Constructing Calligraphy Evaluation Model Based on Writing Movement with LSTM Network

Wang, Zhaoyi; Lyu, Ruimin; Liu, Xinya; Ze, Yuefeng; Xie, Zhenping; Yan, Tao

doi:10.1109/cscwd54268.2022.9776173

Cited by 2 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Calligraphy evaluation can be simply understood as evaluating the quality of writing. According to whether there is reference calligraphy characters, it can be divided roughly into two categories: template-free [1][2][3] and template-based [4][5][6][7][8][9]. Template-free methods mostly combine calligraphy writing experience and directly evaluate the writing quality based on the morphological features of calligraphy characters.…”

Section: Calligraphy Evaluationmentioning

confidence: 99%

“…Template-based methods are more suitable for calligraphy teaching scenarios. Wang [4] uses sensors to obtain motion data during the calligraphy writing process, and uses LSTM to calculate the feature differences between the target calligraphy images and the reference calligraphy images as evaluation results. Sun [9] calculated the overall calligraphy feature difference between the target calligraphy characters and the reference calligraphy characters, as well as the overall morphological feature differences at different angles, and combined the two differences as the evaluation results.…”

Section: Calligraphy Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

Intelligent Evaluation Method of Calligraphy Characters Based on Deep Stroke Extraction

2023

acss

View full text Add to dashboard Cite

Calligraphy is an important part of Chinese culture. And calligraphy education is the main way to spread calligraphy culture. Intelligent calligraphy evaluation can reduce dependence on experienced calligraphy teachers and effectively enhance the development of calligraphy culture. However, traditional intelligent calligraphy evaluation methods mostly focus on the whole and lack fine-grained analysis, which cannot form effective evaluation results. In this paper, we propose an intelligent evaluation method for calligraphy characters based on deep stroke extraction. By disassembling calligraphy character strokes, a more fine-grained evaluation of the writing results of a single stroke can be achieved. This method consists of two main parts: stroke extraction module that extracts single strokes through a structure deformable image registration-based stroke extraction model; evaluation module that provides detailed quantitative evaluation results from the whole character, radicals and single strokes. The experimental results show that our method can extract strokes of complex calligraphy characters and provide detailed evaluation results of calligraphy characters effectively.

show abstract