#0022

Resize Image With Nearest Neighbor

MediumML22 min15 XP

Problem

Given a 2D NumPy array image and new height/width, return a resized image using nearest-neighbor sampling.

Why This Matters

Image resizing looks magical in libraries until you implement the coordinate mapping yourself.

Function Signature

def resize_nearest(image, new_h, new_w):

Examples

Example 1

Input[[1, 2], [3, 4]] resized to 4x4

Outputeach original pixel becomes a 2x2 block

Nearest neighbor copies the closest source pixel without interpolation.

Constraints

Input is a 2D grayscale image.
Use nearest-neighbor sampling, not averaging.
Return shape must be (new_h, new_w).

CodePython

import numpy as np

def resize_nearest(image, new_h, new_w):
    """
    Input example: [[1, 2], [3, 4]] resized to 4x4
    Return: each original pixel becomes a 2x2 block
    Return the value from the function so the tests can compare it. Do not print unless the prompt asks.
    """
    # write your code here
    return image

import numpy as np

def resize_nearest(image, new_h, new_w):
    """
    Input example: [[1, 2], [3, 4]] resized to 4x4
    Return: each original pixel becomes a 2x2 block
    Return the value from the function so the tests can compare it. Do not print unless the prompt asks.
    """
    # write your code here
    return image

Visible browser tests run here when available.

Testcases1 visible / 4 hidden categories

Upscale 2x2 to 4x4

Input[[1,2],[3,4]] -> 4x4

Expected2x2 blocks of each value

Each source pixel covers a larger region in the output.

Hidden Test Categories

Downscale imageResize to same shapeNon-square target shapeSingle-row image