• Journal of the Korean earth science society
• /
• v.30 no.5
• /
• pp.550-564
• /
• 2009

When magnetotelluric (MT) data are obtained in the vicinity of the coast, the surrounding seas make it difficult to interpret subsurface structure, especially the deep part of the subsurface. We introduce an iterative method to correct the sea effect, based on the previous topographic correction method that removes the distortion due to topographic changes in seafloor MT data. The method first corrects the sea effect in observed MT impedance, and then inverts corrected response in a model space without the sea. Due to mutual coupling between the sea and the subsurface structure, the correction and inversion steps are iterated until the changes in each result become negligible. The method is tested for 1- and 2-D structures using synthetic MT data produced by 3-D forward modeling including surrounding seas. In all cases, the method closely recovers the true structure assumed to generate synthetic responses after a few iterations.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.32 no.10
• /
• pp.367-375
• /
• 1983

An extended Kalman filter (EKF) algorithm for estimating aerodynamic parameters from missile flight data is evaluated using simulated test data. The algorithm includes a general purpose 6-DOF missile airframe suitable for representing a variety of missile configurations. The EKF is demonstrated to be well suited as a postflight analysis tool for extracting large numbers of airframe parameters from flight test measurements. A structure identification algorithm is evaluated using synthetic measurement data. This algorithm used in conjunction with the parameter identification algorithm, can select that model from a family of candidate models which most likely produced the synthetic measurement data.

• Genomics & Informatics
• /
• v.11 no.2
• /
• pp.76-82
• /
• 2013

Over the past decade or so, dramatic developments in our ability to experimentally determine the content and function of genomes have taken place. In particular, next-generation sequencing technologies are now inspiring a new understanding of bacterial transcriptomes on a global scale. In bacterial cells, whole-transcriptome studies have not received attention, owing to the general view that bacterial genomes are simple. However, several recent RNA sequencing results are revealing unexpected levels of complexity in bacterial transcriptomes, indicating that the transcribed regions of genomes are much larger and complex than previously anticipated. In particular, these data show a wide array of small RNAs, antisense RNAs, and alternative transcripts. Here, we review how current transcriptomics are now revolutionizing our understanding of the complexity and regulation of bacterial transcriptomes.

• Proceedings of the KSRS Conference
• /
• 2008.10a
• /
• pp.441-444
• /
• 2008

This paper describes the development of a circularly polarized microstrip antenna, as a part of the Circularly Polarized Synthetic Aperture Radar (CP-SAR) sensor which is currently under developed at the Microwave Remote Sensing Laboratory (MRSL) in Chiba University. CP-SAR is a new type of sensor developed for the purpose of remote sensing. With this sensor, lower-noise data/image will be obtained due to the absence of depolarization problems from propagation encounter in linearly polarized synthetic aperture radar. As well the data/images obtained will be investigated as the Axial Ratio Image (ARI), which is a new data that hopefully will reveal unique various backscattering characteristics. The sensor will be mounted on an Unmanned Aerial Vehicle (UAV) which will be aimed for fundamental research and applications. The microstrip antenna works in the frequency of 1.27 GHz (L-Band). The microstrip antenna utilized the proximity-coupled method of feeding. Initially, the optimization process of the single patch antenna design involving modifying the microstrip line feed to yield a high gain (above 5 dBi) and low return loss (below -10 dB). A minimum of 10 MHz bandwidth is targeted at below 3 dB of Axial Ratio for the circularly polarized antenna. A planar array from the single patch is formed next. Consideration for the array design is the beam radiation pattern in the azimuth and elevation plane which is specified based on the electrical and mechanical constraints of the UAV CP-SAR system. This research will contribute in the field of radar for remote sensing technology. The potential application is for landcover, disaster monitoring, snow cover, and oceanography mapping.

• Information Systems Review
• /
• v.24 no.4
• /
• pp.97-118
• /
• 2022

As generative adversarial network (GAN) based oversampling techniques have achieved impressive results in class imbalance of unstructured dataset such as image, many studies have begun to apply it to solving the problem of imbalance in structured dataset. However, these studies have failed to reflect the characteristics of structured data due to changing the data structure into an unstructured data format. In order to overcome the limitation, this study adapted CycleGAN to reflect the characteristics of structured data, and proposed hybridization of synthetic minority oversampling technique (SMOTE) and the adapted CycleGAN. In particular, this study tried to overcome the limitations of existing studies by using a one-dimensional convolutional neural network unlike previous studies that used two-dimensional convolutional neural network. Oversampling based on the method proposed have been experimented using various datasets and compared the performance of the method with existing oversampling methods such as SMOTE and adaptive synthetic sampling (ADASYN). The results indicated the proposed hybrid oversampling method showed superior performance compared to the existing methods when data have more dimensions or higher degree of imbalance. This study implied that the classification performance of oversampling structured data can be improved using the proposed hybrid oversampling method that considers the characteristic of structured data.

• Journal of Broadcast Engineering
• /
• v.26 no.4
• /
• pp.453-462
• /
• 2021

A lot of license plate data is required for car number recognition. License plate data needs to be balanced from past license plates to the latest license plates. However, it is difficult to obtain data from the actual past license plate to the latest ones. In order to solve this problem, a license plate recognition study through deep learning is being conducted by creating a synthetic license plates. Since the synthetic data have differences from real data, and various data augmentation techniques are used to solve these problems. Existing data augmentation simply used methods such as brightness, rotation, affine transformation, blur, and noise. In this paper, we apply a style transformation method that transforms synthetic data into real-world data styles with data augmentation methods. In addition, real license plate data are noisy when it is captured from a distance and under the dark environment. If we simply recognize characters with input data, chances of misrecognition are high. To improve character recognition, in this paper, we applied the DeblurGANv2 method as a quality improvement method for character recognition, increasing the accuracy of license plate recognition. The method of deep learning for license plate detection and license plate number recognition used YOLO-V5. To determine the performance of the synthetic license plate data, we construct a test set by collecting our own secured license plates. License plate detection without style conversion recorded 0.614 mAP. As a result of applying the style transformation, we confirm that the license plate detection performance was improved by recording 0.679mAP. In addition, the successul detection rate without image enhancement was 0.872, and the detection rate was 0.915 after image enhancement, confirming that the performance improved.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.49 no.12
• /
• pp.963-969
• /
• 2021

The wind data measured from local meteorological masts is used to evaluate wind speed distribution and energy production in the specified site for wind farm However, wind data measured from meteorological masts often contain missing information or insufficient desired height or data length, making it difficult to perform wind turbine control and performance simulation. Therefore, long-term continuous wind data is very important to assess the annual energy production and the capacity factor for wind turbines or wind farms. In addition, if seasonal influences are distinct, such as on the Korean Peninsula, wind data with seasonal characteristics should be considered. This study presents methodologies for generating synthetic wind that take into account fluctuations in both wind speed and direction using the hidden Markov model, which is a statistical method. The wind data for statistical processing are measured at Maldo island in the Kokunnsan-gundo, Jeonbuk Province using the Automatic Weather System (AWS) of the Korea Meteorological Administration. The synthetic wind generated using the hidden Markov model will be validated by comparing statistical variables, wind energy density, seasonal mean speed, and prevailing wind direction with measurement data.

• Smart Media Journal
• /
• v.8 no.2
• /
• pp.46-57
• /
• 2019

In recent research, deep learning-based methods have achieved state-of-the-art performance in various computer vision tasks. However, these methods are commonly supervised, and require huge amounts of annotated data to train. Acquisition of data demands an additional costly effort, particularly for the tasks where it becomes challenging to obtain large amounts of data considering the time constraints and the requirement of professional human diligence. In this paper, we present a data level synthetic sampling solution to learn from small and imbalanced data sets using Generative Adversarial Networks (GANs). The reason for using GANs are the challenges posed in various fields to manage with the small datasets and fluctuating amounts of samples per class. As a result, we present an approach that can improve learning with respect to data distributions, reducing the partiality introduced by class imbalance and hence shifting the classification decision boundary towards more accurate results. Our novel method is demonstrated on a small dataset of 2789 tomato plant disease images, highly corrupted with class imbalance in 9 disease categories. Moreover, we evaluate our results in terms of different metrics and compare the quality of these results for distinct classes.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.20 no.2
• /
• pp.187-196
• /
• 2017

The size of raw data has dramatically increased due to the recent trend of Synthetic Aperture Radar(SAR) development plans for high resolution and high definition image acquisition. The large raw data has an impact on satellite operability due to the limitations of storage and transmission capacity. To improve the SAR operability, the SAR raw data shall be compressed before transmission to the ground station. The Block Adaptive Quantization (BAQ) algorithm is one of the data compression algorithm and has been used for a long time in the spaceborne SAR system. In this paper, an optimization method of BAQ threshold table is introduced using real SAR raw data to prevent the degradation of signal quality caused by data compression. In this manner, a new variation estimation strategy and a new threshold method for block type decision are introduced.

• International Journal of Internet, Broadcasting and Communication
• /
• v.11 no.4
• /
• pp.37-42
• /
• 2019

In this paper, we explore the details of three classic data augmentation methods and two generative model based oversampling methods. The three classic data augmentation methods are random sampling (RANDOM), Synthetic Minority Over-sampling Technique (SMOTE), and Adaptive Synthetic Sampling (ADASYN). The two generative model based oversampling methods are Conditional Generative Adversarial Network (CGAN) and Wasserstein Generative Adversarial Network (WGAN). In imbalanced data, the whole instances are divided into majority class and minority class, where majority class occupies most of the instances in the training set and minority class only includes a few instances. Generative models have their own advantages when they are used to generate more plausible samples referring to the distribution of the minority class. We also adopt CGAN to compare the data augmentation performance with other methods. The experimental results show that WGAN-based oversampling technique is more stable than other approaches (RANDOM, SMOTE, ADASYN and CGAN) even with the very limited training datasets. However, when the imbalanced ratio is too small, generative model based approaches cannot achieve satisfying performance than the conventional data augmentation techniques. These results suggest us one of future research directions.