Image augmentation applies simple transforms to images to create more training examples and help computer vision models generalize better.

A common augmentation pipeline might include horizontal flips, brightness scaling, and random cropping, each applied in a controlled way.

Requirements

Implement the function

Rules:

• Apply transforms strictly in the order: flip → brightness → crop.
• Horizontal flip means reversing the columns: [:, ::-1].
• Brightness scaling uses the formula $p' = p \cdot b$ for each pixel value p and factor b.
• Cropping uses crop = [top, left, height, width] and returns the sub-image.
• Use only NumPy.

Example

Output: