I would like to use the CNN pre-trained model in feature extraction but I don't know what features are extracted from that. Please let me know!
Here is a good post about how features are extracted using CNN: https://towardsdatascience.com/cnn-application-on-structured-data-automated-feature-extraction-8f2cd28d9a7e
"I don't know what features are extracted from that."
It depends on what dataset was that CNN pre-trained on. E.g. if the model was pre-trained on a dataset consisting of dogs and cats, the features should contain useful information about dogs and cats.
In practice it turned out that models pre-trained on ImageNet dataset extract features which are a good starting point for training on other custom datasets.