1、Generation of Chinese Character Based on Human Vision and Prior Knowledge of Calligraphy,报告人： 史操作者： 史操、肖建国、贾文华、许灿辉单位： 北京大学计算机科学技术研究所,NLP & CC 2012:,基于人类视觉和书法先验知识的汉字自动生成,Abstract,本文根据中国书法先验知识对汉字笔画及结构进行建模，提出一种五层框架用于表征笔画及部件间的相关性。通过分析书法家特有的书法技巧，拆分书法字为本文提出的五层框架提供充足的笔画及部件样本。根据 Marr 提出的对人类视觉的若干假设，本文提出了可量化的
2、视觉美学准则用于指导汉字自动生成。使用贝叶斯统计原理，可将整个汉字生成过程描述为一个贝叶斯动态模型。模型中状态方程控制对笔画、部件的线性变换，并且将本文提出的视觉美学准则应用于观测方程之中。,Previous Work,1. Handwriting Imitation 2. Samples Based Character Generation ( Font Companies ),Previous Work,1. Handwriting Imitation,Previous Work,2. Samples Based Character Generation,Yan Zhenqings Cal
3、ligraphy,Generated by our algorithm,Previous Work,2. Samples Based Character Generation,Yan Zhenqings Calligraphy,Generated by our algorithm,Previous Work,2. Samples Based Character Generation,Yan Zhenqings Calligraphy,Generated by our algorithm,Modeling Prior Knowledge of Calligraphy,Five layers fr
4、amework to represent Chinese character,Modeling Prior Knowledge of Calligraphy,Modeling Stroke and DFS,Modeling Prior Knowledge of Calligraphy,Modeling Prior Knowledge of Calligraphy,Modeling Radical and DFR,(a) (b) (c) (d),Automatic Generation Algorithm of Chinese Character,David Marrs “Vision”inve
5、stigates the spatial arrangement of objects in an image1. Average local intensity;2. Average size of similar objects;3. Local density of the objects;4. Local orientation of the objects;5. Local distances associated with the spatial arrangement of similar objects;6. Local orientation associated with
6、the spatial arrangement of similar objects.,Automatic Generation Algorithm of Chinese Character,David Marrs “Vision”investigates the spatial arrangement of objects in an image1. Average local intensity;2. Average size of similar objects;3. Local density of the objects;,Automatic Generation Algorithm
7、 of Chinese Character,David Marrs “Vision”investigates the spatial arrangement of objects in an image1. Average local intensity;2. Average size of similar objects;3. Local density of the objects;,Automatic Generation Algorithm of Chinese Character,David Marrs “Vision”investigates the spatial arrange
8、ment of objects in an image1. Average local intensity;2. Average size of similar objects;3. Local density of the objects;,Automatic Generation Algorithm of Chinese Character,David Marrs “Vision”investigates the spatial arrangement of objects in an image1. Average local intensity;2. Average size of s
9、imilar objects;3. Local density of the objects;,Automatic Generation Algorithm of Chinese Character,David Marrs “Vision”investigates the spatial arrangement of objects in an image4. Local orientation of the objects;6. Local orientation associated with the spatial arrangement of similar objects.,Auto
10、matic Generation Algorithm of Chinese Character,David Marrs “Vision”investigates the spatial arrangement of objects in an image4. Local orientation of the objects;6. Local orientation associated with the spatial arrangement of similar objects.,Automatic Generation Algorithm of Chinese Character,Davi
11、d Marrs “Vision”investigates the spatial arrangement of objects in an image5. Local distances associated with the spatial arrangement of similar objects;,Automatic Generation Algorithm of Chinese Character,David Marrs “Vision”investigates the spatial arrangement of objects in an image5. Local distan
12、ces associated with the spatial arrangement of similar objects;,Automatic Generation Algorithm of Chinese Character,Automatic Generation Algorithm of Chinese Character,Step 1: Input prior information of target character and absent radicals.Step 2: If all radicals to compose the character exist, then
13、 turn to Step 4.Step 3: Compose the absent radicals using the Bayesian dynamic models, similar to (11)(14).Step 4: Compose the target character using the Bayesian dynamic models.Step 5: Output the target character.,Experiments and Discussion,Experiments and Discussion,Experiments and Discussion,谢 谢 ！ http:/,
