• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.1
• /
• pp.63.2-63.2
• /
• 2019

Eight hundred thousand asteroids in the solar system have been identified so far under extensive sky surveys. Kilometer to sub-km sized asteroids, however, are still waiting for discovery, and their size and orbital distribution will provide a better understanding of the collisional and dynamical evolution of the solar system. In order to study the number of asteroids which is detectable with 1.6 m telescope and their orbital distribution, we conducted a small observation campaign as a part of Deep Ecliptic Patrol of the Southern Sky (DEEP-South) project, which is an asteroid survey in the southern hemisphere with Korea Microlensing Telescope Network (KMTNet). We observed the ecliptic plane near opposition ($2^{\circ}{\times}2^{\circ}$ field of view centering on ${\alpha}=22h40m31s$, ${\delta}=-08^{\circ}22^{\prime}58^{{\prime}{\prime}}$) in August 2018, and identified 464 moving objects by visual inspection. As a result, 266 of 464 moving objects turn out to be previously unknown asteroids, and their signal to noise ratio is below two on numerous occasions. Most of the newly detected objects are main belt asteroids (MBAs), while three Hildas, one Jupiter trojan, and two Hungarias are also identified. In this meeting, we report the differences in the orbital distributions between the previously known asteroids and newly discovered ones using statistical methods. We also talk about the observational bias of this survey and suggest future works.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.1
• /
• pp.52.1-52.1
• /
• 2019

The Helioseismic and Magnetic Imager (HMI) is the instrument of Solar Dynamics Observatory (SDO) to study the magnetic field and oscillation at the solar surface. The HMI image is not enough to analyze very small magnetic features on solar surface since it has a spatial resolution of one arcsec. Super resolution is a technique that enhances the resolution of a low resolution image. In this study, we use a method for enhancing the solar image resolution using a Deep-learning model which generates a high resolution HMI image from a low resolution HMI image (4 by 4 binning). Deep learning networks try to find the hidden equation between low resolution image and high resolution image from given input and the corresponding output image. In this study, we trained a model based on a very deep residual channel attention networks (RCAN) with HMI images in 2014 and test it with HMI images in 2015. We find that the model achieves high quality results in view of both visual and measures: 31.40 peak signal-to-noise ratio(PSNR), Correlation Coefficient (0.96), Root mean square error (RMSE) is 0.004. This result is much better than the conventional bi-cubic interpolation. We will apply this model to full-resolution SDO/HMI and GST magnetograms.

• Journal of Ocean Engineering and Technology
• /
• v.34 no.6
• /
• pp.442-453
• /
• 2020

Underwater color images suffer from low visibility and color cast effects caused by light attenuation by water and floating particles. This study applied single image enhancement techniques to enhance the quality of underwater images and compared their performance with real underwater images taken in Korean waters. Dark channel prior (DCP), gradient transform, image fusion, and generative adversarial networks (GAN), such as cycleGAN and underwater GAN (UGAN), were considered for single image enhancement. Their performance was evaluated in terms of underwater image quality measure, underwater color image quality evaluation, gray-world assumption, and blur metric. The DCP saturated the underwater images to a specific greenish or bluish color tone and reduced the brightness of the background signal. The gradient transform method with two transmission maps were sensitive to the light source and highlighted the region exposed to light. Although image fusion enabled reasonable color correction, the object details were lost due to the last fusion step. CycleGAN corrected overall color tone relatively well but generated artifacts in the background. UGAN showed good visual quality and obtained the highest scores against all figures of merit (FOMs) by compensating for the colors and visibility compared to the other single enhancement methods.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.1525-1527
• /
• 2013

본 논문에서는 저시력자의 개선된 독서 환경을 제공하는 시각보조기기를 위한 실시간 영상처리 SoC(System on Chip) 하드웨어 구조 설계에 대해서 기술한다. 기존의 시각보조기기는 화면 영상이 실제 움직임보다 늦게 출력되는 잔상 현상이 발생하며, 색 변환 기능도 제한적이다. 따라서 본 논문에서 제안하는 실시간 영상처리 SoC 하드웨어 구조는 데이터 연산을 최소화함으로써 잔상 현상이 감소되며, 저시력자를 위한 다양한 색상 모드를 지원한다. 제안하는 영상처리 SoC 하드웨어 구조는 Core-A 모듈, Memory Controller 모듈, AMBA AHB bus 모듈, ISP(Image Signal Processing) 모듈, TFT-LCD Controller 모듈, VGA Controller 모듈, CIS Controller 모듈, UART 모듈, Block Memory 모듈로 구성된다. 시각보조기기를 위한 실시간 영상처리 SoC 하드웨어 구조는 Virtex4 XC4VLX80 FPGA 디바이스를 이용하여 검증하였으며, TSMC 180nm 셀 라이브러리로 합성한 결과 동작주파수는 54MHz, 게이트 수 197k이다.

• Applied Microscopy
• /
• v.50
• /
• pp.23.1-23.9
• /
• 2020

We propose an effective deep learning model to denoise scanning transmission electron microscopy (STEM) image series, named Noise2Atom, to map images from a source domain 𝓢 to a target domain 𝓒, where 𝓢 is for our noisy experimental dataset, and 𝓒 is for the desired clear atomic images. Noise2Atom uses two external networks to apply additional constraints from the domain knowledge. This model requires no signal prior, no noise model estimation, and no paired training images. The only assumption is that the inputs are acquired with identical experimental configurations. To evaluate the restoration performance of our model, as it is impossible to obtain ground truth for our experimental dataset, we propose consecutive structural similarity (CSS) for image quality assessment, based on the fact that the structures remain much the same as the previous frame(s) within small scan intervals. We demonstrate the superiority of our model by providing evaluation in terms of CSS and visual quality on different experimental datasets.

• Journal of the Korean Society for Precision Engineering
• /
• v.13 no.12
• /
• pp.30-37
• /
• 1996

On-line process control equipment for CO$_{2}$ laser cutting is not available for industrial applications. The major part of the industrial laser cutting machines are adjusted off-line by highly educated engineers. The quality inspection of the sample is visual and referred to different quality scales. Due to the lack of automation the potential laser users hesitate to implement the cutting method. The first step toward an automation of the process is the development of a process monitoring system and the research is cincentrated on the area of on-line quality monitoring during CO$_{2}$ laser cutting. The method is based on the detection of the emitted light from the cutting front by photo diode. The developed monitoring system consists of the OP Amplifier, A/D convertor, power supply and PC. The signal from the photo diode has been undertaken from Fourier analysis and statistical analysis with real time. The photograph of striation pattern was taken by metallurgical microscope. As a result, it is possible to predict the striation pattern according to the beam traveling speed.

• International journal of advanced smart convergence
• /
• v.12 no.1
• /
• pp.92-100
• /
• 2023

Wireless communication technology is becoming increasingly prevalent in smart factories, but the rise in the number of wireless devices can lead to interference in the ISM band and obstacles like metal blocks within the factory can weaken communication signals, creating radio shadow areas that impede information exchange. Consequently, accurately determining the radio communication coverage range is crucial. To address this issue, a Radio Environment Map (REM) can be used to provide information about the radio environment in a specific area. In this paper, a technique for estimating an indoor REM usinga mobile robot and machine learning methods is introduced. The mobile robot first collects and processes data, including the Received Signal Strength Indicator (RSSI) and location estimation. This data is then used to implement the REM through machine learning regression algorithms such as Extra Tree Regressor, Random Forest Regressor, and Decision Tree Regressor. Furthermore, the numerical and visual performance of REM for each model can be assessed in terms of R2 and Root Mean Square Error (RMSE).

• Journal of Ecology and Environment
• /
• v.46 no.1
• /
• pp.18-30
• /
• 2022

Background: Bees and flowering plants associations were initially began during the early Cretaceous, 120 million years ago. This coexistence has led to a mutual relationship where the plant serves as food and in return, the bee help them their reproduction. Animals pollinate about 75% of food crops worldwide, with bees as the world's primary pollinator. In general, bees rely on flower scents to locate blooming flowers as visual clue is limited and also their host plants from a distance. In this review, an attempt is made to collect some relevant 107 published papers from three scientific databases, Google Scholar, Scopus, and Web of Science database, covering the period from 1959 to 2021. Results: Flowering plants are well documented to actively emit volatile organic compounds (VOCs). However, only a few of them are important for eliciting behavioral responses in bees. In this review, fifty-three volatile organic compounds belonging to different class of compounds, mainly terpenoids, benzenoids, and volatile fatty acid derivatives, is compiled here from floral scents that are responsible for eliciting behavioral responses in bees. Bees generally use honest floral signals to locate their host plants with nectar and pollen-rich flowers. Thus, honest signaling mechanism plays a key role in maintaining mutualistic plant-pollinator associations. Conclusions: Considering the fact that floral scents are the primary attractants, understanding and identification of VOCs from floral scent in plant-pollinator networks are crucial to improve crop pollination. Interestingly, current advances in both VOCs scent gene identification and their biosynthetic pathways make it possible to manipulate particular VOCs in plant, and this eventually may lead to increase in crop productivity.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.29-32
• /
• 2020

Due to the lack of improper image acquisition process, noise induction is an inevitable step. As a result, objective image quality assessment (IQA) plays an important role in estimating the visual quality of noisy image. Plenty of IQA methods have been proposed including traditional signal processing based methods as well as current deep learning based methods where the later one shows promising performance due to their complex representation ability. The deep learning based methods consists of several convolution layers and down sampling layers for feature extraction and fully connected layers for regression. Usually, the down sampling is performed by using max-pooling layer after each convolutional block. We reveal that this max-pooling causes information loss despite of knowing their importance. Consequently, we propose a better IQA method that replaces the max-pooling layers with strided convolutions to down sample the feature space and since the strided convolution layers have learnable parameters, they preserve optimal features and discard redundant information, thereby improve the prediction accuracy. The experimental results verify the effectiveness of the proposed method.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.2
• /
• pp.70.4-70.4
• /
• 2019

Image super-resolution (SR) is a technique that enhances the resolution of a low resolution image. In this study, we use three SR models (RCAN, ProSRGAN and Bicubic) for enhancing solar SDO/HMI magnetograms using deep learning. Each model generates a high resolution HMI image from a low resolution HMI image (4 by 4 binning). The pixel resolution of HMI is about 0.504 arcsec. Deep learning networks try to find the hidden equation between low resolution image and high resolution image from given input and the corresponding output image. In this study, we trained three models with HMI images in 2014 and test them with HMI images in 2015. We find that the RCAN model achieves higher quality results than the other two methods in view of both visual aspects and metrics: 31.40 peak signal-to-noise ratio(PSNR), Correlation Coefficient (0.96), Root mean square error (RMSE) is 0.004. This result is also much better than the conventional bi-cubic interpolation. We apply this model to a full-resolution SDO/HMI image and compare the generated image with the corresponding Hinode NFI magnetogram. As a result, we get a very high correlation (0.92) between the generated SR magnetogram and the Hinode one.