What are the most important data augmentation methods for classification?

I have a classification problem of 7 classes with around 10k images. i need a list of common data augmentation methods with their details.

any survey on data augmentation methods or publications will also be helpful.


In my experiences, if you trained a lot of the same datasets the convergences will be slowed to handle and sometime will be deaded with overfitting problems, and it hard for you to manipulate.