• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2001.11b
• /
• pp.181-185
• /
• 2001

본 논문에서는 오디오 워터마킹 기술을 방송 동기화 시스템에 적용하여 방송시스템을 자동화하기 위한 시스템을 구현했다. 방송시스템은 대개의 경우 1개의 중앙방송과 다수의 지역방송국으로 구성된다. 방송동기화 시스템은 방송소재가 중앙방송에서 지방방송으로 전환하는 시점을 알리는 신호에 워터마크를 삽입하여 방송하고 다수의 지방 방송에서는 워터마크를 검출해서 방송소재의 전환을 자동으로 하고자 한다. 본 논문에서 구현한 방송동기화 시스템은 방송시스템의 특성을 중시하여 설계 제작되었다. 방송시스템의 오류는 바로 방송사고로 이어지는 만큼 동기신호를 잘못 검지하는 오류는 없어야 한다. 즉 워터마크가 없는 오디오에서 워터마크가 있다고 검출되는 경우가 0% 이어야 한다. 본 논문에서 제안된 알고리즘으로 실제 방송되는 라디오 신호를 갖고 실험해본 결과 위의 조건을 만족하고 다양한 환경 즉 압축 및 A/D, D/A, 노이즈 삽입에 대한 신뢰성 테스트를 수행하여 우수한 결과를 얻었다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.72-75
• /
• 2022

야외 환경을 카메라로 촬영한 일반 영상에서 텍스트 이미지를 검출하고 인식하는 기술은 로봇 비전, 시각 보조 등의 기반이 되는 기술로 활용될 수 있어 매우 중요한 기술이다. 하지만 저해상도의 텍스트 이미지의 경우 텍스트 이미지에 포함된 노이즈나블러 등이 더 두드러지기 때문에 텍스트 내용을 인식하는 것이 어렵다. 이에 본 논문은 일반 영상에서의 저해상도 한글 및 영어 텍스트에 대한 이미지 초해상화를 통해 텍스트 인식 정확도를 개선하였다. 트랜스포머에 기반한 모델로 한글 및 영어 텍스트에 대한 이미지 초해상화를 수행하였으며, 영어 및 한글 데이터셋에 대해 제안한 초해상화 방법을 적용했을 때 그렇지 않을 때보다 텍스트 인식 성능이 개선되는 것을 확인하였다.

• Journal of Broadcast Engineering
• /
• v.18 no.4
• /
• pp.643-646
• /
• 2013

We present an improved AdaBoost algorithm to avoid overfitting phenomenon. AdaBoost is widely known as one of the best solutions for object detection. However, AdaBoost tends to be overfitting when a training dataset has noisy samples. To avoid the overfitting phenomenon of AdaBoost, the proposed method divides positive samples into K clusters using k-means algorithm, and then uses only one cluster to minimize the training error at each iteration of weak learning. Through this, excessive partitions of samples are prevented. Also, noisy samples are excluded for the training of weak learners so that the overfitting phenomenon is effectively reduced. In our experiment, the proposed method shows better classification and generalization ability than conventional boosting algorithms with various real world datasets.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.14 no.6
• /
• pp.35-42
• /
• 2000

In this paper, AE(Acoustic Emission) signals detected from the growth of the electrical tree in an epoxy resin under ac high-voltage application were analysed to diagnose the insulation deterioration of cast-resin power transformer. Frequency spectra of AE signals generated from the magnetizing and the load currents in the actual operating cast-resin power transformer of 500[kVA] under distribution system of 22.9[kV] were also analysed to distinguish the AE signals due to void discharges from the magnetic circuit noises in the core of the transformer. As the experimental results, we could distinguish the AE signals whether those signals were caused due to the void discharges or due to the magnetic circuit noises by analyzing the frequency spectrum of AE signals. The frequency spectra of AE signals generated from the cast-resin power transformer in operation due to both the magnetizing and the load currents appeared in the range of 40-120[kHz], but the frequency band of AE signals emitted from the void discharges in an epoxy resin sample was about 50[kHz] to 230[kHz].

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2003.10a
• /
• pp.115-115
• /
• 2003

최근 회전기계에 대한 세계적인 기술추세를 볼 때 해를 거듭할수록 더욱 정밀해지고 고속화에 대한 요구가 한층 증대되고 있으며 이러한 측면에서 여러 분야에 우수한 장점을 지니고 있는 능동적 자기베어링의 회전축계 활용에 대한 연구가 활발히 진행되고 있다. 특히 프랑스를 중심으로 한 미국, 일본 등이 이에 대한 연구를 활발하게 수행하여 고속 공작기계 스핀들을 비롯한 고진공 펌프 및 각종 터어빈, 압축기 등에 실용화시키는 단계에 있다. 그러나 국내 관련기술에 대한 연구는 연구소, 학계를 중심으로 실험실적인 기초연구로서 부분적으로 수행하는 단계에 있으며 관련분야 활용을 위한 본격적인 연구는 수행되지 못하고 있는 실정이고 이를 실용화시키기 위한 일환으로 능동적 자기 베어링의 회전 축계를 구성하고 있는 요소기술 중 하나인 전력 증폭기의 개발이 이뤄져야 할 필요성이 있는데 본 논문에서는 네트웍 기반 전력 증폭기 개발을 시도하였다. 전력 증폭기는 크게 리니어 앰프와 스위칭 앰프로 구분된다. 리니어 앰프의 경우 회로가 간단하고 노이즈가 비교적 작다는 장점이 있지만 전력손실 및 발열이 크기 때문에 에너지 측면에서 저 효율이라는 점과 따로 방열판을 부착해야 한다는 단점을 가지고 있고, 스위칭 앰프의 경우 전력손실이 작은 반면, 회로가 비교적 복잡하고 노이즈의 발생 가능성이 높다는 단점이 있다. 본 논문에선 위 두 가지 방식을 혼합한 혼합형 전력 증폭기로 설계하였다. 또한 기존에 전력증폭기의 경우 상위 주제어기로부터 제어량을 아날로그 신호로 통신하기 때문에 발생할 수 있는 EMI 노이즈신호에 대한 대책을 세워야 하는데 본 연구를 통해 개발된 전력증폭기는 따로 보조제어기(TMS320LF2406A)를 두어 상위 주제어기에 전력증폭기의 상태값을 궤환할 수 있도록 명령 신호체계를 전체 시스템의 샘플링 시간을 고려하여 비교적 전송 속도가 빠른 CAN(Controller Area Network)으로 구축하여 주제어기와 전력 증폭기간에 양방향 통신이 가능하도록 하였다. 이로써 전력증폭기의 상태정보를 알 수 있다. 따라서 본 논문에서는 칩 설계기술의 발전으로 가격대 성능비가 우수한 DSP(TMS320LF2406A)를 이용하여 과거의 아날로그 방식의 명령신호체계를 디지털 신호체계로 바꿈으로써 네트웍을 통해 전력증폭기의 상태진단 가능성을 검증한다.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.28 no.5
• /
• pp.391-399
• /
• 2017

At present the general trend of modern electronics is toward smaller packages and high performance. As an antenna requires high powers, the EMC(Electromagnetic Compatibility) problems of the transmission line stage is becoming crucial day by day. In this paper, a transmission line excited by the electromagnetic fields from an infinitesimal electric dipole antenna is analyzed using the modified telegrapher's equations. The analytical equations are derived for arbitrarily positioned electric dipole with reference to a transmission line. To verify our approach, the induced voltage and current at the terminal were computed by the proposed approach and compared with those obtained by the electromagnetic simulation solver. Furthermore, the induced currents at the terminal of a transmission line excited by the electric dipoles at various positions were investigated using our approach.

• Journal of Broadcast Engineering
• /
• v.18 no.4
• /
• pp.543-549
• /
• 2013

Motion artifacts of central and autonomic nervous system signals degrades the performance of the bio-signal based human factor analysis. Firstly, we propose a defining method of motion artifact section by analyzing successive image frames. Motion artifact section is defined when the amount of motion is greater than the pre-defined threshold. In here, the amount of motion is estimated by first derivation of image frames at temporal domain. Secondly, we propose another defining method of motion artifact section through designing 2D Gaussian probability density function model by analyzing feature vectors of one cycle of signal such as length and amplitude. The defined motion artifact sections are interpolated on the basis of 1D Gaussian function. At result of applying the method into photoplethysmography signal, we confirmed that the calculated heartbeat rate from the restored photoplethysmography came up to the one from electrocardiography. Also, we found that the video based method generated relatively more false acceptance of motion artifact section and the probability density function based method generated relatively more false rejection of motion artifact section.

• Journal of Korean Society for Geospatial Information Science
• /
• v.13 no.4 s.34
• /
• pp.3-9
• /
• 2005

The extraction of traffic information based on image processing is under broad research recently because the method based on image processing takes less cost and effort than the traditional method based on physical equipment. The main purpose of the algorithm based on image processing is to extract vehicles from an image correctly. Before the extraction, the algorithm needs the pre-processing such as background subtraction and binary image thresholding. During the pre-processing much noise is brought about because roadside tree and passengers in the sidewalk as well as vehicles are extracted as traffic flow. The noise undermines the overall accuracy of the algorithm. In this research, most of the noise could be removed by extracting the exact road area which does not include sidewalk or roadside tree. To extract the exact road area, traffic lanes in the image were used. Algorithm speed also increased. In addition, with the ratio between the sequential images, the problem caused by vehicles' shadow was minimized.

• Journal of Information Technology and Architecture
• /
• v.9 no.3
• /
• pp.323-331
• /
• 2012

With the advancement of the area of knowledge and information, Intellectual Property, especially, patents have captured attention more and more emergent. The increasing need for efficient way of patent information search has been essential, but the prevailing patent search engines have included too many noises for the results due to the Boolean models. This has occasioned too much time for the professional experts to investigate the results manually. In this paper, we reveal the differences between the conventional document search and patent search and analyze the limitations of existing patent search. Furthermore, we propose a specialized in patent search, so that the relationship between the keywords within each document and their significance within each patent document search keyword can be identified. Which in turn, the keywords and the relationships have been appointed a ranking for this patent in the upper ranks and the noise in the data sub-ranked. Therefore this approach is proposed to significantly reduce noise ratio of the data from the search results. Finally, in, we demonstrate the superiority of the proposed methodology by comparing the Kipris dataset.

• Journal of KIISE:Software and Applications
• /
• v.34 no.7
• /
• pp.667-680
• /
• 2007

The bio-data processing is used for a suitable purpose with bio-signals, which are obtained from human individuals. Recently, there is increasing demand that the bio-data has been widely applied to various applications. However, it is often that the number of data within each class is limited and the number of classes is large due to the property of problem domain. Therefore, the conventional pattern recognition systems and classification methods are suffering form low generalization performance because the system using the lack of data is influenced by noises of that. To solve this problem, we propose a modified additive factor model for bio-data generation, with two factors; the class factor which affects properties of each individuals and the environment factor such as noises which affects all classes. We then develop a classification system through defining a new similarity function using the proposed model. The proposed method maximizes to use an information of the class classification. So, we can expect to obtain good generalization performances with robust noises from small number of datas for bio-data. Experimental results show that proposed method outperforms significantly conventional method with real bio-data.