This repository contains the implementation of an innovative tool designed for real-time panoramic imaging in an optical microscopy environment. The tool aims to overcome limitations associated with conventional panoramic image construction techniques, offering an efficient and high-quality solution through the use of computer vision algorithms.

Motivation

In the field of optical microscopy, analysts commonly rely on a macroscopic reference frame to locate specific details in samples. Often, computational tools are employed to construct high-resolution panoramic images from lower-resolution captures. However, this traditional process is tedious, time-consuming, and often yields low-quality results in terms of image resolution.

Key Features

Continuous Image Acquisition: The tool employs an innovative strategy of continuously forming the image while traversing the sample by moving the microscope stage.
Panoramic Construction Algorithm: An advanced algorithm based on computer vision tools has been developed for efficient real-time panoramic image construction.
Testing and Validation: The repository includes tests conducted in simulated environments and a real optical microscope. Simulated tests have shown promising results, demonstrating the ability to reconstruct images accurately. Tests on the optical microscope have yielded encouraging results in terms of speed and image quality.
Identification of Intrinsic Aberrations: Tests on the optical microscope have also allowed the identification of sources of intrinsic instrument aberrations, contributing to improving the technique's robustness.

Compatibility

This project has been tested and is compatible with the following environment:

Operating System: Windows
Python Version: 3.x

Clonning de repository

1 - Clone the repository

2 - Create a virtual enviroment with:

python3 -m venv /path/name_venv

3 - Activate venv with (Windows):

.\.name_venv\Scripts\Activate.ps1

4 - Install dependencies

py -m pip install -r requirements.txt

User Interface

The tool features a user-friendly interface built with QML and Qt.

Video demo

To launch the tool with the graphical user interface, run:

cd micro_stitch

python main.py

Contributions and Issues

Contributions are welcome! Please fork the repository and create a pull request. If you encounter any issues, please open a new issue here.

License

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