be located in the target/ directory.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.