Computer Generated Algorithm of Sand Painting Based on Convolutional Neural Network

The neural style transfer of sand painting is to transform the style with the artistic image of sand painting into another image, and to create a new art image with appreciation while maintaining the semantic content of the image. This paper introduces a sand painting image simulation algorithm based on convolutional neural network. The algorithm utilizes the output of the VGG16 network as an advanced feature of the image to achieve the content of the target image by separating and recombining the content features and style features of the input image. The fusion of sand painting form and high quality computer art works proves the correctness and effectiveness of the algorithm.