Fresco Sketch Generation Based On Deep Learning

The protection of the endangered ancient fresco is urgent.Sketch drawing of fresco is an important task of protecting fresco.Traditionally rely on manual sketch drawing method complex process,time-consuming,inefficient,very detrimental to the protection of fresco.Therefore,the introduction of computer technology,especially image processing technology is particularly important.Based on the classic VGG16 convolutional neural network,this paper designs a convolutional neural network that can extract the image sketch,and realizes the imageto-image sketch-drawing extraction with the help of the idea of image edge detection and deep learning.In this paper,the open data set is used to train the convolution neural network,and then the original fresco data is input into the neural network to obtain the sketch drawing of the fresco.The convolution neural network,which is designed for line drawing,has achieved excellent results in the test of the public data set,and has obtained good results in the drawing of the mural line diagram,meanwhile,it provides a new way for the work of fresco protection.