Two-dimensional extended attribute grammar method for the recognition of hand-printed chinese characters

doi:10.1016/0031-3203(90)90092-Y

Pattern Recognition

Volume 23, Issue 7, 1990, Pages 685-695

https://doi.org/10.1016/0031-3203(90)90092-Y Get rights and content

Abstract

This paper introduces the design and realization of two-dimensional extended attribute grammar method (2D EAG for short) for the recognition of hand-printed Chinese characters. In this method, 2D EAGs are used to carry both overall character shape information and local features of samples, and to conduct top-down matching and bottom-up reduction. The strategies of separating element drawing and grammar processing, autonomous component drawing, two dimensional redundant reduction, and structure analogizing are proposed.

References (7)

M. Zhao
Two-dimensional extended attribute grammars method for the recognition of hand printed chinese characters
J.W. Tai
A syntactic-semantic approach for describing Chinese characters
Comput. Process. Chinese Oriental Languages
(May 1984)
D.E. Knuth
Semantics of context-free languages

There are more references available in the full text version of this article.

Cited by (13)

Hippocampus-heuristic character recognition network for zero-shot learning in Chinese character recognition
2022, Pattern Recognition
Citation Excerpt :
In the past few decades, lots of efforts have been made for radical-based Chinese character recognition. Radical-based Chinese character recognition can be considered to consist of two major problems, radical extraction [13,14] and structure analysis [15,16]. For the first problem, traditional methods usually adopt CNN as the encoder to extract high-level visual features of radicals from input Chinese characters.
The recognition of Chinese characters has always been a challenging task due to their huge variety and complex structures. The current radical-based methods fail to recognize Chinese characters without learning all of their radicals in the training stage. To this end, we propose a novel Hippocampus-heuristic Character Recognition Network (HCRN), which can recognize unseen Chinese characters only by training part of radicals. More specifically, the network architecture of HCRN is a new pseudo-siamese network designed by us, which can learn features from pairs of input samples and use them to predict unseen characters. The experimental results on the recognition of printed and handwritten characters show that HCRN is robust and effective on zero/few-shot learning tasks. For the printed characters, the mean accuracy of HCRN outperforms the state-of-the-art approach by 23.93% on recognizing unseen characters. For the handwritten characters, HCRN improves the mean accuracy by 11.25% on recognizing unseen characters.
Radical analysis network for learning hierarchies of Chinese characters
2020, Pattern Recognition
Citation Excerpt :
Moreover, the CNN can be applied to online characters by first transforming the online handwriting trajectory into image-like representations such as AMAP[28], path signature maps [29] and directional feature maps [30]. The use of radicals for Chinese character recognition has been researched for decades, and radical-based Chinese character recognition can be considered to consist of two major problems, namely, radical extraction [31,32] and structure analysis [33,34]. The two problems can be solved sequentially or globally.
Chinese characters have a valuable property, this is, numerous Chinese characters are composed of a compact set of fundamental and structural radicals. This paper introduces a radical analysis network (RAN) that makes full use of this valuable property to implement radical-based Chinese character recognition. The proposed RAN employs an attention mechanism to extract radicals from Chinese characters and to detect spatial structures among the radicals. Then, the decoder in RAN generates a hierarchical composition of Chinese characters based on the knowledge of the extracted radicals and their internal structures. The method of treating a Chinese character as a composition of radicals rather than as a single character category is a human-like method that can reduce the size of the vocabulary, ignore redundant information among similar characters and enable the system to recognize unseen Chinese character categories, i.e., zero-shot learning. Through experiments, we assess the practicality of RAN for recognizing Chinese characters in natural scenes. Furthermore, a RAN framework can be proposed for scene text recognition with the extension of a dense recurrent neural network (denseRNN) encoder, a multihead coverage attention model and HSV representations. The proposed approach achieved the best performance in the ICPR MTWI 2018 competition.
Optical recognition of handwritten Chinese characters by hierarchical radical matching method
2001, Pattern Recognition
In this paper, a radical-based OCR system for the recognition of handwritten Chinese characters is proposed. In our approach, a recursive hierarchical scheme is developed to perform radical extraction first. Character features and radical features are then extracted for matching. Last, a hierarchical radical matching scheme is devised to identify the radicals embedded in an input Chinese character and recognize the input character accordingly. Experiments for radical extraction are conducted on 1856 characters. The successful rate of radical extraction is 92.5%. The average time for radical extraction is 0.65 second per character. Experiments for matching process are conducted on two sets: training set and testing set, each set includes 900 characters. The overall recognition rate in our experiments is 98.2 and 80.9% (for training set and testing set, respectively). The average recognition time of our hierarchical radical matching scheme is 0.274 s. There are totally 4716 radicals in 1800 characters. In average, one character consists of 2.62 radicals. Each character will match 7.28 radical templates in average. Thus, each radical will match 2.77 radical temples. The experimental results reveal that our proposed method is feasible, flexible, and effective.
Theoretical aspects of syntactic pattern recognition
1995, Pattern Recognition
In this paper, a pessimistic view of syntactic pattern recognition is developed. This paper describes several aspects of syntactic pattern recognition from various points of view including the relation between the set of patterns and grammars, the semantics of a grammar, the expressive power of a grammar, grammatical inference, comparison between a syntactic method and a statistical method and a comparison of computing costs between a syntactic method and a prototype matching method. In every aspect a syntactic method faced with difficulties that are very hard to overcome.
Optical recognition of handwritten Chinese characters: Advances since 1980
1993, Pattern Recognition
The work on the optical recognition of handwritten Chinese characters which has been reported since 1980 is reviewed. The various methods used are classified according to the feature and comparison methods employed. Particular attention is paid to models developed under the recently introduced connectionist paradigm. As an area with sufficient challenge, considerable application potential, and extensive test data, handwritten Chinese character recognition is a good testbed for new recognition algorithms. It therefore enables connectionist models to be compared with more traditional recognition methods.
A genetic algorithm approach to chinese handwriting normalization
1997, IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

View all citing articles on Scopus

^☆: Supported by National Natural Science Foundation of China, No. 6883024.

View full text

Two-dimensional extended attribute grammar method for the recognition of hand-printed chinese characters☆

Abstract

Two-dimensional extended attribute grammars method for the recognition of hand printed chinese characters

A syntactic-semantic approach for describing Chinese characters

Comput. Process. Chinese Oriental Languages

Semantics of context-free languages