Write efficient Python code, minimizing loops with vectorization where possible.the input image is $512$x$512$ and the tiles are $32$x$32.$ Both are RGB images, so the input image will have the shape $(512,512,3)$ and each tile will have the shape $(32,32,3).$do not normalize the tiles by dividing by $255.$ Keep the values as integers between $0$ and $255.$You can save the resulting nparray as an $\text{'}.$png$\text{'}$ or $\text{'}.$jpg$\text{'}$ image like this: Image.fromarray$($result$).$save$(\text{'}$out$.$png$\text{'}).$ import numpy as np

import matplotlib.pyplot as plt

from sklearn.neighbors import KDTree

from PIL import Image

import urllib.request

from keras.datasets import cifar$100$ as tiles$\_$dataset. tiles $=$ tiles$\_$dataset.load$\_$data$()[0][0].$ First, compute the mean colors of all the tiles and construct a kd$-$tree.

Now generate $10$ random colors and query the kd$-$tree for the nearest neighbours. For each of the $10$ colors, imshow $($in a horizontal row$)$: the original color, the nearest mean color and the tile corresponding to the mean color. $2.$ Make a mosaic of the image based on that. $3.$ imshow the portion of your mosaic that corresponds to the top$-$left $8$x$8$ tiles.