• Economic and Environmental Geology
• /
• v.55 no.5
• /
• pp.551-561
• /
• 2022

Geophysical exploration methods are very useful for generating high-resolution images of underground structures, and such methods can be applied to investigation of buried cultural properties and for determining their exact locations. In this study, image feature extraction and image segmentation methods were applied to automatically distinguish the structures of buried relics from the high-resolution ground-penetrating radar (GPR) images obtained at the center of Silla Kingdom, Gyeongju, South Korea. The major purpose for image feature extraction analyses is identifying the circular features from building remains and the linear features from ancient roads and fences. Feature extraction is implemented by applying the Canny edge detection and Hough transform algorithms. We applied the Hough transforms to the edge image resulted from the Canny algorithm in order to determine the locations the target features. However, the Hough transform requires different parameter settings for each survey sector. As for image segmentation, we applied the connected element labeling algorithm and object-based image analysis using Orfeo Toolbox (OTB) in QGIS. The connected components labeled image shows the signals associated with the target buried relics are effectively connected and labeled. However, we often find multiple labels are assigned to a single structure on the given GPR data. Object-based image analysis was conducted by using a Large-Scale Mean-Shift (LSMS) image segmentation. In this analysis, a vector layer containing pixel values for each segmented polygon was estimated first and then used to build a train-validation dataset by assigning the polygons to one class associated with the buried relics and another class for the background field. With the Random Forest Classifier, we find that the polygons on the LSMS image segmentation layer can be successfully classified into the polygons of the buried relics and those of the background. Thus, we propose that these automatic classification methods applied to the GPR images of buried cultural heritage in this study can be useful to obtain consistent analyses results for planning excavation processes.

• Journal of Korea Water Resources Association
• /
• v.57 no.7
• /
• pp.437-449
• /
• 2024

Rainfall-runoff prediction studies using deep learning while considering catchment attributes have been gaining attention. In this study, we selected two models: the Transformer model, which is suitable for large-scale data training through the self-attention mechanism, and the LSTM-based multi-state-vector sequence-to-sequence (LSTM-MSV-S2S) model with an encoder-decoder structure. These models were constructed to incorporate catchment attributes and predict the inflow of 10 multi-purpose dam watersheds in South Korea. The experimental design consisted of three training methods: Single-basin Training (ST), Pretraining (PT), and Pretraining-Finetuning (PT-FT). The input data for the models included 10 selected watershed attributes along with meteorological data. The inflow prediction performance was compared based on the training methods. The results showed that the Transformer model outperformed the LSTM-MSV-S2S model when using the PT and PT-FT methods, with the PT-FT method yielding the highest performance. The LSTM-MSV-S2S model showed better performance than the Transformer when using the ST method; however, it showed lower performance when using the PT and PT-FT methods. Additionally, the embedding layer activation vectors and raw catchment attributes were used to cluster watersheds and analyze whether the models learned the similarities between them. The Transformer model demonstrated improved performance among watersheds with similar activation vectors, proving that utilizing information from other pre-trained watersheds enhances the prediction performance. This study compared the suitable models and training methods for each multi-purpose dam and highlighted the necessity of constructing deep learning models using PT and PT-FT methods for domestic watersheds. Furthermore, the results confirmed that the Transformer model outperforms the LSTM-MSV-S2S model when applying PT and PT-FT methods.

• Journal of Intelligence and Information Systems
• /
• v.25 no.3
• /
• pp.43-62
• /
• 2019

At one time, the anomaly detection sector dominated the method of determining whether there was an abnormality based on the statistics derived from specific data. This methodology was possible because the dimension of the data was simple in the past, so the classical statistical method could work effectively. However, as the characteristics of data have changed complexly in the era of big data, it has become more difficult to accurately analyze and predict the data that occurs throughout the industry in the conventional way. Therefore, SVM and Decision Tree based supervised learning algorithms were used. However, there is peculiarity that supervised learning based model can only accurately predict the test data, when the number of classes is equal to the number of normal classes and most of the data generated in the industry has unbalanced data class. Therefore, the predicted results are not always valid when supervised learning model is applied. In order to overcome these drawbacks, many studies now use the unsupervised learning-based model that is not influenced by class distribution, such as autoencoder or generative adversarial networks. In this paper, we propose a method to detect anomalies using generative adversarial networks. AnoGAN, introduced in the study of Thomas et al (2017), is a classification model that performs abnormal detection of medical images. It was composed of a Convolution Neural Net and was used in the field of detection. On the other hand, sequencing data abnormality detection using generative adversarial network is a lack of research papers compared to image data. Of course, in Li et al (2018), a study by Li et al (LSTM), a type of recurrent neural network, has proposed a model to classify the abnormities of numerical sequence data, but it has not been used for categorical sequence data, as well as feature matching method applied by salans et al.(2016). So it suggests that there are a number of studies to be tried on in the ideal classification of sequence data through a generative adversarial Network. In order to learn the sequence data, the structure of the generative adversarial networks is composed of LSTM, and the 2 stacked-LSTM of the generator is composed of 32-dim hidden unit layers and 64-dim hidden unit layers. The LSTM of the discriminator consists of 64-dim hidden unit layer were used. In the process of deriving abnormal scores from existing paper of Anomaly Detection for Sequence data, entropy values of probability of actual data are used in the process of deriving abnormal scores. but in this paper, as mentioned earlier, abnormal scores have been derived by using feature matching techniques. In addition, the process of optimizing latent variables was designed with LSTM to improve model performance. The modified form of generative adversarial model was more accurate in all experiments than the autoencoder in terms of precision and was approximately 7% higher in accuracy. In terms of Robustness, Generative adversarial networks also performed better than autoencoder. Because generative adversarial networks can learn data distribution from real categorical sequence data, Unaffected by a single normal data. But autoencoder is not. Result of Robustness test showed that he accuracy of the autocoder was 92%, the accuracy of the hostile neural network was 96%, and in terms of sensitivity, the autocoder was 40% and the hostile neural network was 51%. In this paper, experiments have also been conducted to show how much performance changes due to differences in the optimization structure of potential variables. As a result, the level of 1% was improved in terms of sensitivity. These results suggest that it presented a new perspective on optimizing latent variable that were relatively insignificant.

• Journal of Bio-Environment Control
• /
• v.13 no.4
• /
• pp.217-225
• /
• 2004

This study was performed to provide the basic knowledge about the mushroom cultivation facilities. Classified current status of cultivation facilities in Gyeongnam province was investigated by questionnaire. The structure of Pleurotus eryngii cultivation facilities can be classified into the simple and permanent frame type. The simple frame structures were mostly single-span type, on the other hand, the permanent frame structures were more multi-span than simple structures. And the scale of cultivation facilities was very different regardless of structural type. But as a whole, the length, width and ridge height were prevailing approximately 20.0 m, $6.6\~7.0m$ and $4.6\~5.0m$ range, respectively. The floor area was about $132\~160\;m^2$, and floor was built with concrete to protect mushrooms from various harmful infection. The roof slope of the simple and permanent type showed about $41.5^{\circ}\;and\;18.6\~28.6^{\circ}$, respectively. The width and layer number of growing bed for mushroom cultivation were around $1.2\~1.6m$, 4 layers in common, respectively. Most of year round cultivation facilities were equipped with cooler, heater, humidifier, and ventilating fan. Hot water boiler was the most commonly used heating system, the next was electric heater and then steam boiler. The industrial air conditioner has been widely used for cooling. And humidity was controlled mostly by ultra-wave or centrifuging humidifier. But some farmers has been using nozzle system for auxiliary purpose. More then $90\%$ of the mushroom house had the independent environment control system. The inside temperature was usually controlled by sensor, but humidity and $CO_2$ concentration was controlled by timer for each growing stage. The capacity of medium bottle was generally 850 cc and 1100cc, some farms used 800 cc, 950 co and 1,250 cc. Most of mushroom producted has been usually shipped to both circulating company and joint market.

• Horticultural Science & Technology
• /
• v.18 no.3
• /
• pp.353-359
• /
• 2000

'Yumyeong' is one of the most popular peach varieties in Korea. This study was conducted to monitor the developments of cells and tissues, and the changes in sugar contents during fruit growth and development. At bloom, there were two rows of vascular tissues, and the number and the position of internal vascular bundles were consistent during the fruit growth, however, the number of vascular tissues were increased and the distribution was irregular in the flesh tissues. The tissues between the inner integument and the internal vascular bundles showed different development characteristics from other parenchyma cells which were consisted of small and dense cells containing tannins. Such observation suggested that the stone of peach was consisted of inner epidermis and cells in the internal vascular tissues. The outer epidermis consisted of single layer cells at bloom differrentiated into 1-2 layers by horizontal cell divisions on 14 days after full bloom. On 30 days after full bloom, the epidermis was consisted of 5-6 layers by vertical cell divisions. The cell layers consisting the outer epidermis were gradually decreased to 1-2 layers at maturity. The observations on the changes in the epidermis confirmed that some of the cells consisting the hypodermis of peach fruit were originated from the cells of outer epidermis. Tylosis was observed from 35 days after full bloom and the size and number of tylosis were increased until full fruit maturity. The sucrose content showed the tendency of sharp increase from 50 days to 120 days after full bloom then decrease slightly. After when stone hardening ended, other solids showed a gradual decrease tendency from 80 days after full bloom.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2011.04a
• /
• pp.1-2
• /
• 2011

Hybrid rockets have lately attracted attention as a strong candidate of small, low cost, safe and reliable launch vehicles. A significant topic is that the first commercially sponsored space ship, SpaceShipOne vehicle chose a hybrid rocket. The main factors for the choice were safety of operation, system cost, quick turnaround, and thrust termination. In Japan, five universities including Hokkaido University and three private companies organized "Hybrid Rocket Research Group" from 1998 to 2002. Their main purpose was to downsize the cost and scale of rocket experiments. In 2002, UNISEC (University Space Engineering Consortium) and HASTIC (Hokkaido Aerospace Science and Technology Incubation Center) took over the educational and R&D rocket activities respectively and the research group dissolved. In 2008, JAXA/ISAS and eleven universities formed "Hybrid Rocket Research Working Group" as a subcommittee of the Steering Committee for Space Engineering in ISAS. Their goal is to demonstrate technical feasibility of lowcost and high frequency launches of nano/micro satellites into sun-synchronous orbits. Hybrid rockets use a combination of solid and liquid propellants. Usually the fuel is in a solid phase. A serious problem of hybrid rockets is the low regression rate of the solid fuel. In single port hybrids the low regression rate below 1 mm/s causes large L/D exceeding a hundred and small fuel loading ratio falling below 0.3. Multi-port hybrids are a typical solution to solve this problem. However, this solution is not the mainstream in Japan. Another approach is to use high regression rate fuels. For example, a fuel regression rate of 4 mm/s decreases L/D to around 10 and increases the loading ratio to around 0.75. Liquefying fuels such as paraffins are strong candidates for high regression fuels and subject of active research in Japan too. Nakagawa et al. in Tokai University employed EVA (Ethylene Vinyl Acetate) to modify viscosity of paraffin based fuels and investigated the effect of viscosity on regression rates. Wada et al. in Akita University employed LTP (Low melting ThermoPlastic) as another candidate of liquefying fuels and demonstrated high regression rates comparable to paraffin fuels. Hori et al. in JAXA/ISAS employed glycidylazide-poly(ethylene glycol) (GAP-PEG) copolymers as high regression rate fuels and modified the combustion characteristics by changing the PEG mixing ratio. Regression rate improvement by changing internal ballistics is another stream of research. The author proposed a new fuel configuration named "CAMUI" in 1998. CAMUI comes from an abbreviation of "cascaded multistage impinging-jet" meaning the distinctive flow field. A CAMUI type fuel grain consists of several cylindrical fuel blocks with two ports in axial direction. The port alignment shifts 90 degrees with each other to make jets out of ports impinge on the upstream end face of the downstream fuel block, resulting in intense heat transfer to the fuel. Yuasa et al. in Tokyo Metropolitan University employed swirling injection method and improved regression rates more than three times higher. However, regression rate distribution along the axis is not uniform due to the decay of the swirl strength. Aso et al. in Kyushu University employed multi-swirl injection to solve this problem. Combinations of swirling injection and paraffin based fuel have been tried and some results show very high regression rates exceeding ten times of conventional one. High fuel regression rates by new fuel, new internal ballistics, or combination of them require faster fuel-oxidizer mixing to maintain combustion efficiency. Nakagawa et al. succeeded to improve combustion efficiency of a paraffin-based fuel from 77% to 96% by a baffle plate. Another effective approach some researchers are trying is to use an aft-chamber to increase residence time. Better understanding of the new flow fields is necessary to reveal basic mechanisms of regression enhancement. Yuasa et al. visualized the combustion field in a swirling injection type motor. Nakagawa et al. observed boundary layer combustion of wax-based fuels. To understand detailed flow structures in swirling flow type hybrids, Sawada et al. (Tohoku Univ.), Teramoto et al. (Univ. of Tokyo), Shimada et al. (ISAS), and Tsuboi et al. (Kyushu Inst. Tech.) are trying to simulate the flow field numerically. Main challenges are turbulent reaction, stiffness due to low Mach number flow, fuel regression model, and other non-steady phenomena. Oshima et al. in Hokkaido University simulated CAMUI type flow fields and discussed correspondence relation between regression distribution of a burning surface and the vortex structure over the surface.