Artistic information extraction from Chinese calligraphy works via Shear-Guided filter

Pengfei Xu, Xia Zheng, Xiaojun Chang, Qiguang Miao, Zhanyong Tang, Xiaojiang Chen, Dingyi Fang

Abstract

Chinese calligraphy is a unique visual art, and is still emitting fascinating artistic brilliance and charm. The artistic beauty of Chinese calligraphy is constituted by two elements, i.e., form and spirit, which attract continual attentions by calligraphy artists. In this paper, we present Shear-Guided filter to extract the artistic information. Shear-Guided filter is designed by combining the shear transformation and the guided filter. It has both the edge-preserving smoothing property from the guided filter and the linear multidirectional property from the shear transformation. Shear-Guided filter can extract more spirit information from the calligraphy images, especially in the half-dry strokes. Furthermore, the previous evaluation parameters consider all pixels equally, which is against the truth. Here we put forward three new parameters to solve this issue. Finally, experiments show that Shear-Guided filter can extract the information of form and spirit more accurately, and the new evaluation parameters have better performances in evaluation.
BibTeX
@article{xu2016artistic,
  author = {Pengfei Xu and Xia Zheng and Xiaojun Chang and Qiguang Miao and Zhanyong Tang and Xiaojiang Chen and Dingyi Fang},
  title = {{Artistic information extraction from Chinese calligraphy works via Shear-Guided filter}},
  journal = {J. Vis. Commun. Image Represent.},
  year = {2016},
  doi = {10.1016/j.jvcir.2016.07.012},
}

Loading related content...

Figure