A robot calligraphy writing method based on style transferring algorithm and similarity evaluation
Autor: | Liang Dongtai, Li Ping, Wu Xiaocheng, Shu-min Xing, Dan Liang |
---|---|
Rok vydání: | 2019 |
Předmět: |
0209 industrial biotechnology
Similarity (geometry) Character (computing) Computer science Mechanical Engineering 010401 analytical chemistry Computational Mechanics Process (computing) Writing process 02 engineering and technology 01 natural sciences 0104 chemical sciences 020901 industrial engineering & automation Calligraphy Writing system Artificial Intelligence Font Robot Engineering (miscellaneous) Algorithm |
Zdroj: | Intelligent Service Robotics. 13:137-146 |
ISSN: | 1861-2784 1861-2776 |
Popis: | This paper proposes a robot calligraphy writing method based on style transfer algorithm, which can transfer the art style of Chinese font and maintain the imitation accuracy of the calligraphy. A prototypical similarity evaluation approach is also presented to analyze the calligraphy writing effect, including the similarity ratio of balance, inclination and writing intensity. The robot writing system mainly consists of a six-DOF robot, manipulator, Chinese brush and PC controller. Firstly, input character images are transformed into target style font through training and style transferring process based on generative adversarial networks method. Then, the description parameters of the character images with target style are extracted to control the robot calligraphy writing process precisely. Experimental results show that the minimum similarity between the styles of robot writing characters and target characters is larger than 0.75. The proposed method shows good calligraphy writing effect, which can be used for various calligraphy training processes and trajectory planning applications. |
Databáze: | OpenAIRE |
Externí odkaz: |