Object Detection and Tracking with YOLO and BetterCam

This repository contains a script for object detection and tracking using the YOLO model and BetterCam for capturing frames. The script performs real-time object detection, identifies different classes, and tracks objects across frames using ByteTrack.

Features

• Real-time Object Detection: Utilizes the YOLO model to detect various objects in real-time.
• Object Tracking: Tracks objects across frames using ByteTrack.
• Customizable Configuration: Allows setting box dimensions through a configuration file.
• Logging: Integrated logging for better tracking and debugging.

Requirements

• Python 3.11.6 (recommended)
• CUDA-compatible GPU (for running YOLO model on GPU)
• Python packages:
  • bettercam
  • Pillow
  • opencv-python
  • numpy
  • torch
  • ultralytics
  • configparser

Python Installation

It is recommended to use Python 3.11.6 for this project. You can download and install it from the following link:

Python 3.11.6 Installer

Installation

1. Download the repository:

   • Go to the repository page in your browser.
   • Click the green "Code" button.
   • Select "Download ZIP".
   • Extract the downloaded ZIP file to your desired location.

2. Navigate to the project directory:

   cd path/to/extracted/repo

3. Create a virtual environment and activate it:

   python -m venv venv
   

   On Linux / Mac Os use

   source venv/bin/activate  
   

   On Windows use

   `venv\Scripts\activate`
   

4. Install the required packages:

   pip install -r requirements.txt

   or

   • you can use the install_packages_windows.bat in Scripts Folder.

   or

   • you can use the install_packages_windows_venv.bat in Scripts Folder.

5. Ensure you have CUDA installed and configured for running the YOLO model on GPU.

Configuration

box_config.ini

Create a box_config.ini file to set the box dimensions for the region of interest:

[BoxDimensions]
size = 320

game.yaml

• The game.yaml file should be in the same directory and look like this:

# paths
path: game  # path to the dataset
train: images  # path to the training images
val: val  # path to the validation images

# class names
names:
  0: player
  1: bot
  2: weapon
  3: outline
  4: dead_body
  5: hideout_target_human
  6: hideout_target_balls
  7: head
  8: smoke
  9: fire
  10: third_person

Usage

• Run the script with:

  python main.py
  

Logging

• The script uses Python's built-in logging module to log important events. Logs will be printed to the console with timestamps, log levels, and messages.

Batch Script for Installation

• You can use the following batch script to install each package separately, but all at once.

install_requirements.bat

@echo off
echo Installing bettercam...
pip install bettercam
if %errorlevel% neq 0 (
    echo Failed to install bettercam
    exit /b %errorlevel%
)

echo Installing Pillow...
pip install Pillow
if %errorlevel% neq 0 (
    echo Failed to install Pillow
    exit /b %errorlevel%
)

echo Installing opencv-python...
pip install opencv-python
if %errorlevel% neq 0 (
    echo Failed to install opencv-python
    exit /b %errorlevel%
)

echo Installing numpy...
pip install numpy
if %errorlevel% neq 0 (
    echo Failed to install numpy
    exit /b %errorlevel%
)

echo Installing torch...
pip install torch
if %errorlevel% neq 0 (
    echo Failed to install torch
    exit /b %errorlevel%
)

echo Installing ultralytics...
pip install ultralytics
if %errorlevel% neq 0 (
    echo Failed to install ultralytics
    exit /b %errorlevel%
)

echo Installing configparser...
pip install configparser
if %errorlevel% neq 0 (
    echo Failed to install configparser
    exit /b %errorlevel%
)

echo All packages installed successfully.
pause

Acknowledgements

• YOLO by Ultralytics

• ByteTrack

• OpenCV

• PyTorch

• Rootkit-Org For always being there to educate and help

Contributing

• Contributions are welcome! Please open an issue or submit a pull request for any improvements or bug fixes.

• If you want to contribute: Recommended Python 3.11.6

• Python 3.11.6 Installer