Let's now talk about applications in computer vision for image detection. Note the following photo of a pavement. It is cracked and damaged:

Image via pxhere.com, CC0
Here is another one, it is cracked but in a different manner.

Image via pxhere.com, CC0
There can be many more such images collectively referred to as datasets. You can find more here: pxhere.com/en/photo/690701.
The previous pavement photos were captured with a Sony Cybershot DSC-RX100M4 20.1 Megapixel digital camera on 03/09/2017. Python can make use of GPUs to accelerate deep learning and classify/identify such images. It is one of the many applications in computer vision.
The following paper is a review on pavement distress detection. It talks about computer vision-based automated pavement distress detection...