Dai Ruwei

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.

An interactive image segmentation method based on dynamic programming and its application in medical image analysis

In this paper we present an interactive image segmentation method based on DP (dynamic programming), and extend the method in two aspects. First, we append a training mechanism to enhance the robustness of our approach to local noise; Second, we integrate the region-based segmentation with DP to improve the accuracy of image segmentation. A number of experiments show that our approach performs well on a variety of medical images.

The control structure of artificial creatures

The aim of this article is to integrate some ideas from the science of complexity, behavior-based AI, and the theory of metasynthesis for intelligence systems, and to design a computational model for a brief implementation of these ideas. Our simulated microworld is a two-dimensional grid containing some resources including food and water, walls, shade, bugs, and an artificial creature. This artificial creature will fulfill a set of goals in a complex, dynamic, and unfriendly environment. The creature consists of a set of self-interested agents, and has life-like characteristics by means of interactions between its lifeless agents, as well as the interactions between the creature and its environment. The experimental result demonstrates the usefulness of this model, and this is only the first step toward our ultimate goal.The aim of this article is to integrate some ideas from the science of complexity, behavior-based AI, and the theory of metasynthesis for intelligence systems, and to design a computational model for a brief implementation of these ideas. Our simulated microworld is a two-dimensional grid containing some resources including food and water, walls, shade, bugs, and an artificial creature. This artificial creature will fulfill a set of goals in a complex, dynamic, and unfriendly environment. The creature consists of a set of self-interested agents, and has life-like characteristics by means of interactions between its lifeless agents, as well as the interactions between the creature and its environment. The experimental result demonstrates the usefulness of this model, and this is only the first step toward our ultimate goal.

Models of reservoir regulation based on RBF neural networks

Reservoir regulation is a very complicated problem with much nonlinear relation. The traditional way of reservoir regulation cannot meet the demands of production. In this paper, a new models of reservoir regulation based on RBF neural network is presented. The historical datum of reservoir regulation is used to train RBF neural network in order to improve the precision of the RBF neural network models of reservoir regulation. The boosting algorithm is used to build an integration-neural network models for reservoir regulation. Experiment results have shown good performance in the actual situation with significant economy benefits.

Study diploid system by a hamiltonian cycle problem algorithm

Complex representation in Genetic Algorithms and pattern in real problems limits the effect of crossover to construct better pattern from sporadic building blocks. Instead of introducing more sophisticated operator, a diploid system was designed to divide the task into two steps: in meiosis phase, crossover was used to break two haploid of same individual into small units and remix them thoroughly. Then better phenotype was rebuilt from diploid of zygote in development phase. We introduced a new representation for Hamiltonian Cycle Problem and implemented an algorithm to test the system.

Using association rules to improve Chinese handwritten character recognition

Association rules are useful for determining correlation between attributes of a relation and have applications in marketing, financial and retail sectors. In this paper, we present an approach for combining handwritten character classifiers based on association rules, which reflect the correlation between the classifiers. The experimental results show that the association rules improve the performances of the integrated system significantly. An experimental comparison of two combination schemes is also provided.