Nearest neighbor interpolation

from __future__ import division
import cv2
import numpy as np
 
# Nearest neighbor interpolation 
def nn_interpolate(image, scale_factor):
 
	# Extract size
	(rows, cols, channels) = image.shape
	scaled_height = rows * scale_factor
	scaled_weight = cols * scale_factor
 
	# Compute ratio
	row_ratio = rows / scaled_height
	col_ratio = cols / scaled_weight
 
	row_position = np.floor(np.arange(scaled_height) * row_ratio).astype(int)
	column_position = np.floor(np.arange(scaled_weight) * col_ratio).astype(int)
 
	# Initialize scaled image
	scaled_image = np.zeros((scaled_height, scaled_weight, 3), np.uint8)
 
	for i in range(scaled_height):
		for j in range(scaled_weight):
			scaled_image[i, j] = image[row_position[i], column_position[j]]
 
	return scaled_image
 
# Read image
image = cv2.imread('fruits.jpg')
nn = nn_interpolate(image, 2)
 
 
cv2.imwrite('original.jpg', image)
cv2.imwrite('Nearest-neighbor-interpolation.jpg', nn)