Advanced Project II

A Computer Vision Approach for Fumarole Detection And Classification

The developed fumarole detection and classification pipeline for Jacobs University Advanced Project II module in MSc Data Engineering.

Requirements

This is a CMake, C++17 based program. The following libraries are required:

1. OpenCV 4.1+
2. Boost 1.60+
3. Eigen3

Installation

Using CMake as the build system:

git clone https://github.com/asadahmedde/advanced-project-2.git
cd advanced-project-2
cd fumarole_localization
mkdir build && cd build
cmake ..
make fumarole_localization

Usage

./fumarole_localization thermal_images_dir (optional_output_dir)

Note: The images must be in greyscale.

CSV files will be saved in the provided output directory or in detector_csv_output in the executable directory by default. Each CSV file will be named with the corresponding name of the image file in the input directory and will contain a list of bounding boxes for the detections + the class label.

License

MIT