Skip to content

This GUI code base can be used to identify dust clouds from a Video or an Image.

License

Notifications You must be signed in to change notification settings

RajithaRanasinghe/Dust-Cloud-Identification

Repository files navigation

Dust Cloud Identification GUI Widget

Dust Cloud Identification Widget can be used to identify and approximately quantify dust clouds emitted by traffic.

Windows executale can be downloaded from This Link

Test Environment : Ryzen 9 5900HX, RAM 32GB, Nvidia RTX3080 Mobile 32GB

Step 1 : Load Image or Video you want to segmentate dust

APP Screenshot 1

Step 2 : Select the Machine Learning Model

APP Screenshot 2

Step 3 : Run !!!

How to create executable ?

  1. Download and install Python

  2. Create python vitual environmnet (environment name = env),

venv

python -m venv env
  1. Activate vitual environmnet, Navigate to the '\env\Scripts'
\env\Scripts\activate.bat
  1. Navigate to 'Dust-Cloud-Identification' folder, and Install required packages from requirements.txt,

requirements

python.exe -m pip install --upgrade pip
pip install -r requirements.txt
  1. Create executable using 'cxfreeze', cxfreeze
cxfreeze -c main.py --icon=icon.ico --target-dir dist --packages=torch --target-name=Dust-Cloud-Identification --base-name Win32GUI
  1. Copy 'app_data', 'output' folders and style.qss file into the 'dist' folder,

  2. Download pre-trained ML models from 'model_download_links.txt' and paste them in 'dist' folder,

  3. Run Dust-Cloud-Identification.exe

How to run code from source files ?

  1. Follow step 1 to 4 from above,

  2. Navigate to Dust-Cloud-Identification' folder and run 'main.py',

main.py

python main.py

Research Paper:

Beyond Conventional Monitoring: A Semantic Segmentation Approach to Quantifying Traffic-Induced Dust on Unsealed Roads