Research On Virtual Restoration And Application Of Damaged Pattern On Silk Clothing Relics

Silk costume relics hold significant cultural and historical memories,and their intricate patterns represent the ingenuity and creativity of ancient people,carrying immense national connotation and research value.However,due to prolonged exposure to physical and chemical factors,silk costume relics have faced discoloration,aging,and damage,which have diminished their visual apparel and integrity,requiring urgent conservation and restoration.Compared to manual ontological restoration,digital restoration of cultural relics has the advantages of non-contact,high accuracy,speed,low cost and multi-program generation,which is an effective aid in restoring and regenerating cultural relics.This paper explores the use of digital technology tools to restore silk costume heritage patterns..The research focuses on three aspects of silk costume relics:image clarification,defective patterns restoration,and pattern extraction and vectorization methods.The paper presents the following research contents and conclusions:(1)Image clarification of silk costume artifactsThe image local degradation model is established for the stain problem on the surface of silk costume relics,and a rough stain mask is obtained.And the image is divided into the sample area and the area with stain,and the pixels in different areas are color clustered to roughly estimate the color line offset in the stain area,and the stain is removed from the image by offset correction.After objective evaluation,the average gradient mean value obtained by the method in this paper is 0.0795,which retains more image details.(2)Restoration of defective patterns on silk relicsIn order to solve the problems of structural lines disconnection and filling erroes after image restoration,a restoration method of reconstructing structural lines to guide image completion is proposed.The structural lines in the defective region are fitted by using the curve fitting,and the boundary function is introduced to improve the priority,The sample block size is adaptively adjusted according to the complexity of the image structure to solve the problems of block effect.After experimental validation,the obtained SSIM average value reaches 97.7% and the PSNR average value is 39.16.The results show that the proposed method satisfies the structural continuity better and reproduces the original patterns on the silk relics.(3)Pattern extraction and vectorizationFor the problem of broken points and burrs on the edges of the pattern extracted by the conventional edge detection operator,we combine median filtering to preprocessing the image,adaptive double thresholding to extract strong and weak edges and remove false edges,then use Laplace algorithm to strengthen the edges,and then use morphological processing to solve the problem of broken edges.Finally,the extracted patterns are vectorized to obtain smooth,closed and editable line drawings for color matching design applications in modern apparel and home textile products.The innovations of this research lies in: the establishment of an image degradation model based on the surface contamination of silk relics,estimation of color line offset by clustering the color lines;Additionally,the proposed image restoration method uses reconstructed structure lines to guide image inpainting and reconstructs missing structure lines using curve fitting.Lastly,the extracted patterns are vectorized to obtain smooth,closed,and editable line drawings.This research provides a valuable contribution to the conservation and restoration of cultural heritage relics.