Research On The Extraction Method Of Dynamic Ink And Glyph Skeleton In The Video Of Brush Movement

The development of Chinese calligraphy had a history of more than 3,000 years,occupying an important position in the traditional art culture of the Chinese nation.It used Chinese characters as a carrier,brushes as its main creative tool,and its brushwork,ink method,and calligraphy as its means of expression.In today’s network age,digitizing the calligraphy creation process with the help of computer technology was of great significance to carry forward Chinese national culture.First of all,in order to digitized the dynamic process and writing effect of brush writing,this paper proposed a method based on video recording to extract the ink and its changes during the writing process.This method took a video in a natural writing state as input,and used the motion foreground detection algorithm to obtain the ink and its change information in the video through image frame fetching,convex hull detection,convex hull filling,and image subtraction,and realized the brush pen image and ink image to separate from each other.This method solved the problems associated with collecting input images through sensors fixed on the brush during digitization of calligraphy,as well as the difficulty of separating the brush tip and ink due to similar colors during the writing process.The experimental results showed that the method was feasible and effective,and could extract the dynamic changes of ink from the video written by the brush.Secondly,in order to further refine the ink image extracted above,this paper proposed two methods for skeleton extraction of Chinese calligraphy images:1.Proposed a method for skeleton extraction of calligraphy Chinese characters based on ink increment.Firstly,based on the images that had obtained a large amount of ink changes in the writing video of brush writing,the increment of the ink image was obtained by calculating the difference between the two adjacent images;secondly,the convex hull and the center of gravity of the convex hull were calculated;finally,connected and mapped to extract the skeleton of Chinese calligraphy.2.This paper proposed a method for extracting skeletons of calligraphy Chinese characters based on ink contours.By subtracting the contours from the original ink images and continuing to subtract the contours from the new ink images,the Chinese characters could be reduced in weight.The related thinning algorithms were used to further processe the ’thinning’ Chinese character to extract its skeleton.Finally,the method of this paper was compared with the methods in other literatures on the extraction effect of calligraphic Chinese character image skeleton.The experimental results showed that the skeleton of Chinese calligraphy images extracted by this method had many advantages such as single pixel,high connectivity,and no false branches.