On-line Chinese Character Recognition Research Articles

Stroke-order-free on-line Chinese character recognition by stroke adjustment of two-layer bipartite weighted matching

Dynamic programming (DP) is good at recognizing on-line standard-stroke-order Chinese characters. We propose a two-stage bipartite weighted matching to rearrange the stroke order of a test on-line Chinese handwriting before recognition. At the first layer, for each stroke pair which matches one stroke in the test handwriting with one stroke in the reference character, we generate a stroke-based vector graph (SVG) and the bipartite weighted matching determines the best stroke permutation of the handwriting, stroke-based matched vector graph (SMVG), from the SVG. At the second layer, we superimpose all SMVGs to form a character-based vector graph (CVG), and the bipartite weighted matching calculates the final stroke mapping between the test handwriting and the reference character. Experimental results reveal that the modified split-and-merge DP matching using the stroke adjustment method recognizes on-line stroke-order-varied handwritings with accuracy 89%, much higher than accuracy 5.2% for the original split-and-merge DP matching.

Chinese character recognition: history, status and prospects

Chinese character recognition (CCR) is an important branch of pattern recognition. It was considered as an extremely difficult problem due to the very large number of categories, complicated structures, similarity between characters, and the variability of fonts or writing styles. Because of its unique technical challenges and great social needs, the last four decades witnessed the intensive research in this field and a rapid increase of successful applications. However, higher recognition performance is continuously needed to improve the existing applications and to exploit new applications. This paper first provides an overview of Chinese character recognition and the properties of Chinese characters. Some important methods and successful results in the history of Chinese character recognition are then summarized. As for classification methods, this article pays special attention to the syntactic-semantic approach for online Chinese character recognition, as well as the metasynthesis approach for discipline crossing. Finally, the remaining problems and the possible solutions are discussed.

Two-layer assignment method for online Chinese character recognition

A method of stroke order and number-free online recognition of Chinese characters is proposed. Both input characters and model characters are represented with complete relational graphs (CRGs). Classification of an input character can be implemented by matching its CRG against every CRG of the model base. However, efficient algorithms for graph matching are not available. Therefore, the graph-matching problem is transformed into a two-layer assignment problem and is solved with the Hungarian method. Two complexity reduction schemes are presented to save computational time. Tests demonstrate the efficiency of the proposed method.

Candidate Selection in On-Line Chinese Character Recognition System Using Voting Scheme

In this paper, a candidate selection method using a voting scheme is proposed for speeding up a on-line Chinese character recognition system. Three steps in this method are described as follows: First, several simple features are extracted from the input ink data, such as peripheral code. Then, the number of votes which denote the coarse matching scores between input features and 5401 reference features of Chinese characters are computed. Finally, the templates of Chinese characters whose votes are higher than a prescribed threshold are selected to perform detailed matching. Some experimental results show that the proposed selection method is a suitable tool for speeding up a character recognition system under the condition of maintaining the recognition rate.

An optimal line segment extraction algorithm for online Chinese character recognition using dynamic programming

In this paper, a dynamic programming algorithm which can extract the line segments from the loci of Chinese character in running hand or cursive writing is proposed. This algorithm tries to find the minimum line segments to fit the cursive stroke of character within the prescribed error bound. Three error measures are presented to construct a globally optimal approximation to the given stroke, including sum of deviation, peak deviation and length error. Extensive experiments are demonstrated to illustrate the performance of the proposed algorithm.

On-Line Handwritten Chinese Character Recognition Directed by Components with Dynamic Templates

Structural matching has been recognized as a promising approach for on-line Chinese character recognition. In order to reduce its great computational complexity and improve its performance, people have been seeking for ways to direct the matching of a whole character by the result of partial matching. In this paper, the authors proposed 45 basic components for 3,755 categories of daily-used Chinese characters to direct the stroke segment matching of whole characters. Since they are always located at either the beginning or the end of the stroke segment string of characters, these components are easy to be extracted and separated from other parts of a character. Besides, in our approach, the reference templates of these components are extracted dynamically from the corresponding segment string of characters when a specific matching is carried out. This strategy avoids building multiple templates for the components of the same kind but at different places of characters. The experiments show that the segment matching computation has been reduced greatly without reducing the correctness of matching.

An approximate string matching algorithm for on-line Chinese character recognition

We model a Chinese character by a linear sequence of straight line segments. The 2-dimensional structure of the character is captured by the use of pseudo strokes and stroke length attributes. An effective string match algorithm has been developed which can tolerate common types of distortions, e.g. connected strokes, missing/extra strokes, and variations in writing sequence. A prototype system with 5401 commonly used Chinese characters has been built, and good recognition performance has been achieved.

On-Line Chinese Character Recognition Via a Representation of Spatial Relationships Between Strokes

Chinese characters are constructed by basic strokes based on structural rules. In handwritten characters, the shapes of the strokes may vary to some extent, but the spatial relations and geometric configurations of the strokes are usually maintained. Therefore these spatial relations and configurations could be regarded as invariant features and could be used in the recognition of handwritten Chinese characters. In this paper, we investigate the structural knowledge in Chinese characters and propose the stroke spatial relationship representation (SSRR) to describe Chinese characters. An On-Line Chinese Character Recognition (OLCCR) method using the SSRR is also presented. With SSRR, each character is processed and is represented by an attribute graph. The process of character recognition is thereby transformed into a graph matching problem. After careful analysis, the basic spatial relationship between strokes can be characterized into five classes. A bitwise representation is adopted in the design of the data structure to reduce storage requirements and to speed up character matching. The strategy of hierarchical search in the preclassification improves the recognition speed. Basically, the attribute graph model is a generalized character representation that provides a useful and convenient representation for newly added characters in an OLCCR system with automatic learning capability. The significance of the structural approach of character recognition using spatial relationships is analyzed and is proved by experiments. Realistic testing is provided to show the effectiveness of the proposed method.

On-Line Chinese Character Recognition by Incorporating Human Knowledge

A stroke order and number free method for on-line recognition of Chinese characters is proposed. Both input characters and the model characters are represented with complete attributed relational graphs (ARGs). The ARGs of the model base are built according to the human knowledge of the segment relation structure of Chinese characters. For the recognition purpose, an optimal matching measure between two ARGs is defined, and the graph matching is formulated as a search problem of finding the minimum cost path in a state space tree, using the A * algorithm. Moreover, to reduce the search time of the A *, besides a heuristic estimate, a novel strategy is employed which again uses the human knowledge of the segment position structure of Chinese characters to prune the tree. Our experimental results demonstrate the efficiency of the proposed method.

Intelligent radical‐based on‐line Chinese character recognition system

An intelligent radical-based on-line Chinese character recognition (OLCCR) system is proposed. We make use of the characteristic that most Chinese characters possess radicals and develop a radicalbased preclassification technique. In the procedure of recognizing radicals, potential candidate radicals are identified from an input character first and then matched with radical templates. Based on the matched radicals, corresponding candidate characters with a smaller matching distance are thus identified. Postprocessing, which utilizes the information of the first and last stable line segments, is incorporated to improve the recognition accuracy. Experiments are carried out to evaluate the performance of the proposed OLCCR system and the average recognition rate is improved from 96.06 to 96.52% if only the first winner is selected and the average recognition time is shortened by about 44%.

On-line Chinese character recognition using ART-based stroke classification

This paper proposes an on-line Chinese character recognition method using Adaptive Resonance Theory (ART) based stroke classification. Strokes, primitive components of Chinese characters, are usually warped into a cursive form and classifying them is very difficult. To deal with such cursive strokes, we consider them as a recognition unit and automatically classify them using an ART-2 neural network. The neural network has the advantage of assembling similar patterns together to form classes in a self-organized manner. This stroke classifier contributes to high stroke recognition rate and less recognition time. A database for character recognition also dynamically constructed with generalized character lists, and a new character can be included simply by adding a new sequence to the list. Character recognition is achieved by traversing the Chinese character database with a sequence of recognized strokes and positional relations between the strokes. To verify the performance of the system, we tested it on 1800 basic Chinese characters used daily in Korea, and obtained a good recognition rate of 93.13%. These results suggest that the proposed system is pertinent to be put into practical use.

On-line chinese character recognition with effective candidate radical and candidate character selections

A useful property of Chinese characters is that most of them possess a radical with a meaning. Our system is motivated by this characteristic found in Chinese characters which leads us to radical-based candidate selection approaches, performed before similarity measurement. The searching ranges of the radicals and the characters can be basically decided by the number of strokes in the input script. From the recognized possible radicals, the candidate characters are limited to those having such radicals and those having no radical. Since some of the candidate radicals may have very high matching costs, we can reduce the recognition time by discarding those unlikely candidate radicals. Furthermore, to speed up the inspection of radicals, we developed a radical extraction algorithm to narrow down the searching scope of the reference templates. The radical extraction algorithm was further improved by eliminating false extracted radicals. With these mechanisms, the number of candidate radicals screened out is 286 out of 726 radical templates and the number of candidate characters to be detailed matched is 123 out of 5401 Chinese characters. Through these efforts, the recognition rate improves to be 96.35% for the first rank and 98.96% for the first 10 result candidates with the speed of 0.427 seconds on average per character on a PC using Intel 386-33 CPU.

Online Chinese character recognition using attributed relational graph matching

A structural method for online recognition of Chinese characters is proposed which is stroke-order free and allows variations in stroke type and stroke number. Both input characters and the model characters are represented with complete attributed relational graphs (ARGs). An optimal matching measure between two ARGs is defined. Classification of an input character can be implemented by inexactly matching its ARG against every ARG of the model base. The matching procedure is formulated as a search problem of finding the minimum-cost path in a state space tree using the A* algorithm. To speed up the search of the A*, in addition to a heuristic estimate, a novel strategy that utilises the geometric position information of strokes of Chinese characters to prune the tree is employed. The efficiency of the method is demonstrated by the promising experimental results.

On-line handwritten Chinese character recognition via a fuzzy attribute representation

Chinese characters are constructed by strokes according to structural rules. Therefore, the geometric configurations of characters are important features for character recognition. In handwritten characters, stroke shapes and their spatial relations may vary to some extent. The attribute value of a structural identification is then a fuzzy quantity rather than a binary quantity. Recognizing these facts, we propose a fuzzy attribute representation (FAR) to describe the structural features of handwritten Chinese characters for an on-line Chinese character recognition (OLCCR) system. With a FAR. a fuzzy attribute graph for each handwritten character is created, and the character recognition process is thus transformed into a simple graph matching problem. This character representation and our proposed recognition method allow us to relax the constraints on stroke order and stroke connection. The graph model provides a generalized character representation that can easily incorporate newly added characters into an OLCCR system with an automatic learning capability. The fuzzy representation can describe the degree of structural deformation in handwritten characters. The character matching algorithm is designed to tolerate structural deformations to some extent. Therefore, even input characters with deformations can be recognized correctly once the reference dictionary of the recognition system has been trained using a few representative learning samples. Experimental results are provided to show the effectiveness of the proposed method.

Attributed string matching by split-and-merge for on-line Chinese character recognition

Consecutive strokes of Chinese characters tend to be connected in fast writing, and this causes a problem for most stroke-based recognition approaches. A recognition scheme for recognizing cursive Chinese characters under the constraint of correct stroke writing orders is proposed. The recognition scheme consists of two phases: candidate character selection and detailed matching. In the former phase, an input script with N strokes is used to split the strokes of each reference character into N corresponding parts. In the latter phase, the connected input strokes are broken into multiple strokes under the guidance of candidate characters. In both phases, dynamic programming is employed for stroke or character matching. Good experimental results prove the feasibility of the proposed approach for cursive Chinese character recognition.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A neural network-based on-line Chinese character recognition system

A neural network-based on-line Chinese character recognition (OLCCR) system is presented. In this paper, a back-propagation neural network model is proposed for solving the pattern-matching problems in OLCCR, instead of those non-neural network-based algorithms. This OLCCR system will enable us to recognize handwritten Chinese characters in real time. Each recognition process includes two main phases: (A) the feature-extraction phase, in which a feature vector based on reference stroke information can be extracted, and (B) the character-matching phase, in which the feature vector is matched with a group of standard Chinese reference character feature vectors. Experimental results show that it takes about 0.25 s to recognize a handwritten character and that a recognition rate of about 91% can be achieved.

An On-Line Chinese Character Recognition System

This paper describes an experimental system for the on-line recognition of handwritten Chinese characters. Constituent shapes of the characters are recognized as they are formed on an electronic tablet. The 72 constituent shapes can define a very large set of Chinese characters. The implementation described recognizes more than 2200 Chinese characters. More symbols can be added with relative ease. Experimental results are given for two writer populations.

Approaches to chinese character recognition

Work on the recognition of Chinese characters is reviewed. All of this work has been reported in just the last nine years, and most of it since 1970. The topics covered are: printed Chinese Character Recognition—Half a dozen groups have tackled this problem, using methods as varied as syntactic description, projection profiles, template matching. Fourier transforms, and hierarchical processing: handwritten Chinese Character Recognition. This is a somewhat more difficult problem. The approaches tried include a stroke generation technique and a statistical Markov chain analysis; on-line Recognition—Several investigators have reported on work in on-line Chinese character recognition. Input for on-line recognition is dynamic, using a suitable device such as a tablet. Because the strokes are isolated in time and their sequence known, this is a substantially easier recognition task: on-line Analysis—Less ambitious than on-line recognition is the on-line analysis of Chinese characters falling short of recognition. Several useful applications have been reported, including CAI and dictionary-lookup. The paper is intended as a companion to a paper (6) on formal models of the structure of Chinese characters and their application in computerized systems.