Traffic Sign Detection

This project implements the classification of of German Traffic Sign Recognition Benchmark dataset using various feature extraction methods and classification methods. The goal of this project was to achieve high accuracy in recognizing signs in real world.

Feature Extraction Methods

1. Raw pixels
2. Color histograms
3. Histogram of Gradients(HoG)

Classification Algorithms

1. Multi Layer Perceptron (MLP)
2. K-Nearest Neighbors (K-NN)
3. Support Vector Machines (SVM)
4. Random Forest
5. Convolutional Neural Network (CNN)