IP Webcam Stream Projects: A Journey Through Computer Vision 📸

This document outlines the various stages and features developed for the IP Webcam Stream application, demonstrating a progression from a basic stream to advanced real-time computer vision capabilities. Each project phase built upon the last, culminating in a versatile multi-feature application.

Project Overview & Evolution 📈

Phase 1: Basic IP Webcam Stream

• Purpose: The foundational project. Establishes a Flask web server to stream live video from your local webcam to a web browser. It provides a simple, low-latency video feed.
• Key Features: Live MJPEG streaming, snapshot capture, full-screen viewing.

Phase 2: Object Counting (Initial Iteration - Deprecated)

• Purpose: An early attempt to add computer vision by counting specific, static objects.
• Status: This specific iteration was superseded by more robust real-time object detection due to limitations in simple counting methods for complex scenes.

Phase 3: Real-time Object Detection

• Purpose: Enhanced the stream with dynamic, real-time object detection using a pre-trained deep learning model (MobileNet-SSD). This allows the application to identify and label multiple objects in motion or static within the video frame.
• Key Features: Real-time bounding boxes and labels for detected objects (e.g., bottles, books, persons, cars).

Phase 4: Real-time Age Detection

• Purpose: Shifted focus to human-centric computer vision, specifically detecting faces and estimating age ranges. This involves a two-step process: first, detecting faces, and then, analyzing those facial regions for age prediction.
• Key Features: Real-time face bounding boxes with predicted age ranges displayed on the stream.

Phase 5: Multi-Feature IP Webcam Stream (Current Version) ✨

• Purpose: The culmination of all previous efforts, this application integrates the basic streaming, real-time object detection, and real-time age detection into a single, user-selectable interface. Users can choose their desired mode directly from the web page.
• Key Features:
  • Selectable Modes: Switch between Basic Stream (No AI), Object Detection, and Age Detection.
  • Snapshot Functionality: Capture and download still images from any active stream.
  • Full Screen Mode: View the webcam feed in full-screen.
  • Dynamic UI Updates: Status messages and feature information update based on the selected mode and model loading status.
  • Webcam Offline Placeholder: Displays a clear "WEBCAM OFFLINE" message if the camera is not accessible.
  • Responsive Design: Styled with Tailwind CSS for basic adaptability.

---

Technologies Used 💻

• Python 3.x
• Flask: A micro web framework for handling web requests and serving the video stream.
• OpenCV-Python: The primary library for all computer vision tasks (webcam access, image processing, DNN inference).
• NumPy: Essential for numerical operations within OpenCV.
• HTML5, CSS3, JavaScript: For the interactive web interface.
• Tailwind CSS: For streamlined UI styling.
• Pre-trained Deep Learning Models:
  • MobileNet-SSD (Caffe model for Object Detection)
  • Caffe-based Face Detector (for Age Detection prerequisite)
  • Caffe-based Age Predictor Model

---

Prerequisites ✅

Before running the application, ensure you have:

• Python 3.x installed.
• pip (Python package installer) for library management.
• Install the necessary Python libraries:

    pip install opencv-python flask numpy

---

!!! CRITICAL STEP: Download All Required AI Models !!! ⚠️

To enable the Object Detection and Age Detection features, you MUST DOWNLOAD several pre-trained deep learning model files and place them in the SAME DIRECTORY as your Python script (ip_webcam_combined.py).

If these files are missing or corrupted, the respective AI features will be disabled, and you will see "Model Error" warnings in the console and on the web interface.

Required Model Files (Total 7 Files):

A) For Object Detection (3 files):

1. MobileNetSSD_deploy.prototxt (Model Architecture)
   • Direct Download
   • How to Download: Click the link. When the plain text content loads, right-click and select "Save As...". Save as MobileNetSSD_deploy.prototxt.
2. MobileNetSSD_deploy.caffemodel (Model Weights)
   • Direct Download
   • How to Download: Clicking this link should trigger the download (approx. 23MB).
3. coco.names (Class Labels)
   • How to Create: Open a text editor (Notepad). Paste the list of 21 objects (background, aeroplane... tvmonitor) and save as coco.names.

B) For Age Detection (4 files):

1. deploy.prototxt (Face Detector Architecture)
   • Direct Download
2. res10_300x300_ssd_iter_140000.caffemodel (Face Detector Weights)
   • Direct Download
3. age_deploy.prototxt (Age Predictor Architecture)
   • Direct Download
4. age_net.caffemodel (Age Predictor Weights)
   • Direct Download

Ensure ALL SEVEN downloaded files are in the SAME FOLDER as your Python script!

---

How to Run the Multi-Feature Application ▶️

1. Save the Python Script: Save the provided Python code as ip_webcam_combined.py.
2. Place Model Files: Ensure all seven model files are in the same directory as your script.
3. Run the Script:

   python ip_webcam_combined.py

4. Access the Web Stream: Open the URL provided in the console (e.g., http://127.0.0.1:5000) in your browser.
5. Select Mode: Use the dropdown menu to select a mode and click "Start Stream."

---

Troubleshooting 🔍

• "Model Error" Messages: Verify all seven files are in the same folder and named correctly.
• "Error initializing camera": Ensure no other apps (Zoom/Skype) are using the camera. Try changing cv2.VideoCapture(0) to index 1 or 2.
• Web page shows "Connecting..." indefinitely: Check the console for errors and test the webcam with a native camera app.
• SyntaxError: Ensure you have copied the entire Python code block exactly as provided.

---

Security Warning ⚠️

This application is for local use only. It does not implement authentication or HTTPS. DO NOT expose this stream to the public internet without implementing proper security protocols.