Research On The Algorithm Of Evaluation For Brush Calligraphy Copy

Chinese calligraphy is not only a visual art with a long history,but also the soul of Chinese art.Copying calligraphy,as a key to calligraphy,is the only way to study calligraphy.And form likeness is the most basic requirement of coping calligraphy.Evaluating imitating calligraphy by means of computer is a very valuable topic,which can make calligraphy learners understand their study effect more timely,so that Chinese calligraphy can be better inherited and developed.Therefore,the main content of this thesis is to judge whether the calligraphy character copied by beginners are similar to that in the copybook of calligraphy,so as to give an evaluation of the copied calligraphy.This thesis choosing a suitable method deals with the original calligraphy by denoising,binarization and image optimization on the basis on the characteristics of calligraphy.In order to solve the problem that the coordinate system of the self-written calligraphy(a beginner's copied calligraphy)and the posts calligraphy(calligraphy character in copybook)is different,an image correction algorithm of calligraphy based on vector skeleton is proposed in this thesis.The algorithm firstly extracts strokes from the self-written calligraphy and the posts calligraphy by a stroke extraction algorithm of calligraphy,and sorts and numbers the strokes based on the number of stroke pixels of posts calligraphy.Then,the skeleton is extracted from all strokes of the self-written calligraphy and the posts calligraphy.The line representing the trend of the stroke skeleton is obtained by a sectional line fitting,and the vector skeleton of the self-written calligraphy and the posts calligraphy are obtained by vectorizing the line.Get the sum of relative angle difference between the self-written calligraphy and the posts calligraphy vector skeleton.And,the correction angle of the self-written calligraphy is got according to the optimization method of relative angle difference sum minimization.Then,the coordinate system of the self-written calligraphy image is reconstructed according to the correction angle,thus the direction correction of the self-written calligraphy image is completed.Finally,the self-written calligraphy image and the posts calligraphy image are cut through the minimum bounding rectangle to complete the normalization correction of calligraphy image.After the correction of calligraphy image,this thesis establishes the evaluation standard of the self-written calligraphy character combining the characteristics of calligraphy study and the standard of calligraphy elementary examination.The evaluation standard of the self-written characters is established,which is the similarity computation of the self-written calligraphy from the two aspects of stroke and structure characteristics of the posts calligraphy.Hence,the computer evaluation of the self-written calligraphy have been implemented.Among them,the similarity computation of stroke features are mainly from four aspects:stroke length,stroke width,start and finish of stroke and stroke trend.And the similarity computation of the self-written calligraphy of structural features contains three parts:the position of center of gravity,the height of calligraphy and the angle characteristics.The experiment proves that this research can give an evaluation of the self-written calligraphy by means of computer.