Font Size: a A A

Computer Simulation Algorithm Design And Realistic Sense Evaluation Of Colored Spun Knitted Fabrics

Posted on:2022-12-17Degree:MasterType:Thesis
Country:ChinaCandidate:D HuoFull Text:PDF
GTID:2481306779487134Subject:Computer Software and Application of Computer
Abstract/Summary:PDF Full Text Request
Textile industry is not only a traditional pillar industry of China's national economy and an important livelihood industry,but also an industry with obvious international competitive advantages.In recent years,textile fabrics made of two or more different dyed fibers after mixing and spinning have been deeply sought after by consumers because of their delicate and soft appearance and unique hazy three-dimensional effect and texture,and formed a unique colored spun industry.However,in the colored spun process,because the color presentation principle of colored spun products does not fully comply with the "color addition characteristics" or "color subtraction characteristics",it is necessary to determine the product effect through repeated trial spinning and visual inspection.With the development of digital and information technology,using computer simulation technology to simulate and visualize the appearance of fabric has become the focus of current research.Due to the particularity and complexity of the colored spun products,this paper focuses on the following contents:(1)Aiming at the problem that it is difficult to effectively distinguish the appearance similarity of colored spun fabrics with a single feature measure,an image similarity measure model combining local and global features is proposed.The model extracts the local and global feature descriptors of the image in RGB and HSV color space respectively,and uses the local and global feature distance to measure the similarity of fabric appearance.The experimental results show that the constructed measurement model can distinguish the similarity of 9 kinds of colored spun fabrics with different color matching schemes and organizational structure,and its category separability index reaches 30.02,which has an ideal discrimination effect.(2)Aiming at the problem that it is difficult to effectively separate the fabric color information with local non-stationary characteristics by global frequency domain filtering in the simulation process,a computer simulation method of color spun knitted fabrics based on dual tree complex wavelet is proposed.This method uses dual tree complex wavelet to extract the low-frequency component of fabric image and characterize its color information.On this basis,the image coloring algorithm is designed and the computer simulation of color spun knitted fabric is realized.The experimental results show that compared with the frequency domain filtering method and wavelet transform method,the simulation effect of this method is improved by 7.37% and 3.47% respectively in the similarity measurement index,7.61 db and 13.63 db respectively in the peak signal-to-noise ratio,0.0115 and 0.0081 respectively in the structural similarity,and it has ideal effectiveness.(3)Aiming at the problem that the color information and texture background information of colored spun knitted fabric are overlapped in the spectrum and it is difficult to separate them effectively in the simulation process,a computer simulation method of color spun knitted fabric based on image translation model is proposed.In this method,pix2 pixhd model is used to train and learn the image conversion process of "fabric color component",so as to construct the color component extraction model of colored spun knitted fabric.On this basis,the image coloring and fusion algorithm is designed,and the computer simulation of color spun knitted fabric is realized.The experimental results show that compared with the frequency domain filtering method,the simulation effect of this method is improved by 11.94%in the constructed similarity measurement index,15.72 db and 0.0168 in the peak signal-to-noise ratio and structural similarity index respectively,and its realism is significantly improved.The research of this paper can promote the digital transformation and upgrading of colored spun industry and provide effective guidance for the design and development process of colored spun products.By significantly reducing the number of physical proofing,the development cycle and production cost of colored spun products can be reduced,so as to meet the market demand of small batch,multiple varieties and fast delivery date.
Keywords/Search Tags:computer simulation, colored spun knitted fabric, realistic evaluation, dual tree complex wavelet, image translation model, local features, global features
PDF Full Text Request
Related items