• Korean Journal of Cognitive Science
• /
• v.7 no.1
• /
• pp.75-84
• /
• 1996

In this paper we emphasize the importance of architectural aspects of designing a handwritten numeral recognition program. and describe two architectural design.First, we describe the modular design of a numeral recognition program, and mention its advantages.In this design, a recognizer is composed of 10 binary subrecognizers each of which is responsible for only one class.Rule-based training and neural-based training are presented.Second, we connect two(or more)recognizers serially which we call pipelining connection.The second recognizer may act as verifier for the patterns recognized by the forst recognizer, or as second chance recognizer for the patterns rejected by the first recognizer.Our experimental results obtained till now show the merits of the proposed architectural designs.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.04a
• /
• pp.557-560
• /
• 2002

본 논문에서 제안하는 분절된 화소들의 특징추출 방법은 이진화 영상에서 수직/수평 화소들의 분절점을 탐색하여 추출하는 특징 탐색기이다. 숫자의 구조적인 면을 고려하여 사소한 부분들도 명확한 특징으로 탐지하여 추출하였고, 이러한 방법은 일반적으로 사용하여지는 특징추출 방법 몇가지를 선택하여 이용하였고, 제안하는 방법과 결합하여 필기체 숫자를 인식하였다. 인식기를 구현하기 위하여 3 개층 구조를 갖는 클러스터 MLP 신경망을 사용하였다. 실험 결과 단순히 일반적인 특징만을 활용하여 얻는 인식률 보다 훨씬 향상됨을 보여주었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.05a
• /
• pp.182-185
• /
• 2008

테이블 탑 디스플레이는 사람에게 친숙한 상호작용의 수단인 손을 인터페이스 수단으로 이용하는 일종의 탁자형 터치스크린이다. 본 논문에서는 이러한 환경에서 사용자 독립적인 온라인 필기 숫자를 인식하는 연구를 수행하였다. 이로 인해 추후 진행될 다중 사용자의 한글, 영문, 특수기호의 인식 가능성을 확인하였다. 실험 과정은 테이블 탑 디스플레이의 표면을 통해 입력된 사용자별 손가락 궤적으로 기준점을 잡고, 각 사용자별 필기궤적에서 대표점 추출과 16-방향 체인코드변환을 수행하였다, 변환된 체인코드의 학습 및 필기숫자 인식에 확률 통계적 모델인 은닉 마르코프 모델을 이용하였다. 실험에 사용된 데이터는 총 300개의 데이터를 사용 하였고, 학습은 10회 복제하여 총 3000개의 데이터로 수행하였다. 각 사용자별 데이터를 100개씩 인식 실험에 사용하여 각각 93%, 94%의 정인식율을 보였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.11a
• /
• pp.9-12
• /
• 2007

테이블 탑 디스플레이는 사람에게 친숙한 상호작용의 매개체인 손을 입력장치로 이용하는 일종의 탁자형 멀티 터치스크린이라고 할 수 있다. 본 논문에서는 이러한 환경에서 손가락 제스쳐를 활용하여 필기 숫자를 인식하는 연구를 수행함으로써 테이블 탑 디스플레이에 적합한 필기 숫자 인식 기술을 개발하였고, 이로 인해 추후 진행될 연속 숫자 혹은 특수기호의 성공적인 인식 가능성을 확인하였다. 실험 과정은 테이블 탑 디스플레이의 표면을 통해 입력된 손가락 궤적을 잡음제거, 대표점 추출등의 전처리 과정을 거쳐 16-방향 체인코드로 변환하고, 변환된 체인코드의 학습 및 필기 숫자 인식에 확률 통계적 모델인 은닉 마르코프 모델을 이용하였다. 학습에는 총 300개 필기 숫자 데이터를 이용하였고, 인식 실험에 사용한 별도의 100개의 필기 숫자 데이터에 대해 97%의 정인식율을 보였다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.7
• /
• pp.675-679
• /
• 2001

In this paper, we present a new competitive learning algorithm called Dynamic Competitive Learning (DCL). DCL is a supervised learning method that dynamically generates output neurons and initializes automatically the weight vectors from training patterns. It introduces a new parameter called LOG (Limit of Grade) to decide whether an output neuron is created or not. If the class of at least one among the LOG number of nearest output neurons is the same as the class of the present training pattern, then DCL adjusts the weight vector associated with the output neuron to learn the pattern. If the classes of all the nearest output neurons are different from the class of the training pattern, a new output neuron is created and the given training pattern is used to initialize the weight vector of the created neuron. The proposed method is significantly different from the previous competitive learning algorithms in the point that the selected neuron for learning is not limited only to the winner and the output neurons are dynamically generated during the learning process. In addition, the proposed algorithm has a small number of parameters, which are easy to be determined and applied to real-world problems. Experimental results for pattern recognition of remote sensing data and handwritten numeral data indicate the superiority of DCL in comparison to the conventional competitive learning methods.

• Journal of KIISE:Software and Applications
• /
• v.34 no.11
• /
• pp.1001-1010
• /
• 2007

HMM tends to output high probability for not only the proper class data but confusable class data, since the modeling power increases as the number of parameters increases. Thus it may not be helpful for discrimination to simply increase the number of parameters of HMM. We proposed two methods in this paper. One is a CMC(Confusion Likelihood Model Selection Criterion) using confusion class data probability, the other is a new recognition method, RCM(Recognition Using Confusion Models). In the proposed recognition method, confusion models are constructed using confusable class data, then confusion models are used to depress misrecognition by confusion likelihood is subtracted from the corresponding standard model probability. We found that CMC showed better results using fewer number of parameters compared with ML, ALC2, and BIC. RCM recorded 93.08% recognition rate, which is 1.5% higher result by reducing 17.4% of errors than using standard model only.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.9
• /
• pp.1358-1367
• /
• 1993

In this paper, OMR system based on image processing is developed which improve the performance of conventional OMR system based on line-scan method. Based on this OMR system, real-time OCR system which recognizes alphanumerics is also developed. We propose the OMCR system which recognize the mark and numerals at the same time. Besides, we improve the input system using constrained 7-segment type handwritten numeral instead of mark to solve the problem caused by miswriting the mark. In summary, we verified the reai-time recognition performance of developed OMCR system using application form for admission, answer sheet for college entrance examination and receipt sheet.

• Journal of KIISE:Software and Applications
• /
• v.28 no.9
• /
• pp.681-687
• /
• 2001

In order to raise a class discrimination power by combining multiple classifiers under the Bayesian decision theory, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables obtained from training data samples should be minimized. Wang and Wong proposed a tree dependence first-order approximation scheme of a high order probability distribution composed of the class and multiple feature pattern variables for minimizing the upper bound of the Bayes error rate. This paper presents an extended high order product approximation scheme dealing with higher order dependency more than the first-order tree dependence, based on the minimization of the upper bound of the Bayes error rate. Multiple recognizers for unconstrained handwritten numerals from CENPARMI were combined by the proposed approximation scheme using the Bayesian formalism, and the high recognition rates were obtained by them.