Face detection refers to the process of automatically detecting, identifying and locating faces in images or videos by using computer vision technology. The purpose of this technology is to distinguish the face in an image or video from other objects and backgrounds, so as to accurately locate and analyze the face features. Face detection can be achieved in many ways, including feature-based method, image-based method and deep learning-based method. In this paper,MTCNN convolutional neural network method based on deep learning is used, and three cascaded neural networks are constructed by using convolutional neural networks of various scales and small networks with different aspect ratios, and the information such as the position, size and direction of one or more faces is successfully identified and located in the image. At present, MTCNN convolutional neural network method has become one of the most popular technologies because of its high efficiency, accuracy and strong robustness.