After analyzing every component of a CNN in detail, it is time to see the general structure of a CNN as a whole. For example, starting from the images as input layers, there will be a certain series of convolutional layers interspersed with a ReLU layer and, when necessary, the standardization and pooling layers. Finally, there will be a series of FC layers before the output layer. Here is an example of a CNN architecture:
The basic idea is to start with a large image and continuously reduce the data step by step until you get a single result. The more the convolution passages you have, the more the neural network will be able to understand and process complex functions.