关键词: 非真实感绘制(NPR), 中国画, HTML5, 分段函数, Kmeans聚类算法

Abstract:

By combining the ideas of random concentration in physical simulation and the piecewise approach,the random process of diffusion density of the middle-long distance segment is performed,then a novel subparagraph ink spread function is proposed for simulating the ink diffusion effect.A novel rendering method of Chinese painting style based on piecewise function is proposed.According to the design key points,the overall framework of the proposed method is designed.Firstly,the mean filter is used for smoothing preprocessing of the image.Secondly,the Kmeans clustering algorithm is adopted for image segmentation so as to simulate the tonal gradation effect expressed by ink shades in Chinese painting.Thirdly,the morphological method is used to detect the edge points.Fourthly,the self-defined piecewise function is applied to the detected edge points to simulate the effect of the diffusion of ink edges,thus obtaining the rendered image of Chinese painting style.Finally,HTML5 is used to implement the proposal. The proposed algorithm can generate a more realistic effect for ink simulation and ink diffusion of Chinese painting style.     

Key words: non-photorealistic rendering;Chinese painting;HTML5;segmented function;K-means clustering algorithm