An intelligent design method for color-coated kirigami based on deep learning

Abstract

The invention discloses an intelligent design method for color jacket paper-cutting based on deep learning, comprising the following steps: A: inputting a pair of high-definition natural pictures; B: performing face recognition on the high-definition natural pictures; C: identifying Face contour segmentation processing is performed on the image part of the human face, and landscape contour separation processing is performed on the image part that does not recognize the human face; D: select a style model for the high-definition natural picture in step A, and then use the selected style model The image is used as a style image, and the part of the image that has been processed by the landscape contour is used as the content image for neural style transfer, and the generated image of the same layer is corrected by a norm function to obtain the final image of neural style transfer. The invention adopts the intelligent technology to replace the traditional paper-cut technology, thereby reducing the design threshold and design cost of paper-cut, and greatly shortening the design and production cycle of paper-cut works.

Description

technical field [0001] The invention relates to the field of image intelligent recognition, in particular to an intelligent design method for color-coated paper-cuts based on deep learning. Background technique [0002] Paper-cutting is a cultural art with a long history in our country. At present, the traditional paper-cutting method in our country is to design, sample, and open the plate manually, which greatly improves the design time of design drawings, layer stripping, and sample casting. In two ways: [0003] 1. Due to different kirigami styles, traditional kirigami methods require manual modeling of kirigami styles, which requires a large number of kirigami samples for training and learning, prolonging the processing time of traditional kirigami techniques; [0004] 2. The main body of the paper-cut is the human face, and everything is designed around the lines of the human face. Therefore, it is necessary to remove the background of the image reasonably, while the ...

AI Technical Summary

Problems solved by technology

Therefore, it is necessary to remove the background of the image reasonably, while the traditional paper-cut method needs to be removed manually, and the contours, lines, and textures of the background are naturally different from those of the human face, which increases the difficulty in processing

[0005] However, with the development of computer image processing technology today, there is still no system that can specifically solve the problem of image layering in the color layering process of Chinese paper-cutting. Therefore, the traditional paper-cutting method has become a limitation in my country for hundreds of years. The bottleneck of paper-cut development

Images