Image Classification Using CNN

The convolutional neural network is one of the most popular deep learning model for image classification. In this Blog, we are designing a CNN model to classify the Cat Vs Dog (weather image is lable is cat or dog)  using Tensorflow.  

This is a simple CNN Network

# Importing the libraries
import tensorflow as tf
from tensorflow.keras.preprocessing.image import ImageDataGenerator

#data sugmentation
# Preprocessing the Training set
train_datagen = ImageDataGenerator(rescale=1./255,
                                rotation_range=40,
                                width_shift_range=0.2,
                                height_shift_range=0.2,
                                shear_range=0.2,
                                zoom_range=0.2,
                                horizontal_flip=True,
                                fill_mode='nearest')
training_set = train_datagen.flow_from_directory('image_data/training',
                                                 target_size = (64, 64),
                                                 batch_size = 32,
                                                 class_mode = 'binary')

Found 198 images belonging to 2 classes.

# Preprocessing the Test set
test_datagen = ImageDataGenerator(rescale = 1./255)
test_set = test_datagen.flow_from_directory('image_data/validation',
                                            target_size = (64, 64),
                                              batch_size = 32,
                                              class_mode = 'binary')

Found 100 images belonging to 2 classes.

## showing some image from training
import matplotlib.pyplot as plt
def plotImages(images_arr):
    fig, axes = plt.subplots(1, 5, figsize=(20, 20))
    axes = axes.flatten()
    for img, ax in zip(images_arr, axes):
        ax.imshow(img)
    plt.tight_layout()
    plt.show()

images = [training_set[0][0][0] for i in range(5)]
plotImages(images)

Model Build Use Only CNN

from tensorflow.keras.layers import Conv2D

# Part 2 - Building the CNN
# Initialising the CNN
cnn = tf.keras.models.Sequential()
# Step 1 - # Adding a first convolutional layer
cnn.add(tf.keras.layers.Conv2D(filters=32,padding="same",kernel_size=3, activation='relu',
## step 2 - #apply maxpool
cnn.add(tf.keras.layers.MaxPool2D(pool_size=2, strides=2)) ## Apply pooing stride
# Adding a second convolutional layer
cnn.add(tf.keras.layers.Conv2D(filters=32,padding='same',kernel_size=3, activation='relu'))
cnn.add(tf.keras.layers.MaxPool2D(pool_size=2, strides=2))
# Step 3 - Flattening
cnn.add(tf.keras.layers.Flatten())
# Step 4 - Full Connection
cnn.add(tf.keras.layers.Dense(units=128, activation='relu'))
tf.keras.layers.Dropout(0.5)
# Step 5 - Output Layer
cnn.add(tf.keras.layers.Dense(units=1, activation='sigmoid'))

# Part 3 - Training the CNN
# Compiling the CNN
cnn.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])

# Training the CNN on the Training set and evaluating it on the Test set
history = cnn.fit(x = training_set, validation_data = test_set, epochs = 2)

Save And Load Model

#save model
from tensorflow.keras.models import load_model
cnn.save('model.h5')

from tensorflow.keras.models import load_model
# load model
model = load_model('model.h5')

# Part 4 - Making a single prediction
import numpy as np
from tensorflow.keras.preprocessing import image
test_image = image.load_img('image_data/test/3285.jpg', target_size = (64,64))
test_image = image.img_to_array(test_image)
test_image=test_image/255
test_image = np.expand_dims(test_image, axis = 0)
result = cnn.predict(test_image)
result

array([[0.5059088]], dtype=float32)

if result[0]<=0.5:
    print("The image classified is cat")
else:
    print("The image classified is dog")
from IPython.display import Image
Image(filename='image_data/test/3285.jpg',height='200',width='200')

The image classified is dog