The full name of CNN is "Convolutional Neural Network". Neural network is a mathematical model or computational model that imitates the structure and function of biological neural network (the central nervous system of animals, especially the brain). Neural network is composed of a large number of artificial neurons, and different networks are constructed according to different connection methods.

CNN is one of them, along with GAN (Generative Countermeasure Network) and Recurrent Neural Network. Neural network can have simple decision-making ability and simple judgment ability like people, and can give better results in image and speech recognition.

CNN is an artificial neural network, and its structure can be divided into three layers:

Convolution layer? -The main function is to extract features.

Maximum pool layer? -The main function is to down-sample without destroying the recognition result.

Full connection layer? -The main function is classification.

Applying it to an image can be understood as putting a filter on the image to find out some features in the image. We need to find many features to distinguish an object, so we will have many filters. Through the combination of these filters, we can get many features.

First of all, the format of a picture saved in a computer is pixels. For example, the length of a picture is 1080 and the width is 1024. The total * * * includes pixels of 1080 * 1024. If it is an RGB picture, it consists of three colors and contains three channels.