• Journal of Korea Water Resources Association
• /
• v.56 no.10
• /
• pp.641-653
• /
• 2023

As climate change and urbanization are causing unprecedented natural disasters in urban areas, it is crucial to have urban flood predictions with high fidelity and accuracy. However, conventional physically- and deep learning-based urban flood modeling methods have limitations that require a lot of computer resources or data for high-resolution flooding analysis. In this study, we propose and implement a method for improving the spatial resolution of urban flood analysis using a deep learning based super-resolution technique. The proposed approach converts low-resolution flood maps by physically based modeling into the high-resolution using a super-resolution deep learning model trained by high-resolution modeling data. When applied to two cases of retrospective flood analysis at part of City of Portland, Oregon, U.S., the results of the 4-m resolution physical simulation were successfully converted into 1-m resolution flood maps through super-resolution. High structural similarity between the super-solution image and the high-resolution original was found. The results show promising image quality loss within an acceptable limit of 22.80 dB (PSNR) and 0.73 (SSIM). The proposed super-resolution method can provide efficient model training with a limited number of flood scenarios, significantly reducing data acquisition efforts and computational costs.

• Proceedings of the KIPE Conference
• /
• 1997.07a
• /
• pp.177-181
• /
• 1997

In the speed control system of motors using the low resolution rotary encoder, the period of encoder pulse becomes longer than the sampling time for speed control in the range of very low speed. Therefore, it is difficult to obtain accurate speed information. In this paper, the speed estimating method at the very low speed region using reduced order torque observer, which has been widely used, is examined. The results of simulation show that the characteristics of the speed control at the very low speed region is improved by using the reduced order torque observer.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.12
• /
• pp.6284-6289
• /
• 2013

With the application of k-space trajectory in a different manner and analyzing the influence of noise and its direction, this study was conducted to obtain high-quality images with minimal influence of noise during an MRI examination for cerebrovascular disease, which has a low signal for imaging. To evaluate influence of the noise of different k-space trajectories, a linear Cartesian coordination trajectory and non-Cartesian coordination trajectory were applied to 38 people who had received a high-resolution MRI examination for the early detection of cerebrovascular disease. As a result, the non-Cartesian coordination trajectory showed a 43.32% lower signal of lumens in the internal carotid artery than a linear Cartesian coordination trajectory, and the noise level was also 50.19% lower in a non-Cartesian coordination trajectory. This result shows that noise occurs less in a non-Cartesian coordination trajectory than a linear Cartesian coordination trajectory, and a non-Cartesian coordination trajectory is more effective in low-signal and low-resolution MRI examination. Therefore, when performing high-resolution MRI examination with a low-signal cerebrovascular system, the use of non-Cartesian coordination k-space trajectory will minimize the influence of noise and provide good images.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.6
• /
• pp.1-8
• /
• 2021

In this paper, we propose reinforced VGG style network structure for low performance embedded system to classify low resolution infrared image. The combination of reinforced VGG style network structure and global average pooling makes lower computational complexity and higher accuracy. The proposed method classify the synthesize image which have 9 class 3,723,328ea images made from OKTAL-SE tool. The reinforced VGG style network structure composed of 4 filters on input and 16 filters on output from max pooling layer shows about 34% lower computational complexity and about 2.4% higher accuracy then the first parameter minimized network structure made for embedded system composed of 8 filters on input and 8 filters on output from max pooling layer. Finally we get 96.1% accuracy model. Additionally we confirmed the about 31% lower inference lead time in ported C code.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.48 no.9
• /
• pp.22-30
• /
• 2011

Recently, the rising demand for intelligent video surveillance system leads to high-performance face recognition systems. The solution for low-resolution images acquired by a long-distance camera is required to overcome the distance limits of the existing face recognition systems. For that reason, this paper proposes a hardware design of an image resolution enhancement algorithm for real-time intelligent video surveillance systems. The algorithm is synthesizing a high-resolution face image from an input low-resolution image, with the help of a large collection of other high-resolution face images, called training set. When we checked the performance of the algorithm at 32bit RISC micro-processor, the entire operation took about 25 sec, which is inappropriate for real-time target applications. Based on the result, we implemented the hardware module and verified it using Xilinx Virtex-4 and ARM9-based embedded processor(S3C2440A). The designed hardware can complete the whole operation within 33 msec, so it can deal with 30 frames per second. We expect that the proposed hardware could be one of the solutions not only for real-time processing at the embedded environment, but also for an easy integration with existing face recognition system.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.27 no.2
• /
• pp.79-93
• /
• 2015

Three-dimensional structures of large ocean rings in the Gulf Stream region are investigated using the HYbrid Coordinate Ocean Model (HYCOM). Numerically simulated flow structures around four selected cyclonic and anticyclonic rings are compared with two different horizontal resolutions: $1/12^{\circ}$ and $1/48^{\circ}$. The vertical distributions of Lagrangian Coherent Structures (LCSs) are analyzed using Finite Size Lyapunov Exponent (FSLE) and Okubo-Weiss parameters (OW). Curtain-shaped FSLE ridges are found in all four rings with extensions of surface ridges throughout the water columns, indicating that horizontal stirring is dominant over vertical motions. Near the high-resolution rings, many small-scale flow structures with size O(1~10) km are observed while these features are rarely found near the low-resolution rings. These small-scale structures affect the flow pattern around the rings as flow particles move more randomly in the high-resolution models. The dispersion rates are also affected by these small-scale structures as the relative horizontal dispersion coefficients are larger for the high-resolution models. The absolute vertical dispersion rates are, however, lower for the high-resolution models, because the particles tend to move along inclined eddy orbits when the resolution is low and this increases the magnitude of absolute vertical dispersion. Since relative vertical dispersion can reduce this effect from the orbital trajectories of particles, it gives a more reasonable magnitude range than absolute dispersion, and so is recommended in estimating vertical dispersion rates.

• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.2
• /
• pp.79-87
• /
• 2011

This paper presents the optimal design of an overlapped ultrasonic sensor ring for high resolution obstacle detection of an autonomous mobile robot. It is assumed that a set of low directivity ultrasonic sensors of the same type are arranged along a circle of nonzero radius at a regular spacing with their beams overlapped. First, taking into account the dead angle region, the entire range of obstacle detection is determined with reference to the center of an overlapped ultrasonic sensor ring. Second, the optimal design index of an overlapped ultrasonic sensor ring is defined as the area closeness of three sensing subzones resulting from beam overlap. Third, the lower and upper bounds on the number of ultrasonic sensors are derived, which can guarantee minimal beam overlap and also avoid excessive beam overlap among adjacent ultrasonic sensors. Fourth, employing a commercial low directivity ultrasonic sensor, an optimal design example of an overlapped ultrasonic sensor ring is given along with the ultrasonic sensor ring prototype mounted on top of a mobile robot. Finally, some experimental results using our prototype ultrasonic sensor ring are given to demonstrate the validity and performance of an optimally overlapped ultrasonic sensor ring for high resolution obstacle detection.

• The Bulletin of The Korean Astronomical Society
• /
• v.41 no.2
• /
• pp.43.4-44
• /
• 2016

The presence of multiple stellar populations is now well established in most globular clusters (GCs) in the Milky Way. The origin of this phenomenon, however, is yet to be understood. In this respect, the study of NGC 2808, an intriguing GC which hosts subpopulations with extreme helium abundances, would help to resolve this problem. In order to investigate chemical abundance patterns among different subpopulations in this GC, we have performed low-resolution spectroscopy for the red-giant-branch (RGB) stars and measured CN & CH bands, and Ca line strength. We have identified at least three subpopulations from the CN abundance distribution. This GC shows CN-CH anti-correlation following the general trend among "normal" GCs. In addition, we have cross-matched our results with the high-resolution data in literature, and found a tight correlation between CN strength and sodium abundance. However, CN is anticorrelated with oxygen abundance, as expected from the well known N(&Na)-O anticorrelation. In this talk, we will discuss the implication of these results.

• Asian Pacific Journal of Cancer Prevention
• /
• v.14 no.11
• /
• pp.6619-6623
• /
• 2013

Background:The aim of the research was to explore a cost effective, fast, easy to perform, and sensitive method for epidermal growth factor receptor (EGFR) mutation testing. Methods: High resolution melting analysis (HRM) was introduced to evaluate the efficacy of the analysis for dectecting EGFR mutations in exons 18 to 21 using formalin-fixed paraffin-embedded (FFPE) tissues and plasma free DNA from 120 patients. Results: The total EGFR mutation rate was 37.5% (45/120) detected by direct sequencing. There were 48 mutations in 120 FFPE tissues assessed by HRM. For plasma free DNA, the EGFR mutation rate was 25.8% (31/120). The sensitivity of HRM assays in FFPE samples was 100% by HRM. There was a low false-positive mutation rate but a high false-negative rate in plasma free DNA detected by HRM. Conclusions: Our results show that HRM analysis has the advantage of small tumor sample need. HRM applied with plasma free DNA showed a high false-negative rate but a low false-positive rate. Further research into appropriate methods and analysis needs to be performed before HRM for plasma free DNA could be accepted as an option in diagnostic or screening settings.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1995.10a
• /
• pp.450-455
• /
• 1995

A three-dimensional numerical tool is developed to calculate the potential distribution, electric field, and conductance for any types of conductance probes immersed in the wavy liquid film with various shapes of its free surface. The tool is validated against various analytical solutions. It is applied to find out the characteristics of the wire-wire probe, the flush-wire probe and the flush-flush probe in terms of resolution, linearity, and sensitivity. The wire-wire probe shows high resolution and excellent linearity for various film thickness, but comparably low sensitivity for low film thickness fixed. The flush-wire probe shows good linearity and high sensitivity for varying film thickness, but resolution degrading with an increase in film thickness. In order to check the applicability of the three types of probes in the real situation, the Korteweg-de Vries(KdV) two-dimensional solitary wave is simulated. The wire-wire probe is strongly affected by the installation direction of the two wires; when the wires are installed perpendicularly to the flow direction, the wire-wire probe shows large distortion of the solitary wave. In order to measure the transverse profile of waves, the wire-wire probes and the flush-wire probes are required to be separately installed 2mm and 2mm, respectively.