In order to solve the low recognition rate of template matching method in license plate character recognition, especially the problem that the similar characters cannot be identified accurately, this paper proposes a method of license plate character recognition based on the fusion of local HOG and layered LBP feature. Firstly, we use the template matching method for preliminary identification of all the characters of license plate. Then, a small edge feature of the biggest difference in the similar characters of the license plate and the template is extracted by using HOG operator, and then the layered texture feature of the same area block of HOG in the similar characters of the original license plate and the template is extracted by using LBP operator. Next, serial feature vectors are constructed with serial fusion of the edge feature and the layered texture feature. Finally, according to the Chi square distance between the feature vectors, we measure the similarity of the license plate characters and the template characters, and then complete the second recognition. The recognition performances of the 11 algorithms are compared through experiments. The results show that this method is very effective to solve the problem of false recognition of similar characters and the recognition rate is improved significantly at the same time, which is as high as 99.52%.