I have seen ways in which we can "prep" images before using them for detection tasks, but I\'m failing to understand the reason behind centralizing, standardizing,
