Landscape Art Image Style Reconstruction Algorithm Based on Machine Learning

Images are visual information intuitively felt by human beings, and painting is one of the windows for people's emotional communication. The painting style expresses the creative background, artistic characteristics and rich emotions of the creator. For a long time, the creation process of artistic paintings requires professional skills and a lot of energy and time. According to the characteristics and styles of the landscape, on the basis of the cycle-consistent generative adversarial network (GAN), an attention mechanism is introduced into the network, and a better image style transfer effect is obtained by assigning higher learning weights to the main features of the image., to enhance some details of the stylized image. This paper adopts the method of machine learning. Due to the use of the Internet, it has become a reality to obtain information regardless of time and region. However, how to effectively use this information and use this information to improve productivity has become an urgent problem to be solved. Machine learning is a One of the effective ways to solve this kind of problem.