Simulation Of Gourd Pyrography Style Based On Convolutional Neural Network

In recent years,deep learning technology has attracted much attention,especially in the field of computer vision multimedia,and its research results are even more prominent.Among them,compared with traditional mathematical modeling methods,the effects of simulation and simulation of sketch painting,landscape painting,oil painting,pencil painting,chalk painting,pyrography,etc.can be further achieved by using deep learning technology with outstanding simulation effects.Using traditional mathematical modeling methods to simulate the art works with a certain style,the pertinence is too strong to generalize the large-scale natural images.However,the use of deep learning technology to achieve non-photorealistic rendering,you can achieve arbitrary stylization of any natural image,so as to achieve the simulation of any picture art style.In this paper,using the deep learning method and the deep convolutional neural network model,this paper proposes a simulation method based on deep neural network for the artistic style migration based on the generalization of the current hoist pyrography simulation method.The pyrotechnic style simulation method proposed in this paper is mainly divided into three different stages.Each stage uses different data sets to train different neural network models,and uses the models that have been trained in each stage to process the target images of each stage.The specific simulation process is to separately perform image semantic segmentation,artistic style migration and image deformation fusion on the target image,and generate ordinary natural photos taken by shooting equipment such as camera into a gourd pyrography with a pyrography style.Through the comparative analysis of the experimental results,it can be found that the hue and texture information of the generated gourd pyrography is very similar to the real gourd pyrography,and the actual effect is also very realistic,with high artistic enjoyability.Therefore,the simulation method proposed in this paper can convert ordinary natural photos into corresponding hoist pyrography with good artistic effects,and finally complete the artistic style migration and simulation task of hoist pyrography.