GAN-based automatic design image generation method

Abstract

The invention discloses an automatic design image generation method based on a GAN. The method comprises the following steps: 1, carrying out the blank elimination and size normalization of an image,and converting all images into a pixel size of a preset pixel * a preset pixel; 2, building and operating a GAN deep learning model; 3, after a plurality of rounds of training iteration, assigning thehyper-parameters to a generator network to generate a new design image; 4, performing corrosion and expansion mathematical calculation on the generated design image to reduce noise points of the design image; and step 5, smoothing the color edge in the image by using median filtering to reduce edge abrupt change. Compared with a traditional GAN method, the method has the advantages that on one hand, the edge of the generated design image is smoother, the color is more attractive, on the other hand, the distribution function of the generated image can be changed according to hyper-parameter adjustment in the network model, and the diversity of the design image is improved.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to View More

AI Technical Summary

Benefits of technology

The technical effect of this new method described in this patented technology allows for better edge-quality images that have fewer streaks or shimmering on top while maintains their original color balance. This means they are easier to see when displayed properly compared against older methods such as GAN.

Problems solved by technology

This patented describes how Artificial Intelligence Algorithms help create unique designs by generating high quality photos without any imperfections like cross-color matchings, detaillessness, etc., while also enhancing their ability to produce different types of shapes.

Images