Image Enhancer AI (.NET Edition)

A cross-platform desktop application and web API for enhancing images using Real-ESRGAN AI models and converting them to modern, efficient formats like WebP and AVIF.

Features

Configuration Profiles: Save and load entire processing configurations for repeatable workflows.
Auto-Load Last Profile: The application automatically loads your last-used profile on startup.
AI-Powered Upscaling: Enhance image resolution by 4x using state-of-the-art Real-ESRGAN models.
Modern Format Conversion: Convert images to high-efficiency formats like .webp and .avif.
Batch Processing: Process entire folders of images, including subfolders, in one go.
Cross-Platform: Runs on Windows, macOS, and Linux.
Dual Interface: Use the simple and intuitive desktop GUI (built with Avalonia) or the headless Web API for programmatic access.
Detailed Progress Display: Track overall and per-folder progress with details on file counts and space savings.
UI/UX Enhancements:
- Drag and Drop support for selecting folders.
- Maximized window on startup for a better view.
- Dark & Light themes.
- Input validation and clear user feedback.
Processing Log: Keep track of all processed images.

Profile Management

The application now supports saving your settings into named profiles.

Saving a Profile: Configure your desired settings (folders, model, options) on the main window. Type a name in the text box next to the "Save as Profile" button and click it. The button will be disabled if the name is empty.
Loading and Deleting Profiles: Click the "Profiles..." button at the top right to open the Profile Manager. From there, you can select a profile to either load its settings or delete it permanently.

Solution Structure

The solution is divided into four projects:

ImageProcessor.Core: A .NET library project that contains the core logic for image processing. It orchestrates the command-line tools to perform image enhancement and conversion.
ImageProcessor.Api: An ASP.NET Core web API that exposes the image processing functionality as a web service.
ImageProcessor.UI: A desktop application built with Avalonia UI that provides a graphical user interface.
image-processor-web: A React-based web application that consumes the API.

How it Works

The application uses a pipeline of command-line tools to process images:

Upscaling (Optional): If enabled, the source image is first passed to realesrgan-ncnn-vulkan to be upscaled.
Conversion (Optional): The (potentially upscaled) image is then converted to either .webp using cwebp.exe or .avif using ffmpeg.exe.
File Handling: The application manages the creation of output folders and the deletion of intermediate and source files.

The API uses a job-based system to handle processing requests asynchronously.

API Endpoints

The API provides endpoints for starting and monitoring processing jobs.

`POST /api/Processing/start`

Starts a new image processing job. The request body must be a JSON object with the processing options. The response will contain a jobId that can be used to track the status of the job.

Example JSON Body:

{
  "inputFolder": "C:\\path\\to\\your\\images",
  "outputFolder": "C:\\path\\to\\output",
  "model": "realesrgan-x4plus",
  "processSubfolders": true,
  "convertToWebP": true,
  "convertToAvif": false,
  "applyUpscale": true,
  "deleteSourceFile": false,
  "includeWebPFiles": false,
  "includeAvifFiles": false
}

`GET /api/Processing/{jobId}/status`

Gets the current status of a processing job.

`GET /api/Processing/{jobId}/history`

Gets the complete progress history of a processing job.

Screenshots

(Note: The screenshot below is outdated and will be updated soon to reflect the latest UI changes.)

Processing Options Explained

The main window provides several options to customize the image processing workflow.

Option	Description
Use input folder as output folder	If checked, the output files will be saved in the same folder as the input files. This disables the output folder selection.
AI Model	Selects the Real-ESRGAN AI model to use for the upscaling process. Different models are trained for specific types of content.
Show Preview	A `ToggleSwitch` to show or hide the image preview panel.
Apply 4x Upscale	If checked, the AI-powered upscaling process is applied to the images. If unchecked, the application will only perform format conversion.
Convert to .WebP	If checked, the final image will be converted to the `.webp` format.
Convert to .AVIF	If checked, the final image will be converted to the `.avif` format using `ffmpeg`.
Process Subfolders	If checked, the application will search for images in all subdirectories of the selected input folder.
Include .WebP Files	If checked, existing `.webp` files in the source folder will be included in the processing queue.
Include .Avif Files	If checked, existing `.avif` files in the source folder will be included in the processing queue.
Delete Source Files	(Use with caution!) If checked, the original source image will be permanently deleted after it has been successfully processed.

💡 Note
Conversion quality settings for WebP and AVIF can be found in the Settings window (click the ⚙️ icon).

Settings

The settings window allows for more advanced configuration.

Setting	Description
WebP Quality	A slider (1-100) to control the quality of the WebP conversion. Higher values result in better quality and larger file sizes.
AVIF Quality	A slider (1-100) to control the quality of the AVIF conversion. Higher values result in better quality and larger file sizes.
Real-ESRGAN Command Line Arguments	Allows customization of the command-line arguments passed to the `realesrgan-ncnn-vulkan` executable for advanced users.

Prerequisites and Downloads

💡 Note
For now, all Microsoft Windows prerequisites are included in the release package, available inside the compressed file: [https://github.com/specializeddevel/ia_image_enhancer_csharp/releases]

To build and run this project from source, you will need the .NET 9 SDK.

Download: Download .NET 9 SDK

Required Files

For the application to function correctly, you need to either build from source and acquire the files listed below, or use a pre-compiled version.

Folder Structure Example (Windows)

This is an example of the required folder structure for running ImageProcessor.UI from the packaged release on Windows. The same structure applies to ImageProcessor.Api.

ImageProcessor.UI
└─ models/                         # realesrgan models
    └─ realesr-animevideov3-x2.bin
    └─ ...      
├─ cwebp.exe                       # Compression Windows executable
├─ ffmpeg.exe                      # Ffmpeg Windows executable
├─ ImageProcessor.UI.exe           # Main ImageProcessor User Interface executable.
└─ realesrgan-ncnn-vulkan.exe      # AI model Windows executable

If you are running the project from source, make sure the required files follow this structure:

ia_image_enhancer_csharp/
└─ ImageProcessor.UI/          
   └─ bin/
      └─ Debug/
         └─ net9.0/
            └─ models/                     # realesrgan models
               └─ realesr-animevideov3-x2.bin
               └─ ...      
            ├─ cwebp.exe                   # Compression Windows executable
            ├─ ffmpeg.exe                  # Ffmpeg Windows executable
            ├─ ImageProcessor.UI.exe       # Main ImageProcessor User Interface executable.
            └─ realesrgan-ncnn-vulkan.exe  # AI model Windows executable

💡 Note The /bin folder will be generated automatically after the first successful build.

1. External Binaries

This application relies on external command-line tools. These must be placed in the main application directory.

Real-ESRGAN (realesrgan-ncnn-vulkan):
- Repository: https://github.com/xinntao/Real-ESRGAN-ncnn-vulkan
- Download: Download the latest release for your platform from the releases page.

⚠️ Important
GPU Requirement: AI upscaling requires a graphics card that supports Vulkan 1.1 or higher.

cwebp (WebP Converter):
- Repository: [https://developers.google.com/webp]
- Download: Download the libwebp package for your platform from the downloads repository.
FFmpeg (AVIF Converter):
- Website: https://ffmpeg.org/
- Download: Download a release build for your platform from the official download page.

💡 Note
Linux/macOS Permissions: On Linux and macOS, you must grant execution permissions to the binaries.
chmod +x ./realesrgan-ncnn-vulkan
chmod +x ./cwebp
chmod +x ./ffmpeg

2. Models Folder

You must have a folder named models in the application's root directory. This folder must contain the .bin and .param files for the Real-ESRGAN models you intend to use.

Getting Started

Clone the repository:

git clone https://github.com/specializeddevel/ia_image_enhancer_csharp.git
cd ia_image_enhancer_csharp

Add Required Files: Follow the instructions in the Required Files section to download and place the necessary binaries and models.
Build the project:
```
dotnet build
```

Run the application:

To run the Desktop UI:
```
dotnet run --project ImageProcessor.UI
```
To run the Web API:
```
dotnet run --project ImageProcessor.Api
```

Technologies Used

.NET 9 / C#: The core application logic and UI are built with the latest version of .NET.
Avalonia UI: A cross-platform UI framework for creating the desktop application.
ASP.NET Core: Used to create the Web API for headless processing.
MVVM Pattern: The Model-View-ViewModel pattern is used to structure the UI code, powered by the CommunityToolkit.Mvvm library.
Swagger / OpenAPI: Used for API documentation and testing in the ImageProcessor.Api project via the Swashbuckle.AspNetCore library.

Contributing

Contributions are welcome! If you have a feature request, bug report, or pull request, please feel free to open an issue or submit a PR.

We follow a feature-branch workflow with Pull Requests for merging changes into the main branch.

License

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

Disclaimer

This project uses external, pre-compiled binaries for image processing. These binaries are provided by their respective authors and are subject to their own licenses. Please ensure you comply with their terms of use.

Name		Name	Last commit message	Last commit date
Latest commit History 48 Commits
ImageProcessor.Api		ImageProcessor.Api
ImageProcessor.Core		ImageProcessor.Core
ImageProcessor.UI		ImageProcessor.UI
image-processor-web		image-processor-web
models		models
.gitignore		.gitignore
ImageProcessor.sln		ImageProcessor.sln
README.md		README.md

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Repository files navigation

Image Enhancer AI (.NET Edition)

Table of Contents

Features

Profile Management

Solution Structure

How it Works

API Endpoints

`POST /api/Processing/start`

`GET /api/Processing/{jobId}/status`

`GET /api/Processing/{jobId}/history`

Screenshots

Processing Options Explained

Settings

Prerequisites and Downloads

Required Files

Folder Structure Example (Windows)

1. External Binaries

2. Models Folder

Getting Started

Technologies Used

Contributing

License

Disclaimer

About

Uh oh!

Releases

Packages

Languages

specializeddevel/ia_image_enhancer_csharp

Folders and files

Latest commit

History

Repository files navigation

Image Enhancer AI (.NET Edition)

Table of Contents

Features

Profile Management

Solution Structure

How it Works

API Endpoints

POST /api/Processing/start

GET /api/Processing/{jobId}/status

GET /api/Processing/{jobId}/history

Screenshots

Processing Options Explained

Settings

Prerequisites and Downloads

Required Files

Folder Structure Example (Windows)

1. External Binaries

2. Models Folder

Getting Started

Technologies Used

Contributing

License

Disclaimer

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Languages

`POST /api/Processing/start`

`GET /api/Processing/{jobId}/status`

`GET /api/Processing/{jobId}/history`

Packages