• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.4D
• /
• pp.687-693
• /
• 2006

Image co-registration is the process of overlaying two images of the same scene, one of which represents a reference image, while the other is geometrically transformed to the one. In order to improve efficiency and effectiveness of the co-registration approach, the author proposed a pre-qualified area matching algorithm which is composed of feature extraction with canny operator and area matching algorithm with cross correlation coefficient. For refining matching points, outlier detection using studentized residual was used and iteratively removes outliers at the level of three standard deviation. Throughout the pre-qualification and the refining processes, the computation time was significantly improved and the registration accuracy is enhanced. A prototype of the proposed algorithm was implemented and the performance test of 3 Landsat images of Korea. showed: (1) average RMSE error of the approach was 0.435 pixel; (2) the average number of matching points was over 25,573; (3) the average processing time was 4.2 min per image with a regular workstation equipped with a 3 GHz Intel Pentium 4 CPU and 1 Gbytes Ram. The proposed approach achieved robustness, full automation, and time efficiency.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.5
• /
• pp.251-258
• /
• 2004

This paper proposes a method for estimating the illuminant chromaticity using the distributions of the camera responses obtained by a CCD camera in a real-world scene. Illuminant estimation using a highlight method is based on the geometric relation between a body and its surface reflection. In general, the pixels in a highlight region are affected by an illuminant geometric difference, camera quantization errors, and the non-uniformity of the CCD sensor. As such, this leads to inaccurate results if an illuminant is estimated using the pixels of a CCD camera without any preprocessing. Accordingly, to solve this problem the proposed method analyzes the distribution of the CCD camera responses and selects pixels using the Mahalanobis distance in highlight regions. The use of the Mahalanobis distance based on the camera responses enables the adaptive selection of valid pixels among the pixels distributed in the highlight regions. Lines are then determined based on the selected pixels with r-g chromaticity coordinates using a principal component analysis(PCA). Thereafter, the illuminant chromaticity is estimated based on the intersection points of the lines. Experimental results using the proposed method demonstrated a reduced estimation error compared with the conventional method.

• Journal of Convergence for Information Technology
• /
• v.11 no.6
• /
• pp.1-6
• /
• 2021

In the field of speech recognition, as the DNN is applied, the use of speech recognition is increasing, but the amount of calculation for parallel training needs to be larger than that of the conventional GMM, and if the amount of data is small, overfitting occurs. To solve this problem, we propose an efficient method for robust voice feature extraction and voice signal noise removal even when the amount of data is small. Speech feature extraction efficiently extracts speech energy by applying the difference in frame energy for speech and the zero-crossing ratio and level-crossing ratio that are affected by the speech signal. In addition, in order to remove noise, the noise of the speech signal is removed by removing the noise of the speech signal with an average predictive improved LMS filter with little loss of speech information while maintaining the intrinsic characteristics of speech in detection of the speech signal. The improved LMS filter uses a method of processing noise on the input speech signal by adjusting the active parameter threshold for the input signal. As a result of comparing the method proposed in this paper with the conventional frame energy method, it was confirmed that the error rate at the start point of speech is 7% and the error rate at the end point is improved by 11%.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.33 no.4
• /
• pp.231-244
• /
• 2015

The accuracy of reference point of VLBI antenna is mandatory to perform collocation of different space geodetic techniques. In this study, we evaluated the optimal methods for the 3D circle fitting to enhance the accuracy of the reference point of VLBI antenna. Two kinds of methodologies for the orthonormal coordinate system with translation of planar observation point and the unitary coordinate transforamation were suggested and their fitting accuracies were evaluated where the orthogonal distance was calculated by residual between observation point and fitting model and the recursive calculation was performed to improve the accuracy of 3D circle fitting. Finally, we found that the methodology for the unitary coordinate transformation is highly appropriate to determine the optimal equation for azimuth-axis and elevation-axis of VLBI antenna. Therefore, the reference point of VLBI antenna with high accuracy can be determined by the intersection of the above two axises (azimuth-axis and elevation-axis). This result is expected to be utilized for a variety of researches for connection between VLBI observation results and the national control point.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.27 no.2
• /
• pp.213-223
• /
• 2009

In geoid modelling field, it is very important the GPS/leveling data because it could be check-out the accuracy of gravimetric geoid and computed the hybrid geoid. In this study, GPS surveying was accomplished in the test area including mountainous area to improve the GPS/leveling data density in Korea. And the geometric geoidal heights was calculated using the GPS/leveling data in the test area and the accuracy of the geoidal heights was analyzed. For this study, GPS surveying was accomplished on the 211 1st and 2nd order benchmarks in Gyeongbuk province and 198 GPS/leveling data were achieved after both baseline analysis and network adjustment. Geometric geoidal heights were calculated using these 198 GPS/leveling data and the accuracy analysis was done by comparison with the geoidal heights from EGM2008 geopotential model. The results showed that the bias and standard deviation computed from 190 GPS/leveling data after gross removal was -0.185$\pm$0.079m. And also, the accuracy analyses according to the benchmark order, baseline length, and altitude were accomplished.

• Explosives and Blasting
• /
• v.18 no.2
• /
• pp.7-13
• /
• 2000

Abstract The characteristics of bed rock in the study area was classified by means of the crack coefficient estimated from the seismic velocities of in-situ and intact rocks. Various statistical methods were investigated in order to minimize the possible errors in estimating the predictive equation of blasting vibration and to enhance the determination coefficient $R^2$, for more reliable estimation. The determination coefficient showed the highest in the analysis for those groups using weighting function with the number of samples. The analysis for the weighting function employed with standard coefficient and variance also enhanced the determination coefficients significantly compared to the others, but the reliability was slightly lower than results obtained former method. Therefore the most reliable predictive equation of blasting vibration was found to be obtained from a regression analysis of the mean vibration level using the weighting of same distance groups within 15m with the same explosive charge weight per delay. The coefficients, K and n 317.4 and -1.66, respectively, when using the square root scaling, and 209.9 and -1.66, respectively, when using the cube root scaling. The analysis also showed that the square root scaling may be used in the distance less than 31m form the blast source, and the cube root scaling in the distance more than 31m for safe design.

• Economic and Environmental Geology
• /
• v.48 no.6
• /
• pp.451-465
• /
• 2015

The free-air anomalies are computed using a data set from various types of gravity measurements in the Korean Peninsula area. The gravity values extracted from the Earth Gravitational Model 2008 are used in the surrounding region. The upward continuation technique suggested by Dragomir is used in the computation of the external free-air anomalies at various altitudes. The integration radius 10 times the altitude is used in order to keep the accuracy of results and computational resources. The direct geodesic formula developed by Bowring is employed in integration. At the 1-km altitude, the free-air anomalies vary from -41.315 to 189.327 mgal with the standard deviation of 22.612 mgal. At the 3-km altitude, they vary from -36.478 to 156.209 mgal with the standard deviation of 20.641 mgal. At the 1,000-km altitude, they vary from 3.170 to 5.864 mgal with the standard deviation of 0.670 mgal. The predicted free-air anomalies at 3-km altitude are compared to the published free-air anomalies reduced from the airborne gravity measurements at the same altitude. The rms difference is 3.88 mgal. Considering the reported 2.21-mgal airborne gravity cross-over accuracy, this rms difference is not serious. Possible causes in the difference appear to be external free-air anomaly simulation errors in this work and/or the gravity reduction errors of the other. The external gravity field is predicted by adding the external free-air anomaly to the normal gravity computed using the closed form formula for the gravity above and below the surface of the ellipsoid. The predicted external gravity field in this work is expected to reasonably present the real external gravity field. This work seems to be the first structured research on the external free-air anomaly in the Korean Peninsula area, and the external gravity field can be used to improve the accuracy of the inertial navigation system.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.5
• /
• pp.8-14
• /
• 2009

In the research field of speech recognition, pinpointing the endpoints of speech utterance even with the presence of background noise is of great importance. These noise present during recording introduce disturbances which complicates matters since what we just want is to get the stationary parameters corresponding to each speech section. One major cause of error in automatic recognition of isolated words is the inaccurate detection of the beginning and end boundaries of the test and reference templates, thus the necessity to find an effective method in removing the unnecessary regions of a speech signal. The conventional methods for speech endpoint detection are based on two linear time-domain measurements: the short-time energy, and short-time zero-crossing rate. They perform well for clean speech but their precision is not guaranteed if there is noise present, since the high energy and zero-crossing rate of the noise is mistaken as a part of the speech uttered. This paper proposes a novel approach in finding an apparent threshold between noise and speech based on Lyapunov Exponents (LEs). This proposed method adopts the nonlinear features to analyze the chaos characteristics of the speech signal instead of depending on the unreliable factor-energy. The excellent performance of this approach compared with the conventional methods lies in the fact that it detects the endpoints as a nonlinearity of speech signal, which we believe is an important characteristic and has been neglected by the conventional methods. The proposed method extracts the features based only on the time-domain waveform of the speech signal illustrating its low complexity. Simulations done showed the effective performance of the Proposed method in a noisy environment with an average recognition rate of up 92.85% for unspecified person.

• Ocean and Polar Research
• /
• v.24 no.4
• /
• pp.491-500
• /
• 2002

Magnetic anomalies in the Dokdo and it's surroundings were investigated with respect to structure and origin of the Dokdo and surrounding seamounts. After normal and diurnal correction of measured magnetic data, crossover correction was applied to reduce errors between sets of magnetic anomalies. The errors from crossover operation result in decrease of about 51%, from 62.2 nT to 30.1 nT in standard deviation. Reduction-to-the-pole, second vertical derivative and analytic signal processing were applied to explore magnetic anomaly signatures in detail. Magnetic anomalies are most complicated in the 1st-Dok seamount, show SWW-NEE linear pattern in the 2nd-Dok seamount and lower to the 3rd-Dok seamount. Different magnetic anomaly patterns in three seamounts imply that three volcanic seamounts were formed at different times and are composed of rocks that were produced in different conditions. It seems that the 3rd-Dok seamount was first to form and followed by the 1st-Dok seamount. The complicated magnetic and second vertical derivative anomaly patterns in the 1st-Dok seamount may be due to subsidiary cones around crater or the presence of intruded magma bodies below sea surface and the Dokdo is probably a marginal subsidiary part of crater.

• Korean Journal of Remote Sensing
• /
• v.27 no.4
• /
• pp.411-420
• /
• 2011

In this study, a digital elevation model (DEM) in tidal flat of Suncheon Bay, one of the most ecological preserved area in the world, was generated from digital aerial stereo-images. The focal lengths for the aerial stereo-images were adjusted using ground control points (GCPs) in order to improve the accuracy of camera parameters. We proposed matching sizes suitable for generating DEM in tidal flat and a method for eliminating excessive position errors using intersection-distance($P_R$) threshold value. The accuracies of the DEM generated from the proposed method as well as the commercial S/W were compared with the elevation profiles measured by Total Station in the filed. As the results, the DEM generated by the proposed method showed better result (maximum deviation is a -21 centimeters) with detailed topography than DEM by the commercial S/W in the region. These results suggest that the DEM of tidal flat, which hardly obtained with the traditional methods, can be generated from digital aerial images by applying the proposed method in this study. We believe that the generated DEM in tidal flat can be an essential data for monitoring the sediment erosion and deposit of the tidal flat.