🚦 Traffic Sign Classification & Real-Time Computer Vision

Computer Vision | MATLAB App Designer | Color Segmentation | Performance Metrics

📖 Overview

This project focuses on the detection and classification of road traffic signs in real-time using a standard webcam. Developed in MATLAB, the system employs advanced image processing techniques to isolate, identify, and validate traffic signs (STOP, Mandatory Left, Mandatory Right) based on color and shape analysis.

🛠️ Technical Features

• Real-Time Processing: Low-latency acquisition and processing of video frames.
• HSV Color Segmentation: Robust detection across different lighting conditions using dynamic thresholds.
• Feature Extraction: Analysis of object properties (Centroid, Bounding Box, Eccentricity) to distinguish between circular and octagonal signs.
• Interactive GUI: Built with MATLAB App Designer, allowing:
  • Live video feed with mask overlays.
  • Dynamic adjustment of HSV parameters.
  • Real-time classification results and historical log.

📈 Performance Analysis

The project includes a validation module that calculates:

• Precision & Recall: Measuring the accuracy of the detection engine.
• Confusion Matrix: Evaluating classification performance across different sign types.
• Distance Sensitivity: Analysis of how resolution and distance affect classification reliability.

📺 Project Demo

Watch the Real-Time Classification Video on YouTube

📂 Repository Structure

• /src: Contains the .mlapp source code and History.mat for metric logging.
• /docs: Full technical report detailing the vision algorithms and test results.
• /media: Links to video demonstrations and system screenshots.

👥 Author

• Tiago Oliveira