• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.12C
• /
• pp.1660-1668
• /
• 2004

This paper proposes a Voice Activity Detection (VAD) algorithm based on Radial Basis Function (RBF) network using dual threshold. The k-means clustering and Least Mean Square (LMS) algorithm are used to upade the RBF network to the underlying speech condition. The inputs for RBF are the three parameters in a Code Exited Linear Prediction (CELP) coder, which works stably under various background noise levels. Dual hangover threshold applies in BRF-VAD for reducing error, because threshold value has trade off effect in VAD decision. The experimental result show that the proposed VAD algorithm achieves better performance than G.729 Annex B at any noise level.

• Journal of Korea Multimedia Society
• /
• v.13 no.11
• /
• pp.1610-1620
• /
• 2010

This paper proposes a motion-adaptive 3-D deinterlacing algorithm based on an adaptive-thresholded motion detection and an interpolation method using binary patterns to compensate motion missing and false motion errors. For efficient motion detection, we adaptively decided a threshold value according to the complexity of image. Many edge-based interpolation algorithms have been proposed to improve the subjective quality. Recently, to efficiently interpolate low angle edge and line, a method using predefined binary patterns has been proposed. In this paper, we propose an improved method by modifying the binary patterns. Simulation results have shown that the proposed method provides better performance than the existing methods.

• Journal of KIISE:Computing Practices and Letters
• /
• v.7 no.5
• /
• pp.440-450
• /
• 2001

This paper discusses an effective algorithm which is proposed for abrupt scene change detection in MPEG bitstream. The proposed algorithm restores DC images by decoding only DC coefficients and estimates the new motion vectors between adjacent DC images and detects scene change by similarity measure between frames. The proposed algorithm calculates similarity measure between adjacent frames, i.e motion compensated inter-frame correlation, and detects scene change by comparing this similarity measure with threshold value independent of sequences. Experimental results show that the proposed algorithm has more than 90% \`recall\` and \`precision\` in almost sequences and these results can be considered better than other algorithms using threshold value dependent of sequences.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.5A
• /
• pp.490-495
• /
• 2010

In this paper, we propose an sequential spectrum sensing algorithm utilizing DFT. The conventional sensing algorithm using FFT contains redundant computation due to the characteristic of FFT which computes all frequency components at one time. The proposed sensing algorithm utilizing DFT computes a frequency component once at a time according to the priority and decides presence of signal. The proposed sensing algorithm can provide similar detection performance to the conventional scheme while computations of the sensing process could be reduced significantly depends on an early detection of signal.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.4
• /
• pp.473-480
• /
• 2004

This paper presents a new edge detection algorithm in color image. The proposed Adaptive morphological Wavelet-CNN algorithm is divided into two parts : The Adaptive morpholog and WCNN(Wavelet Cellular Neural Networks). It detects the optimal edge with applying this color image to WCNN algorithm, after it does level up a boundary side of a color image by using the adaptive morphology as the threshold of an input color image. Also, it is used not a conventional fixed mask edge detection method but variable mask method which is called a variable BBM. Finally, to show the feasibility of the proposed algorithm, this paper provides by simulation that the color image consists of 30.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.11a
• /
• pp.249-252
• /
• 2003

본 논문은 컬러 영상을 명암도에 따른 공간적 객체 분할인 YIQ 모델을 사용하여 객체 분할한 영상의 임계값에 따른 적응적 형태학을 이용하여 영상의 경계면을 레벨 업시킨 후, 이를 웨이브렛에 적용하여 최적의 에지를 검출하였다. 또한, 흑백 영상보다 더 많은 더 정보를 가진컬러 영상을 사용하여, 기존의 영상 에지 검출 알고리즘인 Sobel 에지 검출과 다른 웨이브렛기저 계수를 적용한 에지 검출 방법과 비교하고, 제안된 알고리즘이 기존의 다른 에지 검출보다 우수함을 확인하였다. 특히 에지와 에지의 부분이 가까울 때 정확한 에지를 검출하였으며, 완만한 곡선을 가지고 있는 부분에서 더 우수한 결과 에지를 얻을 수 있음을 확인하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2014.06a
• /
• pp.139-141
• /
• 2014

지능형 차량 시스템에 있어서 교통 표지판 검출/인식은 매우 중요한 요소들 중의 하나이다. 따라서 주행 중인 차량에서 카메라로부터 취득한 영상을 이용하여 교통 표지판을 인식하는 여러 가지 영상인식 알고리즘들이 개발되고 있다. 하지만 이러한 알고리즘은 표지판의 색상 값이 날씨와 시간에 따른 조도와 컬러의 변화에 따라 성능이 크게 변한다는 점에서 어려움을 겪고 있다. 따라서 본 논문은 환경 변화에 강인한 교통 표지판 검출 및 인식 알고리즘을 제안한다. 구체적으로, 표지판 검출을 위하여 제안하는 알고리즘에서는 색상과 형태 정보를 이용하여 교통 표지판 후보군을 찾는다. 여러 색상 임계값에 대하여 영상 피라미드 형태를 만들고, 모든 피라미드 영상들에 대해서 인식 알고리즘을 수행함으로써 실외 빛에 변화에 강인하게 한다. 교통 표지판 후보군을 찾은 후, 후보군들을 Linear SVM을 통해 학습함으로써 교통 표지판인지 아닌지 분류해낸다. 실험 결과는 제안하는 알고리즘이 정확하게 교통 표지판을 인식하고, 동시에 실외 빛의 변화에 상관없이 강인하게 표지판을 인식함을 보여준다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.42 no.4 s.304
• /
• pp.51-58
• /
• 2005

This paper proposes a robust scene change detection technique that uses the weighted chi-square test and the automated threshold-decision algorithms. The weighted chi-square test can subdivide the difference values of individual color channels by calculating the color intensities according to NTSC standard, and it can detect the scene change by joining the weighted color intensities to the predefined chi-square test which emphasize the comparative color difference values. The automated threshold-decision at algorithm uses the difference values of frame-to-frame that was obtained by the weighted chi-square test. At first, The Average of total difference values is calculated and then, another average value is calculated using the previous average value from the difference values, finally the most appropriate mid-average value is searched and considered the threshold value. Experimental results show that the proposed algorithms are effective and outperform the previous approaches.

• Journal of Advanced Navigation Technology
• /
• v.11 no.4
• /
• pp.485-492
• /
• 2007

In this paper, we propose the receiver block error detection of the video codec and the image concealment algorithm using fuzzy inference. The proposed error detection and concealment algorithm gets SSD(Summation of Squared Difference) and BMC(Boundary Matching Coefficient) using the temporal and spatial similarity between corresponded blocks in the two successive frames. Proportional constant, ${\alpha}$, for threshold value, TH1 and TH2, is decided after fuzzy data is generated by each parameter. To examine the propriety of the proposed algorithm, random errors are inserted into the QCIF Susie standard image, then the error detection and concealment performance is simulated. To evaluate the efficiency of the algorithm, image quality is evaluated by PSNR for the error detection and concealed image by the existing VLC table and by the proposed method. In the experimental results, the error detection algorithm could detect all of the inserted error, the image quality is improved over 15dB after the error concealment compare to existing error detection algorithm.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.5
• /
• pp.19-26
• /
• 2008

In this paper, we converted an RGB into an YCbCr image input from CCD camera and then after compute difference two consecutive images, conduct Glassfire Labeling. We extract an image become ware of motion-change, if the difference between most broad(area) and Area critical value more than critical value. We enforce the detection of facial characteristics to an extracted motion-change images by using AdaBoost algorithm to extract an characteristics.