Shigeki Hirobayashi

Noise reduction for periodic signals using high-resolution frequency analysis

The spectrum subtraction method is one of the most common methods by which to remove noise from a spectrum. Like many noise reduction methods, the spectrum subtraction method uses discrete Fourier transform (DFT) for frequency analysis. There is generally a trade-off between frequency and time resolution in DFT. If the frequency resolution is low, then the noise spectrum can overlap with the signal source spectrum, which makes it difficult to extract the latter signal. Similarly, if the time resolution is low, rapid frequency variations cannot be detected. In order to solve this problem, as a frequency analysis method, we have applied non-harmonic analysis (NHA), which has high accuracy for detached frequency components and is only slightly affected by the frame length. Therefore, we examined the effect of the frequency resolution on noise reduction using NHA rather than DFT as the preprocessing step of the noise reduction process. The accuracy in extracting single sinusoidal waves from a noisy environment was first investigated. The accuracy of NHA was found to be higher than the theoretical upper limit of DFT. The effectiveness of NHA and DFT in extracting music from a noisy environment was then investigated. In this case, NHA was found to be superior to DFT, providing an approximately 2 dB improvement in SNR.

Dynamic model to estimate the dependence of gas sensor characteristics on temperature and humidity in environment

Abstract A gas sensor using tin oxide changes its electrical resistance ( R s ) with a change in the concentration of a reducing gas ( C ). However, the R s vs. C correlation is affected significantly by the temperature ( t ) and relative humidity ( h ) of the surrounding atmosphere. We propose a dynamic model to describe the influences of t and h . As is well-known, the R s vs. C correlation was expressed satisfactorily by an equation, log 10 R s = α log 10 C + β ( α and β constant), under fixed conditions of t and h . Analysis of the observed data indicated that the values of α and β under various environmental conditions were distributed along a plane when plotted three-dimensionally against t and h . This fact allows to propose a simple model in which both α and β are linear functions of t and h , respectively. The three constants included in each function were determined experimentally. The resulting equation of α and β were found to describe the R s vs. C correlation well under various conditions of t and h .

Verification of a logarithmic model for estimation of gas concentrations in a mixture for a tin oxide gas sensor response

Abstract Changes in electrical conductivity due to a reaction occurring between the original adsorbents and the gases present in the surrounding atmosphere were measured for a tin dioxide-based gas sensor. In this study, a model of the gas sensor response using a relationship between the gas concentration in a mixture and the sensor resistance is proposed. The values of coefficients are determined by a least-squares fit of measured data. Using two gas sensors having different characteristics, the concentrations of gases in a mixture can be evaluated. The proposed method has been applied to the evaluation of the gas concentrations in ammonia–ethanol and CO–ethanol mixtures. Furthermore, we verified our model for a gas mixture of carbon monoxide, propane and methane gases, namely for a three-gas system. The results of the present study indicated that the evaluation of gas concentrations in a mixture is feasible using the proposed model of gas sensor response, which can be used as an inexpensive monitor for air pollution.

Motion Analysis Using 3D High-Resolution Frequency Analysis

The spatiotemporal spectra of a video that contains a moving object form a plane in the 3D frequency domain. This plane, which is described as the theoretical motion plane, reflects the velocity of the moving objects, which is calculated from the slope. However, if the resolution of the frequency analysis method is not high enough to obtain actual spectra from the object signal, the spatiotemporal spectra disperse away from the theoretical motion plane. In this paper, we propose a high-resolution frequency analysis method, described as 3D nonharmonic analysis (NHA), which is only weakly influenced by the analysis window. In addition, we estimate the motion vectors of objects in a video using the plane-clustering method, in conjunction with the least-squares method, for 3D NHA spatiotemporal spectra. We experimentally verify the accuracy of the 3D NHA and its usefulness for a sequence containing complex motions, such as cross-over motion, through comparison with 3D fast Fourier transform. The experimental results show that increasing the frequency resolution contributes to high-accuracy estimation of a motion plane.

Verification of Individual Identification Method Using Bioelectric Potential of Plant during Human Walking

In this study, we monitored electromagnetic waves generated by human activity and investigated a method for individual identification by looking at the bioelectric potential of a rubber tree. Four subjects were asked to walk in place at a distance of 60 cm from a rubber tree while we measured variations in the bioelectric potential of the tree as produced by the stepping. The results confirmed that electromagnetic waves generated by a human subject walking in place produce a measurable response in the bioelectric potential of a plant. It was also found that this variation in bioelectric potential varies in synchrony with each subjects walking pace. The spectral envelope of the observed signal was approximated using a straight line and the distribution of the coefficients of this line was plotted. Even for a simple straight-line approximation, we demonstrated that this coefficient distribution varies considerably among individual subjects.

Image Denoising With Edge-Preserving and Segmentation Based on Mask NHA

In this paper, we propose a zero-mean white Gaussian noise removal method using a high-resolution frequency analysis. It is difficult to separate an original image component from a noise component when using discrete Fourier transform or discrete cosine transform for analysis because sidelobes occur in the results. The 2D non-harmonic analysis (2D NHA) is a high-resolution frequency analysis technique that improves noise removal accuracy because of its sidelobe reduction feature. However, spectra generated by NHA are distorted, because of which the signal of the image is non-stationary. In this paper, we analyze each region with a homogeneous texture in the noisy image. Non-uniform regions that occur due to segmentation are analyzed by an extended 2D NHA method called Mask NHA. We conducted an experiment using a simulation image, and found that Mask NHA denoising attains a higher peak signal-to-noise ratio (PSNR) value than the state-of-the-art methods if a suitable segmentation result can be obtained from the input image, even though parameter optimization was incomplete. This experimental result exhibits the upper limit on the value of PSNR in our Mask NHA denoising method. The performance of Mask NHA denoising is expected to approach the limit of PSNR by improving the segmentation method.

Image Inpainting on the Basis of Spectral Structure From 2-D Nonharmonic Analysis

The restoration of images by digital inpainting is an active field of research and such algorithms are, in fact, now widely used. Conventional methods generally apply textures that are most similar to the areas around the missing region or use a large image database. However, this produces discontinuous textures and thus unsatisfactory results. Here, we propose a new technique to overcome this limitation by using signal prediction based on the nonharmonic analysis (NHA) technique proposed by the authors. NHA can be used to extract accurate spectra, irrespective of the window function, and its frequency resolution is less than that of the discrete Fourier transform. The proposed method sequentially generates new textures on the basis of the spectrum obtained by NHA. Missing regions from the spectrum are repaired using an improved cost function for 2D NHA. The proposed method is evaluated using the standard images Lena, Barbara, Airplane, Pepper, and Mandrill. The results show an improvement in MSE of about 10-20 compared with the examplar-based method and good subjective quality.

Forecast of stock market based on nonharmonic analysis used on NASDAQ since 1985

Although research involving economic time series forecasting based on virtual market models is frequently conducted, long-term forecasting is difficult due to many factors that affect actual markets. However, as exemplified by the business cycle and Elliot Wave theories in economics, it is assumed that fluctuations in economic time series forecasting have various periodicities, ranging from short-term to long-term. Accordingly, we used a new high-resolution frequency analysis (Non-Harmonic Analysis (NHA)) method, which we have recently developed, to conduct analysis of the periodicity of economic time series forecasting. We also attempted a long-term economic time series forecast by combining multiple periodic signals. In the verification experiment, we analysed the National Association of Securities Dealers Automated Quotations (NASDAQ) closing price data for a time period of approximately 20 years using nonharmonic analysis with an analysis window of the previous 2 years, and forecasted price fluctuations for the following 2 years.

Detection of human activities by inverse filtration of gas sensor response

Abstract In an approach to identifying human activities in an indoor space from the response signals of a combustible gas sensor, an indoor space model was presented in which the response signal and the gas generation due to the human activities were correlated linearly through an indoor gas transfer function. Analysis indicated that: (1) the gas transfer by diffusion can describe essential features of the transfer function in the indoor space, and (2) the inverse filtration based on the space transfer characteristics thus estimated enables one to restore or extract the gas-generation function from the sensor signal output. The model was applied to identify daily human activities in a domestic house. The sensor response output consisted of heavily overlapping signals due to human actions taking place in succession. However, the inverse filtration could eliminate the overlapping (residual) components from the overall output to single out each human action associated with causing gas generation in the indoor space.

Slit Structure as a Countermeasure for the Thermal Deformation of a Metal Mask

A metal mask used for the pattern formation of sputtered films is sometimes deformed during film deposition, because its temperature increases during deposition. In this study, a slit structure as a countermeasure for the deformation is proposed. The mask investigated is made of 0.3-mm-thick stainless steel and is composed of a frame with many beams running inside the frame. The slits in the structure absorb the thermal expansion of the beams, which results in the relaxation of thermal stress in the beams. The relaxation of the stress is analyzed using the finite element method as well as the theory of beam bending. It was experimentally confirmed that a mask furnished with well-designed slits was not deformed.