nikithks / traffic-sign-recognition-using-cnn

Traffic Sign recognittion system using CNN model

Traffic-Sign-Recognition-using-CNN

Traffic Sign recognittion system using CNN model

In this project we developed a Traffic Sign Recognition model using Keras and CNN. The file 'tsr.pynb' contains the method we used for training the Rescognition model 'tsr-fin.h5'.

There are 2 interfaces available, both built using Tkinter. 'ui.py' is simple GUI that takes image as input and gives it's prediction as to which sign is in the image. 'sim.py' works like a simulator. It has a speedometer, that simulates speed of a vehicle and alters it acoordingly when a traffic sign is detected. 'Samples' folder contains sample images of all the signs that the model can identify.

Python packages required to run the simulator: Tkinter, math, random, time, threading, PIL, numpy, Keras.

Output Scrrenshots: