AI_Niewiarowska_lab8_zad1

from keras.datasets import mnist

import matplotlib.pyplot as plt
import numpy as np


def convolution2d(image, kernel):
    m, n = kernel.shape
    if (m == n):
        y, x = image.shape
        y = y - m + 1
        x = x - m + 1
        new_image = np.zeros((y, x))
        for i in range(y):
            for j in range(x):
                new_image[i][j] = np.sum(image[i:i+m, j:j+m]*kernel)
    return new_image


def maxpooling(image):
    m, n = kernel.shape
    if (m == n):
        y, x = image.shape
        x //= 2
        y //= 2
        new_image = np.zeros((y, x))
        for i in range(y):
            for j in range(x):
                ox = j * 2
                oy = i * 2
                new_image[i][j] = np.max(image[oy:oy+2, ox:ox+2])
    return new_image


(X_train, y_train), (X_test, y_test) = mnist.load_data()


kernel = np.array([[1, 0, -1], [1, 0, -1], [1, 0, -1]])
print(kernel)

image1 = X_train[0]
image2 = convolution2d(image1, kernel)
image3 = maxpooling(image2)

plt.subplot(1, 3, 1)
plt.imshow(image1, cmap=plt.get_cmap('gray'))
plt.subplot(1, 3, 2)
plt.imshow(image2, cmap=plt.get_cmap('gray'))
plt.subplot(1, 3, 3)
plt.imshow(image3, cmap=plt.get_cmap('gray'))
plt.show()