Due to the importance of license plate recognition in many parts in human society, several methods are concluded and proposed in this article. In this work, artificial intelligence (AI) technology, deep learning, and neural networks are put into use in different aspects or modules of license plate recognition methods. Current methods can be divided into two categories, one is the traditional methods and the other is method using AI technology. Two-stage method is a typical example for traditional methods. In this work, other methods such as RT-LPDRnet method, which takes YOLO v5 into use, is an example for AI technology in license plate recognition. Now commercial license plate recognition generally adopts the method of end-to-end deep learning, while some methods of character segmentation first and then recognition are still worthy of study. In this work, three evaluation metrics are concluded to help evaluate the performance of previous methods. In addition, application of license plate recognition is proposed such as recognition in the parking lots, entrances and exits of expressway and electronic police system.