• Journal of Korea Multimedia Society
• /
• v.25 no.1
• /
• pp.1-9
• /
• 2022

Ship detection in synthetic aperture radar (SAR) images is an important application in marine monitoring for the military and civilian domains. Over the past decade, object detection has achieved significant progress with the development of convolutional neural networks (CNNs) and lot of labeled databases. However, due to difficulty in collecting and labeling SAR images, it is still a challenging task to solve SAR ship detection CNNs. To overcome the problem, some methods have employed conventional data augmentation techniques such as flipping, cropping, and affine transformation, but it is insufficient to achieve robust performance to handle a wide variety of types of ships. In this paper, we present a novel and effective approach for deep SAR ship detection, that exploits label-rich Electro-Optical (EO) images. The proposed method consists of two components: a data augmentation network and a ship detection network. First, we train the data augmentation network based on conditional generative adversarial network (cGAN), which aims to generate additional SAR images from EO images. Since it is trained using unpaired EO and SAR images, we impose the cycle-consistency loss to preserve the structural information while translating the characteristics of the images. After training the data augmentation network, we leverage the augmented dataset constituted with real and translated SAR images to train the ship detection network. The experimental results include qualitative evaluation of the translated SAR images and the comparison of detection performance of the networks, trained with non-augmented and augmented dataset, which demonstrates the effectiveness of the proposed framework.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.32 no.spc4_2
• /
• pp.401-412
• /
• 2014

This study is purposed on showing an effective process for coordinate transformation of cadastral maps, based on the local datum, in the World Geodetic System (WGS) for Cadastral Reform Project (CRP). The process follows three steps: coordinate adjustment, datum transformation and distortion modeling. The first procedure is that point coordinates on local datum has to be adjusted by those GPS observed point-to-point distances, using trilateration. Secondly, the adjusted coordinates need to be transformed to WGS by applying the Affine model, while the verification of the methodology is implemented under numerical experiments. To conduct this procedure, 195 points in the same coordinate origins in Seoul and 61 points in several different origins in Incheon are used in the estimation. As a result, there are less than 2cm coordinate differences between transformed coordinates and measured ones at everywhere. Also, it is remarkable that the transformation does not depend on either of the particular common points or the sizes of computed region. Therefore, this suggested methodology is expected to easily provide identifications and corrections for points-deviations for improved quality of the cadastral map by distortion modeling through CRP.

• The Korean Journal of Pesticide Science
• /
• v.8 no.3
• /
• pp.220-230
• /
• 2004

This study was conducted of Korean native plants to screen herbicidal activity which could be used for the development of new natural herbicides. Ninety-eight plants were collected from Wan Island, Chollanamdo in Korea and their methanol extracts were obtained. Herbicidal activities of the methanol extracts were determined by seed bioassay using canola (Brassica napus L.) seedlings. Among ninety-eight species, twenty plants were highly herbicidal ($GR_{50}<1,000\;{\mu}g\;g^{-1}$): Abies holophylla MAXIM., Ailanthus altissima (MILL.) SWINGLE, Anthemis nobilis L., Aralia elata SEEM., Artemisia iwayomogi KITAMURA, Asarum sieboldii MIQ., Brassica campestris subsp. napus var. nippo-oleifera MAKINO, Clematis terniflora DC., Crataegus scabrida SARG., Gnaphalium affine D. DON, Jasminum nudiflorum LINDL., Kalopanax pictus (THUNE.) NAKAI, Machilus japonica S. et Z., Myrica rubra S. et Z., Osmunda japonica THUNB., Phytolacca esculenta V. Houtte, Platanus occidentalis L., Quisqualis indica L., Rubus hirsutus THUNB., Yucca smalliana FERN. Fifty plants were shown moderate herbicidal activity $(1,000\;{\mu}g\;g^{-1}, however, twenty-eight plants were not shown any herbicidal activity.

• The Korean Journal of Mycology
• /
• v.25 no.4 s.83
• /
• pp.354-361
• /
• 1997

The antifungal activities of 153 extracts from 51 species of Korean mushroom on six phytopathogenic fungi were investigated. The powder of fruit-body of each mushroom was extracted with petroleum ether, 80% ethanol and distilled water. The water extracts of four mushrooms including Amanita virosa showed antifungal activities on Alternaria alternata, Colletotrichum graminicola, Fusarium solani, Idriella bolleyie. The 80% ethanol extracts of seven mushrooms including Boletus auripes showed antibiotic activities against A. alternata, C. gramincola, Cylindrocarpon destructans, F. solani, F. oxysporum cucumerinum, and I. bolleyie. The petroleum ether extracts of six mushrooms including Amanita citrina showed antibiotic activities against A. alternata, C. destructans, and C. graminicola. The 102 extracts of 34 mushrooms including Agaricus arvensis didn't show antibiotic activities.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.45 no.9
• /
• pp.711-723
• /
• 2017

This paper proposes a flight control system for an organic flight array(OFA) which has a new configuration to consist of multi modularized ducted-fan unmanned aerial vehicles (UAVs). The OFA is able to apply to various missions such as indoor reconnaissance, communication relay, and radar jamming by using capability of hover flight. The OFA has a distinguished advantage due to reconfigurable structure to assemble or separate with respect to its missions or operational conditions. A dynamic modelling of the OFA is derived based on equations of motion of the single ducted-fan modules. In order to apply nonlinear control method, an affine system of attitude dynamics is derived. Moreover, the control system is composed of a back-stepping controller for attitude control and a PID controller for position control. Then the performance of the proposed controller is verified via a numerical simulation under wind disturbance.

• KDI Journal of Economic Policy
• /
• v.36 no.2
• /
• pp.65-101
• /
• 2014

This paper estimates and characterizes expected inflations using an affine term structure model based on the empirical stochastic process of the interest rates in Korea. The empirical results show that the expected inflation which marked above 4% before the global financial crisis has dampened and stabilized after the crisis. Moreover, we investigate the rationality of the various expected inflation measures in terms of the unbiasedness and efficiency and find that unbiasedness is not rejected across the all measures, while the efficiency cannot be empirically warranted. Besides, we run Granger causality tests and conclude that the expected inflations compiled from the Consensus, BOK-Expert have the cross-causality with the long-run actual inflation, while the expected inflation estimated from the term structure model has the cross-causality with the short-run actual inflation. These results connote that expected inflations collected from different sources and methods have their targets and horizons and the central bank needs to watch all of them with a balanced view instead of preferring one to the other.

• Progress in Medical Physics
• /
• v.21 no.2
• /
• pp.153-164
• /
• 2010

To determine the clinical target volumes considering vascularity and cellularity of tumors, the software was developed for mapping of the analyzed biological clinical target volumes on anatomical images using regional cerebral blood volume (rCBV) maps and apparent diffusion coefficient (ADC) maps. The program provides the functions for integrated registrations using mutual information, affine transform and non-rigid registration. The registration accuracy is evaluated by the calculation of the overlapped ratio of segmented bone regions and average distance difference of contours between reference and registered images. The performance of the developed software was tested using multimodal images of a patient who has the residual tumor of high grade gliomas. Registration accuracy of about 74% and average 2.3 mm distance difference were calculated by the evaluation method of bone segmentation and contour extraction. The registration accuracy can be improved as higher as 4% by the manual adjustment functions. Advanced MR images are analyzed using color maps for rCBV maps and quantitative calculation based on region of interest (ROI) for ADC maps. Then, multi-parameters on the same voxels are plotted on plane and constitute the multi-functional parametric maps of which x and y axis representing rCBV and ADC values. According to the distributions of functional parameters, tumor regions showing the higher vascularity and cellularity are categorized according to the criteria corresponding malignant gliomas. Determined volumes reflecting pathological and physiological characteristics of tumors are marked on anatomical images. By applying the multi-functional images, errors arising from using one type of image would be reduced and local regions representing higher probability as tumor cells would be determined for radiation treatment plan. Biological tumor characteristics can be expressed using image registration and multi-functional parametric maps in the developed software. The software can be considered to delineate clinical target volumes using advanced MR images with anatomical images.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.26 no.2
• /
• pp.28-36
• /
• 2022

As port structures are exposed to various extreme external loads such as wind (typhoons), sea waves, or collision with ships; it is important to evaluate the structural safety periodically. To monitor the port structure, especially the rubber fender, a fender segmentation system using a vision sensor and deep learning method has been proposed in this study. For fender segmentation, a new deep learning network that improves the encoder-decoder framework with the receptive field block convolution module inspired by the eccentric function of the human visual system into the DenseNet format has been proposed. In order to train the network, various fender images such as BP, V, cell, cylindrical, and tire-types have been collected, and the images are augmented by applying four augmentation methods such as elastic distortion, horizontal flip, color jitter, and affine transforms. The proposed algorithm has been trained and verified with the collected various types of fender images, and the performance results showed that the system precisely segmented in real time with high IoU rate (84%) and F1 score (90%) in comparison with the conventional segmentation model, VGG16 with U-net. The trained network has been applied to the real images taken at one port in Republic of Korea, and found that the fenders are segmented with high accuracy even with a small dataset.

• Korean Journal of Veterinary Research
• /
• v.35 no.2
• /
• pp.327-336
• /
• 1995

This study was carried out to develop the animal protein(feed for fry) that was isolated, purified and lyophilized the rumen ciliates from the healthy rumen contents which have $10^5-10^6/g$ ciliates and were discarded in abattoirs. The rumen ciliates are non-pathogenic, anaerobic and the weight of this protozoa is 2% of rumen content. The rumen protozoan and bacterial proteins both have a biological value for rats of 80-81, which is higher than the 72 of brewer's yeasts. Furthermore, the true digestibility and net protein utility of the protozoan protein are 91 and 73, much higher than those of bacterial(74 and 60) or yeast(84 and 60) proteins. The amino acids of rumen protozoa is nutritionally superior than the others. The size of rumen ciliates is $30-200{\times}20-110{\mu}m$ and so we had isolated and purified the rumen ciliates from the rumen contents by the physical methods. The purified rumen protozoa was lyophilized with freezing dryer. The results of this experiment were as follows : 1. Population dynamics of protozoan ciliates in slaughtered rumens; % of samples which small ciliates were predominated was 82.5%(52/63) and that of large ciliates was 17.5%(11/63). 1) predominant species of small ciliates were Entodinium ovinum and E nanellum. 2) predominant species of large ciliates were Epidinium ecaudatum and Diploplastron affine. 2. The lyophilized rumen ciliates which were isolated and purified from 1 kg of rumen content at the pH 6.2-6.8 was about 7.0 gram. 3. The nutrient analysis of lyophilized rqmen ciliates(LRC) was as follows: 1) Proximate analysis of the LRC and the composition of fry feed; moisture 8.05%(below 10.0), protein 35.37%(45), fat 5.39%(4.5), fiber 1.23%(below 2.5), ash 2.25%(below 15.0), Ca 0.26%(below 2.0), P 0.14%(below 1.1), energy 4,608.11(fish meal 5000 cal/g) 2) Amino acids (% in crude protein) of the LRC and the rotifer(Brachionus plicatilis); Arg 5.19%(4.50), His 2.50%(1.55), Ile 5.29%(3.45), Leu 8.11%(5.85), Lys 10.34%(6.15), Met 2.25% (0.85), Phe 5.66%(3.80), Thr 5.14% (3.45), Val 4.18%(3.90), Ala 4.13%(3.35), Asp 13.26%(8.25), Glu 16.62%(9.20), Gly 4.23%(3.10), Pro 3.25%(5.05), Ser 4.85%(3.85), Tyr 5.04%(3.05) 3) Fatty acids(% in fat) of the LRC and the rotifer(biological feed ; Brachionus plicatilis); myristic acid(C14:0) 3.27%(0.3), myristoleic acid(C14:1) 0.83%(-), palmitic acid(C16:0) 39.11% (23.5), palmitoleic acid(C16:1) 2.81%(2.0), stearic acid(C18:0) 9.36%(5.6), oleic acid(C18:1) 25.54%(3.5), linoleic acid(C18:2) 15.05%(32.9), linolenic acid(C18:3) 1.74%(9.8). Judging from the above investigated results, the analytical data of proximate analysis, amino acids, fatty acids of the purified and lyophilized rumen protozoa are reasonable for the feed of freshwater fishes(fry and fingerling). But it was disappointed of our expectation that the crude protein of lyophilized rumen ciliates contains low percentage, it was thought that because of the small ciliates(starch digester) in beef cattle rumens which were administered the concentrated feed, is much difficult to isolate and purify than the large ciliates(fiber digester).