LADS-License-Authorization-and-Driver's-Saftey

Check Project details folder for the Detailed information about this project

This is the repository of my project LADS.
Theme : Prevent Accidents due to unauthorized drivers and drowsy driving.
Product : Mobility service providers can get the driver's quality history directly from the data and improve their employees.

For any commercial usage,Read License file!

Download the Necessary Large files Here

Link for files --> Store the downloaded files in the same directory where the repository is cloned.
Link for files(2) --> This is Entire Project, Extract the Files for usage

Recommended Configuration:

4 GB Ram, 4-Core GPU , Storage-150MB

Minimum Configuration:

4 GB Ram,4-Core CPU, Storage - 150MB

INSTALLATION GUIDE:

Use Anaconda for better experience:

git clone https://github.com/saikrithik/LADS-License-Authorization-and-Drivers-Saftey.git
cd LADS-License-Authorization-and-Drivers-Saftey

1. create a virtual environment (Recommended)

conda create -n environment_name python=3.7

(Note) Add environment to jupyter

python -m ipykernel install --user --name=firstEnv

2. activate virtual environment

conda activate environment_name

2. Install requirements.txt

pip install cmake
conda install -c conda-forge dlib
pip install -r requirements.txt

Changes needed to be done before usage.

(Environment Python-3.7)

1. All the paths needed to be changed according to your requirements.
2. Install all the required packages.

Usage Guide.

Single time usage programs

1. To store the driver's/ user's image locally use StoringImages.py(Each click will correspond to each stored photo).
2. Train the image using TrainingImages.py(It trains all the photos that are stored in the database).
   NOTE: (scipy version should be scipy==1.1.0).

You can also skip the above step and directly run the appication and make the storing and training from the below application itself

3. Run Final.py program for usage of LADS application.

Working Photos

Detects Face(IF USER is Authorized it detects eye landamrks)
Data is Encoded at the User end. Only admin can see the history of the user.

If USER is not Authorized It shows following options.

Credits

Thanks to Adrian Rosebrock , sentedex and their amazing tutorials that helped me build this project.

Tools used

1. Python for programming
2. OpenCV,Dlib,FaceRecognition,numpy,Haar-Cascades
3. TransferLearning is used for Modeling