Center Crop a Grayscale Image

Given a rectangular grayscale image image, return its centered cropHeight × cropWidth crop.
The parity guarantees make the unused rows and columns split evenly across opposite sides.

Function
centerCrop(image: int[][], cropHeight: int, cropWidth: int) → int[][]

Examples
Example 1
image = [[1,2,3,4],[5,6,7,8],[9,10,11,12]]
cropHeight = 1
cropWidth = 2
return = [[6,7]]
One row and one column are removed evenly from opposite sides.

Constraints
1 <= image.length, image[i].length <= 200.
1 <= cropHeight <= image.length and 1 <= cropWidth <= image[i].length.
(image.length - cropHeight) and (image[i].length - cropWidth) are even.

The trick is two offsets. Top offset is (rows - cropHeight) / 2. Left offset is (cols - cropWidth) / 2. Because the constraints guarantee both differences are even, integer division is exact, so you never round. Then loop from top to top + cropHeight and slice each row from left to left + cropWidth. In Python that's image[top:top+h] with a row slice inside a comprehension. The common pitfall is using the full dimensions as the end index, or computing the offset from the crop size instead of the leftover. Another slip is swapping height and width, or reading column count from the wrong row. Check it against the example: 3 rows, crop 1, top is 1. 4 columns, crop 2, left is 1. That gives [[6,7]]. If you blank during the live OA, StealthCoder can supply the slicing logic while you verify the offsets yourself.

Memorize the pattern. If you can't, run StealthCoder. The proctor sees the IDE. They don't see what's behind it.