Computer estimation method of Chinese character writing shape beauty degree

Abstract

The invention discloses a computer assessment method of aesthetic degree of writing morphology of Chinese characters. The method is obtained by machine learning based on the scoring of the aesthetic degree of an artificially given writing sample of Chinese characters. The decomposition of strokes and the parameterization of the writing morphology of Chinese characters are firstly carried out, and then the potential relationship between the writing morphology of Chinese characters and the aesthetic degree is obtained by using the image processing and the artificial intelligent method. The user intelligent interaction is introduced during the process of the decomposition of strokes and the parameterization in the first step, thus improving the processing ability of illegible handwriting. The computer assessment method respectively carries out the scoring of the aesthetic degree of each individual stroke in the writing morphology of Chinese characters, the spatial structure in all parts of the character form and the consistency of the writing style of all parts of the character form during the process in the second step of learning the relationship between the writing morphology and the artificial scoring; and finally the score of the overall aesthetic degree of the writing morphology of Chinese characters is given out by integrating the scoring results. The method has the advantage that the method can be automatically implemented by a computer.

Description

technical field [0001] The invention relates to the fields of computer art and aesthetics and artificial intelligence, in particular to a computer evaluation method for the aesthetics of Chinese characters. Background technique [0002] There have been a lot of work to simulate human artistic thinking and further build computer intelligence systems to solve real-world problems. In terms of Chinese characters, the 1995 Proceedings of the International Conference on Computer Processing of Oriental Languages ​​(ICCPOL) conference proceedings (article title "Chinese glyph generation using character composition and beauty evaluation metrics") published a heuristic method to try to quantify The problem of evaluating the aesthetics of Chinese fonts: They defined four rules in Chinese character writing and implemented them into their rule-based aesthetic scoring module; this module simply calculates the corresponding scores for the four rules one by one, and gets their weighted sum ...

AI Technical Summary

Problems solved by technology

Many people who have used expert systems in their work know that advanced expert rules don't always work; and sometimes this is not necessarily due to knowledge blind spots in the expert system itself, or the problem simply cannot be summed up

Images