• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.3
• /
• pp.60-67
• /
• 2009

The phase retrieval problem is a problem of reconstructing a signal or the phase of Fourier transform of the signal from the magnitude of its Fourier transform. In this paper we address the problem of reconstructing an unknown signal from the magnitude of its Fourier transform and the magnitude of Fourier transform of another signal that is given by the addition of the desired signal. After we briefly mention the uniqueness conditions under which a signal can be uniquely specified from the given information and key equations of the iterative algorithm, we present mathematical background that the iterative algorithm converges to the desired signal, present an example that illustrates the performance of the reconstruction algorithm, and show its convergence property.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.2
• /
• pp.342-348
• /
• 2009

This paper introduces a novel denoising algorithm for the partial-discharge(PD) signals from power apparatuses. The developed algorithm includes three kinds of specific denoising sub-algorithms. The first sub-algorithm uses the fuzzy logic which classifies the noise types in the magnitude versus phase PD pattern. This sub-algorithm is especially effective in the rejection of the noise with high and constant magnitude. The second one is the method simply removing the pulses in the phase sections below the threshold count in the count versus phase pattern. This method is effective in removing the occasional high level noise pulses. The last denoising sub-algorithm uses the grouping characteristics of PD pulses in the 3D plot of the magnitude versus phase versus cycle. This special technique can remove the periodical noise pulses with varying magnitudes, which are very difficult to be removed by other denoising methods. Each of the sub-algorithm has different characteristic and shows different quality of the noise rejection. On that account, a parameter which numerically expresses the noise possessing degree of signal, is defined and evaluated. Using the parameter and above three sub-algorithms, an adaptive complex noise rejection algorithm for the on-line PD diagnosis system is developed. Proposed algorithm shows good performances in the various real PD signals measured from the power apparatuses in the Korean plants.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.2
• /
• pp.141-147
• /
• 2010

In this paper, we propose a new speech separation algorithm to extract and enhance the target speech signals from mixed speech signals by utilizing both magnitude and phase information. Since the previous statistical modeling algorithms assume that the log power spectrum values of the mixed speech signals are independent in the temporal and frequency domain, discontinuities occur in the resultant separated speech signals. To reduce the discontinuities, we apply a smoothing filter in the time-frequency domain. To further improve speech separation performance, we propose a statistical model based on both magnitude and phase information of speech signals. Experimental results show that the proposed algorithm improve signal-to-interference ratio (SIR) by 1.5 dB compared with the previous magnitude-only algorithms.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.1E
• /
• pp.84-88
• /
• 1996

In this paper, we propose a novel adaptive digital filter for tonal noise cancellation, with a frequency tracking capability. The proposed algorithm not only estimates the magnitude and phase of the tonal disturbance but also tracks its frequency, which changes in quasi-static manner. The algorithm uses the steepest descent method and the instantaneous frequency approach for the phase/magnitude estimation and frequency tracking, respectively. A number of computer simulations have been carried out in order to demonstrate the feasibility of the proposed ANC algorithm under various conditions.

• Proceedings of the KIEE Conference
• /
• 1997.07a
• /
• pp.198-200
• /
• 1997

This paper consider a simple Nondestructive Testing(NDT) having eddy currnt effect. We analyzed the two dimension modeling of alternative magnetic field. eddy current with voltage source. And, the current magnitude and phase data obtained from each different frequency five object position is used for learning the neural network. Therefore, we can recognize an object position pattern from new input current magnitude, phase data.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.9
• /
• pp.1159-1164
• /
• 2018

Vacuum circuit breakers(VCB) are widely used for current interruption of high-voltage inductive loads such as induction motors. This VCB can be chopped off before the current zero due to its high arc-extinguishing capability. One of the outstanding features of VCB is that it can cut off high frequency re-ignition current more than other circuit breakers. If the transient recovery voltage generated in the arc extinguishing is higher than the dielectric strength of the circuit breaker, a re-ignition phenomenon occurs. The surge voltage of the re-ignition is very high in magnitude and the steepness of the waveform is so severe that it can act as a high electrical stress on the winding. If the high frequency current of one phase affects the other two phases when the re-ignition occurs, it may cause a high surge voltage due to the virtual current chopping. If the magnitude of the voltage allowed in the motor winding is high or the waveform level is too severe, it may lead to insulation breakdown. Therefore, it is necessary to reduce the voltage to within a certain range. In this study, we briefly explain the various phenomena at the time of interruption, analyzed the magnitude of the dielectric strength and the transient recovery voltage at the simultaneous three-phase interruption that can give the greatest influence to the inductive load, proposed a method to reduce the impact.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.27 no.8
• /
• pp.1275-1283
• /
• 1990

Perception of surfaces from range images plays a key role in 3-D object recognition. Recognition of 3-D objects from range images is performed by matching the perceived surface descriptions with stored object models. The first step of the 3-d object recognition from range images is image segmentation. In this paper, an approach for segmenting 3-D range images into symbolic surface descriptions using spatial Gabor filter is proposed. Since the phase of data has a lot of important information, the phase information with magnitude information can effectively segment the range imagery into regions satisfying a common homogeneity criterion. The phase and magnitude of Gabor filter can represent a unique featur vector at a point of range data. As a result, range images are trnasformed into feature vectors in 3-parameter representation. The methods not only to extract meaningful features but also to classify a patch information from range images is presented.

• Phonetics and Speech Sciences
• /
• v.3 no.1
• /
• pp.87-94
• /
• 2011

The decision directed (DD) approach is widely used to determine a priori SNR from noisy speech signals. In conventional speech enhancement systems with a DD approach, a priori SNR is estimated by using only the magnitude components and consequently follows a posteriori SNR with one frame delay. We propose a phase-dependent two-step a priori SNR estimator based on the minimum mean square error (MMSE) in the log-mel spectral domain so that we can consider both magnitude and phase information, and it can overcome the performance degradation caused by one frame delay. From the experimental results, the proposed estimator is shown to improve the output SNR of enhanced speech signals by 2.3 dB compared to the conventional DD approach-based system.

• Journal of Advanced Marine Engineering and Technology
• /
• v.28 no.2
• /
• pp.338-346
• /
• 2004

Experimental data on the effect of the variable gravity magnitude, namely microgravity, normal gravity and hyper-gravity, on flow pattern and frictional pressure drop were obtained during co-current air-water flow in a horizontal tube, The flow patterns were found to depend strongly on the gravity magnitude and certain flow pattern were found to depend on the gas superficial velocity. The effect of the gravity magnitude had an effect on the frictional pressure drop only at low flow rates. The present data are used to evaluate some of existing flow pattern transition and pressure drop models and correlations.

• Progress in Medical Physics
• /
• v.24 no.1
• /
• pp.25-34
• /
• 2013

To investigate effects of phase mask on susceptibility-weighted images (SWI) using voxel-based analyses in normal elderly subjects. A three-dimensional (3D) gradient echo sequence ran to obtain SWIs in 20 healthy elderly subjects. SWIs with two (SWI2) and four (SWI4) phase multiplications were achieved with positive (PSWI) and negative (NSWI) phase masks to investigate phase mask effects. The voxel-based comparisons were performed using paired t-tests between PSWI and NSWI and between SWI2 and SWI4. Differences of signal intensities between magnitude images and SWI4 were larger than those between magnitude images and SWI2s. Differences of signal intensities between magnitude images and PSWIs were larger than those between magnitude images and NSWIs. Moreover, the signal intensities from NSWI2s and NSWI4s were greater than those from PSWI2s and PSWI4s, respectively. More differences of signal intensities between NSWI4 and PSWI4s were found than those between NSWI2s and PSWI2s in the whole brain images. The voxel-based analyses of SWI could be beneficial to investigate susceptibility differences on the entire brain areas. The phase masking method could be chosen to enhance brain tissue contrast rather than to enhance venous blood vessels. Therefore, it is recommended to apply voxel-based analyses of SWI to investigate clinical applications.