• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.36S no.7
• /
• pp.134-135
• /
• 1999

This paper presents a speech enhancement method with spectral subtraction using wavelet, wavelet packet and cosine packet transforms which are known as multiresolutional signal analysis method. The performance of each method is compared with the conventional spectral subtraction method. Performance assessments based on average SNR, cepstral distance and informal subjective listening test are carried out. Experimental result demonstrate that cosine packet shows the best result in objective performance measure as well as subjective shows less musical noise than the conventional spectral subtraction method after removing the noise components.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2004.05b
• /
• pp.64-67
• /
• 2004

Points of sharp variations in images are the most important components when we analyze singularities of images. Therefore a lot of researches for detecting those edges have been continuing even now. However, existing methods do not have excellent performance in the image which exists noise and can not detect edge selectively. In the meantime, the wavelet transform which is presented as a new technique of signal processing field is able to detect multiscale edge and is being applied widely in many fields that analyze singularities such as edge. for this reason, this paper detected image's line-edge elements with 2-D wavelet function, which is independent of line's width, in noise environment.

• Journal of the Institute of Convergence Signal Processing
• /
• v.6 no.1
• /
• pp.33-38
• /
• 2005

Degradation is generated by several causes in the process of digitalization or transmission of data. And its essential cause is noise. Therefore, researches for wavelet-based methods which reconstruct signal degraded by noise have continued. In AWGN(addtive white gaussian noise) environment, the general trend for denoising is to use the thresholding method. Reconstructed signal includes a lot of noise because these methods only consider statistical characteristic regarding noise. In this paper, we present a new method which uses the cumulation of wavelet detail coefficients. As a result, reconstruction of edges and denoising performance are improved. Also we compare existing methods using SNR(signal-to-noise ratio) as the standard of judgement of improvemental effect.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.1
• /
• pp.150-156
• /
• 2007

Noise is generated by several causes, when signal is processed. Hence, it generates error in the process of data transmission and decreases recognition ratio of image and speech data. Therefore, after eliminating those noises, a variety of methods for reconstructing the signal have been researched. Recently, wavelet transform which has time-frequency localization and is possible for multiresolution analysis is applied to many fields of technology. Then threshold-and correlation-based methods are proposed for removing noise. But, conventional methods accept a lot of noise as an edge and are impossible to remove the additive white Gaussian noise (AWGN) and the impulse noise at the same time. Therefore, in this paper we proposed new wavelet-based algorithm for reconstructing degraded signal by noise and compared it with conventional methods.

• Journal of Broadcast Engineering
• /
• v.11 no.2 s.31
• /
• pp.164-173
• /
• 2006

The data structure of the concentric mosaic can be regarded as a video sequence with a slowly panning camera. We take a concentric mosaic with match or alignment of video sequences. Also the concentric mosaic required for huge memory. Thus, compressing is essential in order to use the concentric mosaic. Therefore we need the algorithm that compressed data structure was maintained and the scene was decoded. In this paper, we used 3D lifting transform to compress concentric mosaic. Lifting transform has a merit of wavelet transform and reduces computation quantities and memory. Because each frame has high correlation, the complexity which a scene is detected form 3D transformed bitstream is increased. Thus, in order to have higher performance and decrease the complexity of detecting of a scene we executed 3D lifting and then transformed data set was sequently compressed with each frame unit. Each frame has a flexible bit rate. Also, we proposed the algorithm that compressed data structure was maintained and the scene was decoded by using property of lifting structure.

• Journal of KIISE:Software and Applications
• /
• v.28 no.2
• /
• pp.91-103
• /
• 2001

내용-기반 문서 검색의 필요성이 급속히 증가하고 있다. 기존의 OCR-기반 텍스트 변환 방법은 명백한 한계를 갖고 있기 때문에 영상-기반 매칭 방법이 대안으로서 인기를 얻고 있다. 새로운 매칭방법은 빠른 속도와 좋은 검색 성능의 두 가지 요구사항을 충족해야 한다. 이 논문은 웨이브렛의 좋은 특성을 기반으로 개발된 한글 단어에 대한 영상-기반 매칭 알고리즘을 제안한다. 실험은 고품질과 저품질 단어 영상을 가지고 수행하였으며, 실험 결과 제안한 알고리즘이 검색 성능과 속도 면에서 우수함을 확인하였다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.3
• /
• pp.221-229
• /
• 2004

The Continuous Wavelets Transform project signal f(t) to "Time-scale"plan utilizing the time varied function which called "wavelets". This Transformation permit to analyze scale time dependence of signal f(t) thus the local or global scale properties can be extracted. Moreover, the signal f(t) can be reconstructed stably by utilizing the Inverse Continuous Wavelets Transform. In this paper, the EEG signal is analyzed by wavelets coherence method and the De-noising procedure is represented.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.55-58
• /
• 2001

기존에 MPEG-1, 2나 H.261/263 등에 사용되고있는 움직임 예측 방법은 블록 기반의 코딩 방식으로 단순히 영상의 움직임을 선형 모델로 간주하고 영상을 일정한 크기의 사각형으로 나누어, 각 사각형이 어느 곳으로 이동하는지를 추정하는 방식이었다. 그러나 이러한 방식은 블록 당 하나의 움직임 벡터만으로 처리하기 때문에 블록 내의 복잡한 움직임은 추정할 수 없을 뿐만 아니라, 보다 일반적인 움직임인 회전, 뒤틀림, 확대, 축소 등을 추정할 수 없다. 또한, 영상의 내용에 상관없이 일정한 크기의 블록으로 나누어 처리하기 때문에 주어진 영상에 최적화된 움직임 추정을 수행하기가 어렵다. 특히 저속 비트율에서는 이러한 점들이 크게 부각된다. 이러한 점들을 극복하기 위해서 여기에서는 삼각형 메쉬를 이용한 공간 변환 방법을 이용하였다. 여기에 영상을 영상에 따라 적응적으로 분해하여 처리할 수 있는 웨이브렛 패킷 부호화를 사용하여 에너지가 많고 적음에 따라 초기 제어점의 격자를 조절하여 좀 더 우수한 성능을 얻을 수 있다.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.11b
• /
• pp.617-622
• /
• 2002

This paper presents that the Wavelet Transform can be used to detect the various local defects in a gearbos. Two types of defects which are broken tooth and localized wear, are experimented and the signals are collected by accerometer and analyzed. Because of the complecity of the signals acquired from sensor, it is needed to identify the interesting signal. The natural frequencies of shafts and the gear mesh frequency(GMF) is calculated theretically. DWT, CWT and the aplication are used to extract a gear-localized defect feature from the vibration signal of the gearbox with the defective gear. The results shows the transform is more effective to detect the failures than the Fourier Transform.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2000.11a
• /
• pp.148-152
• /
• 2000

Detection of tool wear is very important in automated manufacturing. This paper presents tool condition monitoring system based on the wavelet analysis of the AC servo motro current in drilling and milling process. The current measurement system is relatively simple and its mounting will not affect machining operations. The discrete wavelet transform was used to decompose the current signal of a spindle AC servo motor in time - frequency domain. The feature vectors were extracted from the decomposed signals and compared for normal and wear condition. The results show the possibility for the effective application of wavelet analysis to tool condition monitoring.