🧠 CNN-Based Framework for Missing Person Identification and Violence Detection Using Deep Learning and Distributed Processing

📌 Project Title

CNN-Based Framework for Missing Person Identification and Violence Detection Using Deep Learning and Distributed Processing

🧠 Introduction / Overview

This project leverages deep learning and distributed computing to solve two critical problems in surveillance systems:

1. Missing Person Identification – Detect and recognize missing individuals using CNN-based face recognition.
2. Violence Detection – Analyze video frames to detect violent activities using a trained CNN model.

The project integrates a user-friendly UI with a backend powered by Apache Spark for distributed processing, making it scalable for large-scale surveillance footage analysis.

---

⚙️ Working / Functionality

Core Features:

• Face Recognition: Detect and match faces using a trained CNN model.
• Violence Detection: Classify actions as violent or non-violent using video analysis.
• UI Panel: A simple Python UI for interacting with both functionalities.
• Distributed Processing: Spark-based data processing for efficient performance on larger datasets.
• Report Generation: Create summaries/statistics of detections.

Components:

• ui.py: Main UI interface.
• missing_person_detection.py: Contains face recognition logic.
• violence_detection.py: Contains violence classification logic.
• spark_processing.py: Spark-based processing logic.
• report_generation.py: Generates detection reports.
• start_cluster.py: Starts Spark cluster for distributed mode.

---

🧱 Architecture / Flow

                 ┌────────────────────────────┐
                 │       User Interface       │
                 │     (Tkinter-based UI)     │
                 └────────────┬───────────────┘
                              │
                 ┌────────────▼───────────────┐
                 │ Input Media (Images/Videos)│
                 └────────────┬───────────────┘
                              │
      ┌───────────────────────▼──────────────────────┐
      │            Preprocessing Module              │
      │  (Resizing, Normalization, Face Extraction)  │
      └───────────────────────┬──────────────────────┘
                              │
 ┌────────────────────────────▼──────────────────────────────┐
 │                  Deep Learning Inference                  │
 │  ┌────────────────────────┐  ┌──────────────────────────┐ │
 │  │  Face Recognition (CNN)│  │ Violence Detection (CNN) │ │
 │  └────────────────────────┘  └──────────────────────────┘ │
 └────────────────────────────┬──────────────────────────────┘
                              │
                ┌─────────────▼─────────────┐
                │ Distributed Processing    │
                │     (Apache Spark)        │
                └─────────────┬─────────────┘
                              │
                ┌─────────────▼─────────────┐
                │     Report Generation     │
                │    (PDF, Stats, Logging)  │
                └─────────────┬─────────────┘
                              │
                ┌─────────────▼─────────────┐
                │      Output Results       │
                │   (UI + Logs + Reports)   │
                └───────────────────────────┘

---

🔧 Technologies Used

• Python
• OpenCV
• PyTorch / facenet-pytorch
• Apache Spark (PySpark)
• Tkinter (for UI)
• NumPy, Matplotlib
• FPDF (PDF Report)

---

🛠️ Setup Instructions

🔹 Prerequisites

• Python 3.8+
• Java 8+
• Apache Spark
• Virtual environment (recommended)

🔹 Install Dependencies

pip install -r requirements.txt

🙌 Credits

Project Developer
B Akash Krishna – B.Tech AI & ML, KTU – 6th Semester
📧 Email: akash199699@gmail.com
🔗 GitHub: @akash199699

This project was developed as part of the mini project under the university curriculum. Special thanks to our mentors for their guidance, and teammates Anandhu S Kumar and Jewel Saji for their collaboration.