• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.11a
• /
• pp.923-926
• /
• 2004

본 논문에서는 현재까지 도입된 세 가지 종류의 차량번호판의 구조적 특징을 파악하여 이를 구분하고, 반복레이블링기법을 이용하여 각각의 번호판에서 일련번호를 영역화하는 기법을 제안한다. 또한 차량번호판이 가지는 구조적인 특징을 이용하여 용도기호, 차종기초 및 지역명을 영역화하는 기법을 제안한다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2012.10a
• /
• pp.78-81
• /
• 2012

In this paper, we proposed the new vehicle license plates detection method which is available in a various fields, including vehicle access control, illegal parking and speeding vehicle crack down. First, we binarize an image by using difference of gaussian filter to find a sequence of numbers of plates. Second, we determine the plate region by labeling repeatedly using the morphological characteristics of the plates. Finally, we use a projective transformation for correcting the distortion that occurs because of the camera or the location of the vehicle.

• Journal of KIISE
• /
• v.41 no.9
• /
• pp.699-706
• /
• 2014

As XML documents are getting bigger and more complex, a keyword-based search method that does not require structural information is needed to search these large XML documents. In order to use this method, not only all keywords expressed as nodes in the XML document must be labeled for indexing but also structural information should be well represented. However, the existing labeling methods either have very simple information of XML documents for index or represent the structural information which is difficult to deal with the increase of XML documents' size. As the size of XML documents is getting larger, it causes either the poor performance of keyword search or the exponential increase of space usage. In this paper, we present the Repetitive Prime Labeling Scheme (RPLS) in order to improve the problem of the existing labeling methods for keyword-based search of large XML documents. This method is based on the existing prime number labeling method and allows a parent's prime number to be used at a lower level repeatedly so that the number of prime numbers being generated can be reduced. Then, we show an experimental result of the comparison between our methods and the existing methods.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.21 no.3
• /
• pp.419-432
• /
• 2019

Most of deep learning model training was proceeded by supervised learning, which is to train labeling data composed by inputs and corresponding outputs. Labeling data was directly generated manually, so labeling accuracy of data is relatively high. However, it requires heavy efforts in securing data because of cost and time. Additionally, the main goal of supervised learning is to improve detection performance for 'True Positive' data but not to reduce occurrence of 'False Positive' data. In this paper, the occurrence of unpredictable 'False Positive' appears by trained modes with labeling data and 'True Positive' data in monitoring of deep learning-based CCTV accident detection system, which is under operation at a tunnel monitoring center. Those types of 'False Positive' to 'fire' or 'person' objects were frequently taking place for lights of working vehicle, reflecting sunlight at tunnel entrance, long black feature which occurs to the part of lane or car, etc. To solve this problem, a deep learning model was developed by simultaneously training the 'False Positive' data generated in the field and the labeling data. As a result, in comparison with the model that was trained only by the existing labeling data, the re-inference performance with respect to the labeling data was improved. In addition, re-inference of the 'False Positive' data shows that the number of 'False Positive' for the persons were more reduced in case of training model including many 'False Positive' data. By training of the 'False Positive' data, the capability of field application of the deep learning model was improved automatically.

• The Journal of the Korea Contents Association
• /
• v.21 no.6
• /
• pp.701-715
• /
• 2021

The drastic advance of recent deep learning technologies is heavily dependent on training datasets which are essential to train models by themselves with less human efforts. In comparison with the work to design deep learning models, preparing datasets is a long haul; at the moment, in the domain of vision intelligent, datasets are still being made by handwork requiring a lot of time and efforts, where workers need to directly make labels on each image usually with GUI-based labeling tools. In this paper, we overview the current status of vision datasets focusing on what datasets are being shared and how they are prepared with various labeling tools. Particularly, in order to relieve the repetitive and tiring labeling work, we present an interactive smart image annotating system with which the annotation work can be transformed from the direct human-only manual labeling to a correction-after-checking by means of a support of automatic labeling. In an experiment, we show that automatic labeling can greatly improve the productivity of datasets especially reducing time and efforts to specify regions of objects found in images. Finally, we discuss critical issues that we faced in the experiment to our annotation system and describe future work to raise the productivity of image datasets creation for accelerating AI technology.

• 한국어정보학회:학술대회논문집
• /
• 2016.10a
• /
• pp.276-279
• /
• 2016

의미역 결정에서 하나의 의미 논항이 둘 이상의 의미역을 가지는 경우는 복수의 레이블을 할당하기 때문에 어려운 문제이다. 본 논문은 복수의 의미역을 가지는 항의 의미역 결정을 위한 새로운 자질을 제안한다. 복수의 의미역을 결정하기 위해서 체언보다 선행되어 나타나는 용언에 대한 자질을 추가하였다. 또한 문장의 용언에 따라 의미역을 결정하기 위해서 문장 내의 용언 수만큼 각각에 용언에 대한 의미역을 결정할 수 있도록 반복적으로 레이블링하는 방법을 제시하였다. 본 논문의 실험 결과로 제안한 방법은 74.90%의 성능(F1)을 보였다.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.276-279
• /
• 2016

의미역 결정에서 하나의 의미 논항이 둘 이상의 의미역을 가지는 경우는 복수의 레이블을 할당하기 때문에 어려운 문제이다. 본 논문은 복수의 의미역을 가지는 항의 의미역 결정을 위한 새로운 자질을 제안한다. 복수의 의미역을 결정하기 위해서 체언보다 선행되어 나타나는 용언에 대한 자질을 추가하였다. 또한 문장의 용언에 따라 의미역을 결정하기 위해서 문장 내의 용언 수만큼 각각에 용언에 대한 의미역을 결정할 수 있도록 반복적으로 레이블링하는 방법을 제시하였다. 본 논문의 실험 결과로 제안한 방법은 74.90%의 성능(F1)을 보였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.990-993
• /
• 2019

인공지능(AI)은 최근 다양한 산업과 사회에서 패러다임을 바꾸고 있지만, 최첨단 AI 가 제조업에서는 즉각적인 성과를 보이지 못 하고 있다. 다시 말해, Industry 4.0 시점에서 기존의 접근 방법과 차별화되는 실용적인 방법론이 필요하다. 여기서 중요한 점은 '어떤' 데이터를 '어떻게' 활용하여 '어느' 부분에 적용할 것 인가이다. 제조업은 게임과 같이 가상의 캐릭터가 하나의 객체 단위로 구동되는 것이 아니라 수많은 하드웨어가 물리적으로 조합되어 연동한다. 따라서, 현실 세계에서는 물리적 마모, 고장 등으로 인해 엔지니어의 개입 없이 수천만 번 이상의 반복 학습이 불가능하다. 또, 제조업은 학습을 위한 방대한 양의 데이터를 수집하고 레이블링 하는 것이 매우 어렵다. 이 두 가지 한계를 극복할 수 있는 방법은 현실과 매우 유사한 환경을 시뮬레이션으로 재연 후 강화 학습을 사용하는 것이다. 제조 분야에서 아주 복잡한 환경 중 하나로 이송 설비가 있으며, 본 논문에서는 그리드 분류 시스템을 개발하고 강화 학습을 적용시킬 수 있는 환경을 설계한다.

• Journal of KIISE:Software and Applications
• /
• v.32 no.7
• /
• pp.625-635
• /
• 2005

In this paper, we propose a hybrid approach for segmenting the lungs efficiently and automatically in chest CT images. The proposed method consists of the following three steps. first, lungs and airways are extracted by two- and three-dimensional automatic seeded region growing and connected component labeling in low-resolution. Second, trachea and large airways are delineated from the lungs by two-dimensional morphological operations, and the left and right lungs are identified by connected component labeling in low-resolution. Third, smooth and accurate lung region borders are obtained by refinement based on image subtraction. In experiments, we evaluate our method in aspects of accuracy and efficiency using 10 chest CT images obtained from 5 patients. To evaluate the accuracy, we Present results comparing our automatic method to manually traced borders from radiologists. Experimental results show that proposed method which use connected component labeling in low-resolution reduce processing time by 31.4 seconds and maximum memory usage by 196.75 MB on average. Our method extracts lung surfaces efficiently and automatically without additional processing like hole-filling.