Design of Graphic Design Assistant System Based on Artificial Intelligence

Creation Of A Graphic Design Assistant System Based On Artificial Intelligence

Graphic design works should have concise language and intense artistry, which can effectively reflect the author’s artistic style. Although multimedia is developing rapidly and three-dimensional (3D) technology is widely used today, graphic design images also have high application value (Tan & Yang, 2021). Artificial intelligence (AI) computer-aided design has been gradually applied to graphic design-aided systems, which is closely related to graphic design modernization. With the development of AI technology, machine learning based on design image data makes AI-aided design possible. In this regard, deep learning (DL) has shown great application potential in design image analysis and generation (Beik, 2020). The mainstream graphic design methods in the market can no longer meet graphic designers’ growing design inspiration. More and more, graphic designers put forward new requirements for design forms. With the development of technology, graphic design tools are becoming more and more perfect, which allows graphic designers to realize their dream designs, achieve more special effects, and expand their conceptual choices.

Design is an innovative, open and repetitive work to solve problems. In the design process, designers will constantly and better study issues and evaluate better schemes, and improve design (Dong et al., 2021), so iteration is constructed as the most effective innovation pice. It provides a mechanism to support design innovation (Zheng et al., 2019). Bhatti et al. (2020) applied region segmentation to artistic image style, used a down-sampling method to express it step by step, then used a canny operator to refine the edges, and finally filled the corresponding colors and textures according to the blocks. Kim et al. (2022) proposed a mosaic method of image blocks, which first matched the original texture with the existing texture, found the texture mapping between the source image and the target image, and then cut off the redundant boundaries. Long and Han (2020) proposed a new component-based model and a room classification method using data obtained from visual sensors. Zhang et al. (2018) used convolutional neural network (CNN) to identify the items placed in the room in the image, in order to determine the room type which was used for the service robot working in the home environment, so that it could identify the room it visited. Zhang et al. proposed a method of clustering architectural forms. The evaluation indexes of this clustering algorithm can be defined artificially, such as lighting, cooling, and heating load. With the deepening of image graphics based on DL, AI has been applied to practice, and the development of computer vision has become more and more rapid. However, the problem is feature extraction of 3D models (Zhang, Cai et al., 2018).

The development of human science and technology is a process of constantly improving its needs. The fundamental purpose of developing and applying science and technology is to continuously meet people’s living and spiritual needs, among which spiritual needs are critical. For graphic designers, the satisfaction of spiritual needs is the realization of design ideas. The existing graphical design system can provide a graphical platform for users, but it cannot offer more graphical geometric basis for users in the graphical process. Therefore, in this paper, the author proposes a new AI-based graphic design assistant system, which can complete the aesthetic analysis of geometric element proportion in graphic design, thus helping users to perceive the feeling of visual aesthetic and providing users with design support related to geometric aesthetic principles.