Shoplift Detection System

This repository contains a Python-based project that utilizes YOLOv8 (You Only Look Once) object detection model for detecting objects from videos, RTSP streams, or GIF files. It provides a user-friendly Graphical User Interface (GUI) built using Tkinter.

Demo

Check out the Shoplift Detection System in action below:

Features

• Detect objects in videos in real-time.
• Supports detection in video files (.mp4, .avi) or GIF files (.gif).
• Accepts RTSP stream URLs for live detection.
• Displays detection results visually, allows saving annotated frames to the tested/ directory.

---

Requirements

Before running the application, ensure you have installed all necessary dependencies.

1. Python 3.7+ installed on your system.
2. Install required Python libraries using pip:

pip install ultralytics opencv-python-headless tk pillow

---

How It Works

1. Model Integration: The application uses a pre-trained YOLOv8 model (best_model_v7.pt) to detect objects.
2. GUI Features:
   • Upload video or GIF file, process it, and display real-time detections.
   • Enter an RTSP stream URL to analyze the live stream.
3. Frame Output: The app saves all processed and annotated frames in the tested directory.

---

Installation Steps

1. Clone the repository:

   git clone https://github.com/chetanmundhe2911/shoplifting_detection.git
   cd shoplift-detection-system

2. Place the YOLOv8 model file best_model_v7.pt in the same directory as the script.
3. Install dependencies:

   pip install -r requirements.txt

4. Run the application:

   streamlit run shoplift_detection.py

---

Usage Instructions

1. Uploading a Video or GIF:

• Click the "Upload Video/GIF" button in the GUI.
• Select a video file or GIF file from your system.
• For videos:
  • The program processes the file frame-by-frame using the YOLOv8 model.
  • Annotated frames (with detection boxes) are displayed in real-time.
  • Frames with detections are saved in the tested/ directory.
• For GIFs:
  • The GIF is displayed in a separate window as-is (no detection is currently conducted for GIFs).

2. RTSP Stream:

• Enter the RTSP stream URL in the RTSP Stream URL input field.
• Click the "Start RTSP Stream" button and watch real-time detection results.

---

Adding GIF Demo Preview on GitHub Repo

To display a demo GIF on your main README.md file, follow these steps:

1. Create or record a demo GIF showcasing your application functionalities.

2. Save the GIF (e.g., demo.gif) in your repository's root directory.

3. Add the following markdown code in the README.md file to embed:

   ## Demo
   
   ![Shoplift Detection Demo](demo.gif)

Example Embedded Demo

---

Folder Structure

shoplift-detection-system/
├── best_model_v7.pt         # YOLOv8 trained weights
├── shoplift_detection.py    # Main application script
├── requirements.txt         # Python dependencies
├── README.md                # Documentation (current file)
├── demo.gif                 # Demo GIF showcasing app usage
└── tested/                  # Directory for saving annotated frames

---

License

This project is open-source under the MIT License. Feel free to use, modify, and distribute as needed.

---

Author

Developed by Chetan Mundhe.

Contributions and feedback are welcome! 😊