Attribute prediction of Pokémon images based on convolutional neural network

Image recognition has long been a fundamental research problem in computer vision, where the goal is to predict the class of a given image. Existing researches have focused on the category level while pay no attention on predicting object fine attribute information of object. To this end, in this paper, in order to investigate the classification effect of convolution neural network on Pokémon image data set, we use convolution neural network to classify the attributes of Pokémon, which can judge the attributes of a certain elf by its image. We used the pictures of the Pokémon in a scorching game "Pokémon " on the market as our data set. Specifically, we first a structure of the convolutional neural network and apply it to the Pokémon image data set, which can realize automatic recognition and attribute classification of elf images. Then the data enhancement method is used to preprocess the image data, which is combined with a convolutional neural network to improve the effect of attribute recognition and classification. Results show the effectiveness of our proposed method.