Fractal Art Graphic Generation Based on Deep Learning Driven Intelligence

Abstract

Science and technology have greatly promoted social progress and development, and their impact in the field of art and design is also enormous. Driven by AI (Artificial intelligence), this article proposes a fractal art graphic generation model based on DL (Deep learning). By minimizing the total loss function of content information and style information, the model iteratively optimizes a random noise image, so that the random image finally retains both the content information of the content map and the texture information of the style map. Moreover, by designing loss functions with different fusion degrees to meet the requirements of different feature extraction, the random gradient descent method is used to iteratively update the pattern generation effect to realize the fusion generation of fractal art graphics. The results show that the F1 value of the model can reach 96.21%, which is about 12% better than that of the AlexNet model. It is about 7% better than ResNet model. Compared with other classic DL models, the proposed model has better performance, certain reliability and practicability. It provides more basis for CAD fractal art graphic generation based on DL.