Setup & Installation

Prerequisites

  • Python 3.7+

  • pip (Python package installer)

  • Git (for cloning the repository)

  • (Optional for CARLA mode) CARLA Simulator installed and running.

Cloning the Repository

git clone https://github.com/batman-911/ADAS.git
cd ADAS # Or your project's root directory

Installing Dependencies

The project uses a requirements.txt file to manage Python dependencies.

pip install -r requirements.txt

This will install packages like OpenCV, PyTorch, PyQt6, Ultralytics, NLTK, etc.

Downloading Pre-trained Models

The system relies on pre-trained models for perception tasks. A script is provided to download these models.

chmod +x download_models.sh
bash download_models.sh

This script (download_models.sh) downloads models like best.pt (for object detection), crestereo_combined_iter2_480x640.onnx (for disparity estimation), and ultra_falst_lane_detection_culane_288x800.onnx (for lane detection) into the wights/ directory.

Setting Up Environment Variables

A setup script is provided to add the project directory to PYTHONPATH.

# Make sure you are in the project's root directory
chmod +x setup.sh
bash setup.sh

This script (setup.sh) primarily ensures that Python can find the project’s modules.

NLTK Data for Chatbot

The NLP chatbot requires NLTK data (punkt, wordnet). The NLP/adas_chatbot.py script attempts to download these automatically upon first run if they are missing.