Using TensorFlow Library, this project is on an image classifier using Convolutional Neural Network for traffic signs in Singapore. The aim of the model is to find applicability in the development of Autonomous vehicles in identifying traffic information during drives excluding highway. The final model pipeline can be found in the file name "Final_Model_Pipeline.ipynb". Instructions:

• Upload a photo image into the folder called "taget"
• Run the notebook containing the final model pipeline
• There are two models avaiable to choose from: Adam Optimizer Model and RMS Propagation Optimizer Model

Label	Signs
0	END OF EXP WAY
1	NO JAYWALKING
2	TP CAMERA ZONE
3	SPD LIMIT 90
4	TURN LEFT
5	ERP
6	SPD LIMIT 70
7	U TURN LANE
8	SPLIT WAY
9	STOP
10	SPD LIMIT 50
11	CURVE RIGHT ALIGNMENT MARKER
12	ZEBRA CROSSING
13	RAIN SHELTER
14	NO ENTRY
15	KEEP LEFT
16	PARKING AREA FOR MOTORCARS
17	PEDESTRIAN USE CROSSING
18	RESTRICTED ZONE AHEAD
19	CURVE LEFT ALIGNMENT MARKER
20	START OF EXP WAY
21	GIVE WAY
22	NO VEH OVER HEIGHT 4.5
23	SPD LIMIT 40
24	SLOW SPEED
25	ROAD HUMP
26	NO LEFT TURN
27	ONE WAY RIGHT
28	ONE WAY LEFT
29	SLOW DOWN
30	MERGE
31	NO RIGHT TURN

There are two code-base jupyter notebooks:

1. Image processing - Preprocessing of images to prepare different types of processed images to be fit to the model
2. Development of model - Developed a CNN to be trained on various types of processed images from raw (gray scaled) to normalised(color scaled)

Libraries need for code to run, please ensure to install any missing modules using: "pip install " import pandas as pd import random import pickle import tensorflow as tf import keras from keras.models import Sequential from keras.layers import Dense, Flatten,BatchNormalization, Dropout, Lambda, Conv2D, MaxPool2D from tensorflow.keras.utils import plot_model import numpy as np import matplotlib.pyplot as plt import cv2 (pip install opencv-python) import os from sklearn.preprocessing import LabelEncoder