• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.15-17
• /
• 2021

Handwriting recognition research is mainly focused on deep learning techniques and has achieved a great performance in the last few years. Especially, handwritten Thai digit recognition has been an important research area including generic digital numerical information, such as Thai official government documents and receipts. However, it becomes also a challenging task for a long time. For resolving the unavailability of a large Thai digit dataset, this paper constructs our dataset and learns them with some variants of the CNN model; Decision tree, K-nearest neighbors, Alexnet, LaNet-5, and VGG (11,13,16,19). The experimental results using the accuracy metric show the maximum accuracy of 98.29% when using VGG 13 with batch normalization.

• Korean Journal of Heritage: History & Science
• /
• v.46 no.2
• /
• pp.150-165
• /
• 2013

This study aims to investigate the original position, the writer of signboard handwriting and the period of Wibongmun and Joyangru in order to restore Wibongmun and Joyangru which have been designated as tangible cultural properties (有形文化財). They also have to be moved in the Gangwon Provincial Office. Wibongmun and Joyangru were established as government offices in chuncheon(春川官衙) and they were used as attached buildings in Chunceon (春川離宮) in 1890. Wibongmun was moved to Gangwon Provincial Office 5 times and Joyangru was moved twice. In order to move them back to the original place, by using the topographic map made by the Japanese Government-General in Korea, we find out Joyangru was located in the exit of Gangwon Provincial Office and greenhouse and we also figure out Wibongmun was located in the garden. While we study historical evidence on handwriting, we also find out the handwriting of Joyangmun was written by Songhaong (松下翁) Jo, Yun-Hyeong (曺允亨). Especially, Joyangru had played a role as a government office and it may be called 'Joyangru' after reconstructing 'Joyangmun' when attached buildings were established. Through this study, we found that the first period and reason of establishing Wibongmun and Joyangru was at least before 1788. Through this study, we can find the period of both and its historic meaning more clearly.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06b
• /
• pp.275-280
• /
• 2008

본 논문에서는 한글 글립(glyph)의 형태학적 분석을 통해 자동으로 획을 분할하는 방법을 제안한다. 제안된 방법은 thinning된 한글 글립의 골격(skeleton) 이미지를 기반으로, 획 분리, 획 병합, 그리고 획 볼륨 복원의 세가지 단계를 거쳐 한글의 기본 획들을 추출해 낸다. 실험 결과, 트루타입폰트(TrueType Font)에 대해서는 80%, 손글씨(Handwriting) 글립에 대해서는 72%의 획 분할 정확도를 보였다. 본 논문에서 제안한 방법으로 획득된 획 정보를 이용하여, 향후 한글 손글씨 생성을 위한 연구를 하고자 한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.790-792
• /
• 2005

본 연구에서는 3차원 온라인 필기의 효과적인 인식 방법을 제안한다. 3차원 필기 시 pen-up/pen-down 정보의 구분이 없이 입력하도록 하여 사용자가 편리하게 필기하도록 하고 구분의 부정확함으로 인해 발생하는 오류를 줄인다. 또한, 기존의 2차원 필기 인식 모델을 개선하여 3차원 필기 데이터의 특성을 반영하게 함으로써 경제적이며 안정적인 인식이 가능하다. 실험 결과 제안된 인식 방법을 통해 pen-up/pen-down 정보의 구분이 없는 3차원 필기 숫자에 대해 $91.6\%$의 인식 성능을 얻었으며, 특히 인식 모델의 개선을 통해 여러획으로 이루어진 글자의 경우 높은 인식 성능의 향상을 보임을 확인하였다.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.9
• /
• pp.166-177
• /
• 1994

In this paper, an implemenation of on-line continuous recognizer for cursive Hangul handwriting is explained. For the Hangul recognition system, we propose a high speed string matching. The editing process in our proposed string matching is accomplished by single editing path. And the matching results are stored in a heap structure and we decide the user comfortibility of unceasing writing during recognition owing to the high speed matching. In the experimental result, a recongition rate of 86.36% at 1.75 second/character over 21,076 characters collected from 50 persons are abtained. And it is shown that the proposed recognition system is operated properly for the on-line recognition for cursive handwring between graphemes.

• Communications for Statistical Applications and Methods
• /
• v.28 no.6
• /
• pp.583-594
• /
• 2021

In this paper, we propose wavelet-based procedures to identify the difference between images, including portraits and handwriting. The proposed methods are based on a novel combination of multiscale methods with a regularization technique. The multiscale method extracts the local characteristics of an image, and the distinct features are obtained through the regularized regression of the local characteristics. The regularized regression approach copes with the high-dimensional problem to build the relation between the local characteristics. Lytle and Yang (2006) introduced the detection method of forged handwriting via wavelets and summary statistics. We expand the scope of their method to the general image and significantly improve the results. We demonstrate the promising empirical evidence of the proposed method through various experiments.

• The Journal of Korean Academy of Sensory Integration
• /
• v.16 no.1
• /
• pp.14-24
• /
• 2018

Objective : The purpose of this study was to identify the effect of Interactive Metronome (IM) training on postural control and hand writing performance of children with Attention Deficit Hyperactivity Disorder (ADHD). Methods : Participant was a third grade elementary school student diagnosed with ADHD. ABA design was used and a total of 30 sessions were held for 3 sessions every week for a total of 10 weeks. In the intervention period, IM training was conducted for 40~50 minutes before intervention for writing, and the writing task was carried out. We evaluated the handwriting legibility and speed. Before baseline A and within a month after A' phase, Clinical Observation of Motor and Postural Skills (COMPS) was evaluated to examine the changes in postural control of the student. Results : After the IM intervention, the postural control of the student improved in the score of slow movement, finger-nose touching, and asymmetrical tonic neck reflex. The handwriting legibility and speed has also tended to increase during the intervention period, but it has not significantly changed. Conclusion : This study could be used as an evidence that the IM training aimed at postural control and handwriting ability could enhance the ability to improve postural control and thereby provide fundamental knowledge for future studies.

• Journal of Sensor Science and Technology
• /
• v.12 no.3
• /
• pp.128-138
• /
• 2003

In this paper, we proposed the algorithm for locating destination address block (DAB) from automatic system to classify packages. For locating DAB, because the size of obtained images is are very large, we select the region of interesting (ROI) to reduce time carrying into algorithm. After selecting the ROI, proposed algorithm is carried out within the ROI. We extract the outline of the handwriting part of the DAB and the rest components within the obtained ROI using thresholding. We carry out labeling to extract each connected component for extracted outline and the rest components. We extract the outline of the handwriting part of the DAB using the geometrical characteristic of the outline of the handwriting part of the DAB among many connected components. The last, we extract the locating DAB using the outline of the handwriting part of the DAB.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10b
• /
• pp.452-454
• /
• 1999

온라인 필기 문자의 경우에는 필기의 변형이 심하고 문자간의 분리가 힘들기 때문에 인식률이 낮은 실정이다. 본 논문에서는 영문 대문자의 자유로운 필기를 인식할 수 있는 방법으로 영문 대문자의 필기시에 발생하는 획간 조합의 특징을 사용하여 인식하는 알고리즘을 제안한다.

• Journal of KIISE:Software and Applications
• /
• v.30 no.5_6
• /
• pp.430-443
• /
• 2003

The objective of this paper is to evaluate most useful feature vector space using the angle, velocity and location features from gesture trajectory which extracted hand regions from consecutive input images and track them by connecting their positions. For this purpose, the gesture tracking algorithm using color and motion information is developed. The recognition module is a HMM model to adaptive time various data. The proposed algorithm was applied to a database containing 4,800 alphabetical handwriting gestures of 20 persons who was asked to draw his/her handwriting gestures five times for each of the 48 characters.