• Proceedings of the KSRS Conference
• /
• 2008.03a
• /
• pp.219-223
• /
• 2008

초분광 영상은 기존의 다중분광 영상보다 많은 파장대의 영상을 취득하기 때문에 다양한 분야의 연구에 이용되고 있다. 하지만 밴드별로 취득하는 파장대가 짧고 밴드수가 많아, 밴드간의 높은 상관관계 및 노이즈 밴드가 존재한다. 이로 인해 기존에 알려진 분석기법의 적용결과가 제대로 도출되지 않는다. 따라서 초분광 영상을 이용할 경우, 노이즈가 많이 포함된 밴드를 제거한 후 영상분석을 하는 것이 보다 효율적이다. 본 연구에서는 초분광 영상(Hyperspectral Image)의 전처리 과정 중 노이즈 밴드 제거에 초점을 맞추었으며, 이를 위해 프랙탈 차원을 이용하였다. 프랙탈 차원 측정방법 중 삼각기둥 표면적 기법을 이용하였다. 프랙탈 차원을 측정하고, Continuum Removal 기법을 이용하여 경향을 살펴보았다. 경험적으로 구한 임계값을 통해 상대적으로 정보량이 적은 밴드를 노이즈 밴드로 판단하여 제거하였다. 실험 영상으로는 EO-1 위성에서 취득되는 Hyperion 초분광 영상을 사용하였다. 실험 결과 프랙탈 분석을 통해 Hyperion 초분광 영상의 노이즈 밴드를 자동으로 추출하여 제거할 수 있음을 확인하였다.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.23 no.5
• /
• pp.455-463
• /
• 2003

Adaptive noise cancellation techniques are ideally suitable for reducing spatially varying noise due to the grain structure of material in ultrasonic nondestructive evaluation. Grain noises have an un-correlation property, while flaw echoes are correlated. Thus, adaptive filtering algorithms use the correlation properties of signals to enhance the signal-to-noise ratio (SNR) of the output signal. In this paper, a multi-stage adaptive noise cancellation (MANC) method using adaptive least mean square error (LMSE) filter for enhancing flaw detection in ultrasonic signals is proposed.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.11a
• /
• pp.179-181
• /
• 2009

In this paper, we propose an adaptive noise detection and modified Gaussian removal algorithm using local statistics for impulse noise. In order to determine constraints for noise detection, the local mean, variance, and maximum values are used. In addition, a modified Gaussian filter that integrates the tuning parameter to remove the detected noises. Experimental results show that our method is significantly better than a number of existing techniques in terms of image restoration and noise detection.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.15 no.1
• /
• pp.44-50
• /
• 2001

The largest problem in applying the elecbical partial discharge measurement method the transformer that has been operated until now is the removal of external corona noise In this thesis, a methcd was studied. to rneasme only fue partial discharge sIgnal due to the defoct in transfonrer except the external corona noise. To find out the types of partial discharge and corona noise within a transfomr, a partial discharge was made in use of a needle-plane electrodes within a model transfonner and, at the same time, an external corona noise was generated in use of a rod-sphere electrcdes in the air around the transformer. Both of a partial clischarge signal caused from an intemat defect within a transformer and an external noise were found at the rogowski coil which was located at transformer earth wire. When the external corona noise, which was separately measured in use of an antenna sensor out of transfonner, was removed from the signal measured on rogowski coil, the signal caused by partial discharge within a transformer would effectively be acquired.quired.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.10a
• /
• pp.1119-1121
• /
• 2015

본 논문에서는 트위터 내용을 통해 이벤트를 탐지하는 시스템에서 지명 검출 정확도를 개선하는 방법을 제안한다. SNS를 이용한 개인 정보 유출 사례들이 늘어감에 따라 자신의 위치 정보를 공개하기 꺼려하기 때문에 이벤트가 발생한 지역을 검출하기 위해서는 텍스트 내용을 직접 분석해야한다. 그러나 오타나 줄임말, 동형이의어의 사용으로 정확한 지명 검출에 어려움이 발생하였다. 따라서 정확도를 향상시키기 위해 본 논문에서는 두 가지 지명 검출 기법을 제안한다. 지명 단어에서 발생되는 노이즈를 제거하는 지명 노이즈 제거 기법과 랜드 마크를 이용하여 지명 단어를 확정하는 지명 확정 기법이다. 실험 결과 기존 시스템의 정확도 49%에서 지명 노이즈 제거기법은 56%, 지명 확정 기법은 73%로 각각 향상되었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.10a
• /
• pp.1095-1097
• /
• 2015

본 논문에서는 트위터 기반 이벤트 탐지에서의 기계학습을 통한 지명 노이즈제거 방식을 제안한다. 이벤트 탐지 시스템은 트위터 사용자 개개인을 이벤트 탐지의 센서로 이용하여 특정 지명에서 발생하는 이벤트를 탐지하였다. 그러나 지명과 동형이의어 관계의 단어가 탐지되어 이벤트 탐지의 정확도를 낮추는 요인이 된다. 이에 본 논문에서는 먼저 노이즈 관련 데이터베이스 구축을 이용하여 제거 필터링을 진행한 후에 기계학습을 이용해서 지명 유무를 결정하였다. 실험결과 본 논문에서 제시하는 예측기법은 89.6%의 신뢰도로 노이즈제거 기법의 필요성을 보였다.

• Journal of the Korea Computer Graphics Society
• /
• v.27 no.5
• /
• pp.45-54
• /
• 2021

Recently, deep learning-based denoising approaches have been actively studied. In particular, with the advances of blind denoising techniques, it become possible to train a deep learning-based denoising model only with noisy images in an image domain where it is impossible to obtain a clean image. We no longer require pairs of a clean image and a noisy image to obtain a restored clean image from the observation. However, it is difficult to recover the target using a deep learning-based denoising model trained by only noisy images if the distribution of the noisy image is far from the distribution of the clean image. To address this limitation, unpaired image denoising approaches have recently been studied that can learn the denoising model from unpaired data of the noisy image and the clean image. ISCL showed comparable performance close to that of supervised learning-based models based on pairs of clean and noisy images. In this study, we propose suitable normalization techniques for each purpose of architectures (e.g., generator, discriminator, and extractor) of ISCL. We demonstrate that the proposed method outperforms state-of-the-art unpaired image denoising approaches including ISCL.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.605-608
• /
• 2007

The denoising for image restoration based on the Wavelet Packet with absolute average threshold is presented. The Existing method is used standard deviation estimated results in increasing the noise and threshold, and damaging an image quality. In addition that is decreased image restoration PSNR by using the same threshold in spite of changing image because of installing a threshold in proportion of noise size. In contrast, the absolute average threshold with wavelet packet is adapted by changing image to set up threshold by statistic quantity of resolved image and is avoided an extreme impart. The results on the experiment has improved 10% and 5% over than the denoising based on simple wavelet transform and wavelet packet respectively.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.6
• /
• pp.1122-1128
• /
• 2007

The denoising for image restoration based on the Wavelet Packet Transform with AA(Absolute Average) making-threshold is presented. The wavelet packet transform leads to be better in the part of high frequency than wavelet transform to eliminate noise. And the existing threshold determination is used standard deviation estimated results in increasing the noise and threshold, and damaging an image quality. In addition that is decreased image restoration PSNR by using the same threshold in spite of changing image because of installing a threshold in proportion of noise size. In contrast the AA thresholding method with wavelet packet is adapted by changing image to set up threshold by statistic quantity of resolved image and is avoided an extreme impact. The results on the experiment has improved 10% and 5% over than the denoising based on simple wavelet transform and wavelet packet respectively.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.2
• /
• pp.183-190
• /
• 2013

In this paper, we propose a spatially adaptive noise detection and removal algorithm for a single degraded image. Under the assumption that an observed image is Gaussian-distributed, the noise information is estimated by local statistics of degraded image, and the degree of the additive noise is detected by the local statistics of the estimated noise. In addition, we describe a noise removal method taking a modified Gaussian filter which is adaptively determined by filter parameters and window size. The experimental results demonstrate the capability of the proposed algorithm.