• Proceedings of the KIEE Conference
• /
• 1995.11a
• /
• pp.192-194
• /
• 1995

An improved filter for tracking a maneuvering target is presented. The proposed filter consists of two kalman filters based on different dynamic models and double decision logic. The use of double decision logic for the maneuver onset and ending detection leads to reduction in estimation error. This decision rule is based on evidence theory, Dempster-Shafer theory, which is extended in order to be applicable in the tracking problem. Simulation results show that the proposed filter performs better than IMM at a lower computational load.

• Proceedings of the KIEE Conference
• /
• 2007.10a
• /
• pp.107-108
• /
• 2007

In this paper, a cosine-modulated filter bank with a modified synthesis part is proposed for arbitrary scaling of images, whereby down/up-sampling kernels based on a compactly supported sampling function (CSSF) are utilized. Also, an optimized adaptive interpolation technique is incorporated into the filter bank structure to compensate for quality degradation arising in scaled images. Finally, simulation results verify that high quality images with arbitrary sizes can be obtained by applying the proposed approach.

• Journal of Power Electronics
• /
• v.16 no.1
• /
• pp.310-318
• /
• 2016

This paper presents a new orthogonal signals generator (OSG) with DC Offset rejection for implementing a phase locked loop (PLL) in single-phase grid-connected power systems. An adaptive filter (AF) based on the least mean square (LMS) algorithm is used to constitute the OSG in this study. The DC offset in the measured grid voltage signal can be significantly rejected in the developed OSG technique. This generates two pure orthogonal signals that are free from the DC offset. As a result, the DC offset rejection performance of the presented single-phase phase locked loop (SPLL) can be enhanced. A mathematical model of the developed OSG and the principle of the adaptive filter based SPLL (AF-SPLL) are presented in detail. Finally, simulation and experimental results demonstrate the feasibility of the proposed AF-SPLL.

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.980-983
• /
• 1996

This paper presents the method of determining state feedback gains of XPTOS for disk drive servomechanism based BESSEL filter prototype. A typical disk drive actuator can be modeled as second order dynamics for low frequencies. However, the response at higher frequencies shows resonant behavior which cannot be easily modeled. XPTOS consists of the nonlinear control region and the linear control region. In the linear control region, the poles of a second order nominal model of plant must be properly relocated by pole placement technique to attenuate resonant modes at high frequency and to attain minimum time state transition. It is difficult to select position to satisfy this object because velocity feedback gain is subjected to position feedback gain in XPTOS. Here poles of BESSEL filter prototype are selected to determine state feedback gains of XPTOS. Simulation results for disk drive servomechanism using XPTOS having state feedback gains by the proposed method are presented.

• Journal of the Optical Society of Korea
• /
• v.12 no.4
• /
• pp.281-287
• /
• 2008

An acousto-optic tunable filter based 3-D micro surface profile measurement using an equally spaced 5 spectral phase shifting is described. The 5-bucket spectral phase shifting method is compared with a Fourier-transform method in the spectral domain. It can provide a fast measurement capability while maintaining high accuracy since it needs only 5 pieces of spectrally phase shifted imaging data and a simple calculation in comparison with the Fourier transform method that requires full wavelength scanning data and relatively complicated computation. The 3-D profile data of micro objects can be obtained in a few seconds with an accuracy of ${\sim}10nm$. The 3-D profile method also has an inherent benefit in terms of being speckle-free in measuring diffuse micro objects by employing an incoherent light source. Those simplicity and practical applicability is expected to have diverse applications in 3-D micro profilometry such as semiconductors and micro-biology.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.6 no.2
• /
• pp.81-87
• /
• 2013

Motion segmentation is a fundamental technique for analysing image sequences of real scenes. A process of identifying moving objects from data is a typical task in many computer vision applications. In this paper, we propose motion segmentation that generally consists from background subtraction and foreground pixel segmentation. The Approximated Median Filter (AMF) was chosen to perform background modeling. Motion segmentation in this paper covers RGB video data.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.7 no.1
• /
• pp.61-67
• /
• 2014

Motion segmentation is a fundamental technique for analysing image sequences of real scenes. A process of identifying moving objects from data is a typical task in many computer vision applications. In this paper, we propose motion segmentation that generally consists from background subtraction and foreground pixel segmentation. The Approximated Median Filter(AMF) was chosen to perform background modeling. Motion segmentation in this paper covers RGB video data.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.04a
• /
• pp.403-414
• /
• 2005

Most of the currently used spam-filters are based on a Bayesian classification technique, where some serious problems occur such as a limited precision/recall rate and the false positive error. This paper addresses a solution to the problems using a modified Bayesian classifier based on chi-square statistics. The resulting spam-filter is more accurate and flexible than traditional Bayesian spam-filters and can be a personalized one providing some parameters when the filter is teamed from training data.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.8B
• /
• pp.1400-1412
• /
• 2000

In this paper, we propose a technique for image enhancement, especially for denoising and deblocking based on wavelets. In this proposed algorithm, frame wavelet system designed as a optimal edge detector was used. And our theory depends on Lipschitz regularity, spatial correlation, and some important assumptions. The performance of the proposed algorithm was compared with three popular test images in image processing area. Experimental results show that the performance of the proposed algorithm was better than other previous denoising techniques like spatial averaging filter, Gaussian filter, median filter, Wiener filter, and some other wavelet based filters in the aspect of both PSNR and human visual system, The experimental results also show approximately the same capability of deblocking as the previous developed techniques

• Journal of Korea Multimedia Society
• /
• v.24 no.2
• /
• pp.178-185
• /
• 2021

Three-dimensional high-resolution magnetic resonance imaging (MRI) provides fine-level anatomical information for disease diagnosis. However, there is a limitation in obtaining high resolution due to the long scan time for wide spatial coverage. Therefore, in order to obtain a clear high-resolution(HR) image in a wide spatial coverage, a super-resolution technology that converts a low-resolution(LR) MRI image into a high-resolution is required. In this paper, we propose a super-resolution technique through filter learning based on information on the surrounding gradient information in 3D space from 3D MRI images. In the learning step, the gradient features of each voxel are computed through eigen-decomposition from 3D patch. Based on these features, we get the learned filters that minimize the difference of intensity between pairs of LR and HR images for similar features. In test step, the gradient feature of the patch is obtained for each voxel, and the filter is applied by selecting a filter corresponding to the feature closest to it. As a result of learning 100 T1 brain MRI images of HCP which is publicly opened, we showed that the performance improved by up to about 11% compared to the traditional interpolation method.