Research On The Pattern Extraction Of Ceramic Art And Construction Of Image Reproduction System

With the development of high technology,the digitalization of works of art and cultural relics is no longer a problem.Through continuous improvement of the quality of digital printing technology and ink color,there are many digital products on the market,such as digital painting,digital calligraphy,digital paper cutting,digital embroidery and other digital art reproductions.With the rapid development of 3D technology,its application is becoming more and more popular,such as the restoration of cultural relics,the manufacture of simulation,the creation of digital sculpture and the production of ceramic art.However,in the process of copying these three-dimensional artifacts or art,3D printing technology is still more concerned about the shape of the object,the color and the pattern is still drawn by hand.However,in addition to modeling,the pattern is also an important part of art.Chinese traditional patterns have a long history and brilliant achievements.Promote the digital protection of traditional pattern,so that the traditional Chinese culture can be inherited.Based on this background,this thesis focuses on three-dimensional pattern on ceramic art,positions in high quality reproduction and creation,a comprehensive study of the pattern extraction of ceramic art and image stitching technology is conducted in this thesis.The thesis constructs the image reproduction system and realizes the complete digital process from the three-dimensional pattern of shooting,preprocessing,image correction,image transformation from three-dimensional to two-dimensional,image registration,image fusion to generate a complete two-dimensional pattern.Firstly,this thesis reviews the domestic and foreign research status and application field about art reproduction and image stitching technology,introduces the basic flow and the key technology of image mosaic,analyzed and summarized the main problems of the main methods of image registration and image fusion performance.This thesis studies the possible methods of image acquisition in the process of panorama generation,and analyzes the problems existing in the process of image acquisition.The original image acquisition through hardware is not up to the expected quality,It is necessary to carry out effective and accurate image denoising,image correction and so on which is accurate to the calibration accuracy of pixel level image.At the same time,the three models of panorama projection are analyzed.In the process of image acquisition and display,geometric distortion,gray distortion and color distortion are often produced.In this thesis,the distortion of digital image is analyzed,and the corresponding correction method is introduced.According to the method of image acquisition,the image is only a small color difference,so this thesis adopts a simple histogrammatching to adjust the color of the image.Image mosaic technology is to connect the overlap of a group of images captured from the same sensor at different time or different sensors at the same time into a high clear,wide viewing angle of the large image.Because the object of this thesis is the shape of the curved surface,the image of the object is nonlinear with respect to the movement of the camera.The commonly used splicing methods can not be directly applied to the research.In this thesis,a method of calculating homography matrix is proposed to solve the problem.On the basis of previous studies,this thesis analyzes the principle of the main image stitching methods and their advantages and disadvantages.Aiming at the problem that the feature points are few,the thesis proposes a combination of SIFT and MSER algorithm for image stitching algorithm,which can extract more image features for later image registration.In the aspect of feature matching,based on K-D tree search algorithm,the improved BBF search algorithm can improve the feature points searching efficiency;In the aspect of feature purification,the RANSAC algorithm is used to filter out the points,which can eliminate the redundancy and mismatching.In order to remove the possible gap which may occur in the stitching process,an improved multi band fusion algorithm is proposed to solve the problem of image blurring.Experimental results show that the proposed algorithm can be applied to the actual case,and finally get a complete and clear two-dimensional plane pattern.