Image Classification Using Convolutional Neural Networks: A step by step guide

This article was published as a part of the Data Science Blogathon.

Introduction

Convolutional Neural Networks come under the subdomain of Machine Learning which is Deep Learning. Algorithms under Deep Learning process information the same way the human brain does, but obviously on a very small scale, since our brain is too complex (our brain has around 86 billion neurons).

Why CNN for Image Classification?

Image classification involves the extraction of features from the image to observe some patterns in the dataset. Using an ANN for the purpose of image classification would end up being very costly in terms of computation since the trainable parameters become extremely large.

For example, if we have a 50 X 50 image of a cat, and we want to train our traditional ANN on that image to classify it into a dog or a cat the trainable parameters become –
(50*50) * 100 image pixels multiplied by hidden layer + 100 bias + 2 * 100 output neurons + 2 bias = 2,50,302

We use filters when using CNNs. Filters exist of many different types according to their purpose.

Examples of different filters and their effects

Filters help us exploit the spatial locality of a particular image by enforcing a local connectivity pattern between neurons.

Convolution basically means a pointwise multiplication of two functions to produce
a third function. Here one function is our image pixels matrix and another is our filter. We slide the filter over the image and get the dot product of the two matrices. The resulting matrix is called an “Activation Map” or “Feature Map”.

Using a vertical filter here to convolve a 6X6 image

There are multiple convolutional layers extracting features from the image and finally the output layer.

Now there are a lot of other things such as channels, pooling, etc which go into the depth of the theory. Here we will concentrate on the practical.

PRACTICAL: Step by Step Guide

I will be working on Google Colab and I have connected the dataset through Google Drive, so the code provided by me should work if the same setup is being used. Remember to make appropriate changes according to your setup.

Step 1: Choose a Dataset

Choose a dataset of your interest or you can also create your own image dataset for solving your own image classification problem. An easy place to choose a dataset is on kaggle.com.

The dataset I’m going with can be found here.

This dataset contains 12,500 augmented images of blood cells (JPEG) with accompanying cell type labels (CSV). There are approximately 3,000 images for each of 4 different cell types grouped into 4 different folders (according to cell type). The cell types are Eosinophil, Lymphocyte, Monocyte, and Neutrophil.

Here are all the libraries that we would require and the code for importing them.

from keras.models import Sequential

import tensorflow as tf

import tensorflow_datasets as tfds

tf.enable_eager_execution()

from keras.layers.core import Dense, Activation, Dropout, Flatten
from keras.layers.convolutional import Convolution2D, MaxPooling2D
from keras.optimizers import SGD, RMSprop, adam
from keras.utils import np_utils
from sklearn.tree import DecisionTreeClassifier # Import Decision Tree Classifier
from sklearn import metricsfrom sklearn.utils import shuffle
from sklearn.model_selection import train_test_splitimport matplotlib.image as mpimg
import matplotlib.pyplot as plt
import numpy as np
import os
import cv2
import randomfrom numpy import *
from PIL import Image
import theano

Step 2: Prepare Dataset for Training

Preparing our dataset for training will involve assigning paths and creating categories(labels), resizing our images.

Resizing images into 200 X 200

path_test = "/content/drive/My Drive/semester 5 - ai ml/datasetHomeAssign/TRAIN"

CATEGORIES = ["EOSINOPHIL", "LYMPHOCYTE", "MONOCYTE", "NEUTROPHIL"]
print(img_array.shape)IMG_SIZE =200
new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))

Step 3: Create Training Data

Training is an array that will contain image pixel values and the index at which the image in the CATEGORIES list.

training = []def createTrainingData():
  for category in CATEGORIES:
    path = os.path.join(path_test, category)
    class_num = CATEGORIES.index(category)
    for img in os.listdir(path):
      img_array = cv2.imread(os.path.join(path,img))
      new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))
      training.append([new_array, class_num])createTrainingData()

Step 4: Shuffle the Dataset

random.shuffle(training)

Step 5: Assigning Labels and Features

This shape of both the lists will be used in Classification using the NEURAL NETWORKS.

X =[]
y =[]for features, label in training:
  X.append(features)
  y.append(label)
X = np.array(X).reshape(-1, IMG_SIZE, IMG_SIZE, 3)

Step 6: Normalising X and converting labels to categorical data

X = X.astype('float32')
X /= 255
from keras.utils import np_utils
Y = np_utils.to_categorical(y, 4)
print(Y[100])
print(shape(Y))

Step 7: Split X and Y for use in CNN

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 4)

Step 8: Define, compile and train the CNN Model

batch_size = 16
nb_classes =4
nb_epochs = 5
img_rows, img_columns = 200, 200
img_channel = 3
nb_filters = 32
nb_pool = 2
nb_conv = 3

model = tf.keras.Sequential([
    tf.keras.layers.Conv2D(32, (3,3), padding='same', activation=tf.nn.relu,
                           input_shape=(200, 200, 3)),
    tf.keras.layers.MaxPooling2D((2, 2), strides=2),
    tf.keras.layers.Conv2D(32, (3,3), padding='same', activation=tf.nn.relu),
    tf.keras.layers.MaxPooling2D((2, 2), strides=2),
    tf.keras.layers.Dropout(0.5),
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(128, activation=tf.nn.relu),
    tf.keras.layers.Dense(4,  activation=tf.nn.softmax)
])

model.compile(optimizer='adam',loss='sparse_categorical_crossentropy',metrics=['accuracy'])
model.fit(X_train, y_train, batch_size = batch_size, epochs = nb_epochs, verbose = 1, validation_data = (X_test, y_test))

score = model.evaluate(X_test, y_test, verbose = 0 )
print("Test Score: ", score[0])
print("Test accuracy: ", score[1])

 

In these 9 simple steps, you would be ready to train your own Convolutional Neural Networks model and solve real-world problems using these skills. You can practice these skills on platforms like Analytics Vidhya and Kaggle. You can also play around by changing different parameters and discovering how you would get the best accuracy and score. Try changing the batch_size, the number of epochs or even adding/removing layers in the CNN model, and have fun!

The media shown in this article are not owned by Analytics Vidhya and is used at the Author’s discretion.