Research On The Method Of Artificial Intelligence Layout Design Based On Convolutional Neural Network

With the development of artificial intelligence(AI)technology,more and more scholars are exploring the combination of AI and design.The design field will face new opportunities and challenges.The simple and repetitive design operation will be replaced by artificial intelligence,which greatly improves the design efficiency,so the intelligent design is gradually becoming a development direction.At present,the field of visual communication design is particularly affected by artificial intelligence.In addition,automatic processing applications for pictures and color matching have emerged,and many design models for intelligent layout have been proposed.Most of these models take magazine media as the carrier,and realize the automatic layout of pictures and texts through statistical learning framework or a priori design framework.However,statistical learning framework model will face the limitations of black box and information cocoon room,and the design process of prior design framework model still needs human participation,which belongs to semi intelligent application.This research will explore the combination of artificial intelligence and layout design from a new perspective,and propose a new matching model based on posters to solve the above problems.The core of the research is to solve two problems,one is how to make the computer understand the elements of the layout,the other is how to arrange these design elements by computer,that is,to design the layout through computational logic.To solve the first problem,this paper classifies and labels the elements of poster layout,and uses the faster R-CNN algorithm model based on convolutional neural networks to realize the recognition and location of elements.Aiming at the second problem,this paper constructs a matching model based on poster composition,which transforms the design problem into the target matching problem,so as to realize the intelligent layout design.The matching model based on poster composition is composed of layout module and element design module.The layout module is used to define the spatial layout of text and image,and produce the composition template.The element design module is used to plan the generation path of visual elements,complete the design element processing based on template constraints,and output the final scheme.Finally,the result of simulation experiment shows that not all the new layouts produced by layout module are available.The reason may be that there is no evaluation and screening of the layout schemes.In addition,the call of the preset decoration element parameters in the element design module reduces the quality of the layout design scheme,which may be affected by its single form of expression.