Researchand Implementationofan Intelligent Evaluation Method Of Chinese Calligraphy

Chinese characters are the symbol of Chinese civilization,and calligraphy is the essence of Chinese character culture.At present,calligraphy education has entered a special period of popularization and deepening.In view of the current calligraphy teaching,there are intractable problems in the lack of calligraphy teaching teachers,uneven levels and random confusion in teaching content.This topic is divided into three parts: calligraphy Chinese character recognition,calligraphy Chinese character evaluation and French Chinese character auxiliary teaching system.The system evaluates and guides calligraphy practice works completed when no one is on duty,and solves the problem of lack of teachers in calligraphy education.First of all,the diversification of calligraphy image data in form and fonts,and the scarcity of samples have increased the difficulty and challenge of calligraphy Chinese character recognition.This paper uses deep learning image recognition technology to combine calligraphy Chinese character images,and extracts high-level images through convolutional neural networks.Abstract features to complete the task of identifying calligraphic Chinese characters;secondly,there are more subjective judgments in the evaluation of calligraphic Chinese characters,which is a difficult to qualitative evaluation.This article focuses on the global and local features of the skeleton of calligraphic Chinese characters based on image processing technology.After analysis and comparison,the evaluation task of calligraphy Chinese characters is completed by calculating the similarity of features;finally,the calligraphy Chinese characters' auxiliary learning system is completed by outputting the evaluation scores given and giving reasonable practice suggestions.The main research contents of this paper are as follows:First: This article uses deep learning methods in the recognition module to compare the network structure of Google Net and Res Net networks.By performing offline handwritten Chinese character experiments on these two networks,95.97% and 96.63%recognition results are obtained.The experimental results It shows that Res Net network has high recognition accuracy of handwritten Chinese characters.At the same time,the Res Net network model obtained through experiments also has high recognition accuracy for the recognition of calligraphy Chinese characters.Second: This paper proposes a skeleton extraction algorithm based on Chinese characters,a comprehensive evaluation method for a single calligraphic Chinese character image,that is,local feature and global feature analysis.The local features start from the calligraphy Chinese character strokes and analyze the similarity of the corners of the Chinese character strokes;the global features start from the overall structure of the Chinese characters,and the Chinese characters are split into the overall "?" pattern,and the split is calculated based on the Hu moment and Pearson coefficient Similarity.At the same time,the reliability of the system proposed in this paper is verified bycomparing manual scoring.Third: Constructing an auxiliary learning system for calligraphy and Chinese characters.The system realizes scoring and evaluation feedback on the calligraphy practice results input by users,so that users can more intuitively discover the deficiencies of their own works and correct them after completing the calligraphy practice.