• FLOWER RESEARCH JOURNAL
• /
• v.18 no.4
• /
• pp.284-290
• /
• 2010

This study was carried out to breed a self-compatible variety of Tiger Lily from the seedlings originated from Chinju city, Gyeongsangnam-do. The morphological characteristics, ploidy, and the resistance of seedlings to virus infection were investigated. A progeny test was also conducted to examine whether the propagated progenies had the same characteristics as the mother plant. The self-compatible diploid lily variety developed in this study showed a tall type like native triploid lily, Lilium tigrinum, and bulbils were formed on both lily varieties. The morphological characteristics of the flowers and leaves were not much different between the varieties, while the size of the bulbils and bulbs showed significant differences. The percentage of seed generation by self-pollination was 72.6% for the self-compatible variety, while there was no seed generated for native triploid lily. The number of chromosome was 2n = 26 (x = 13, diploid) for the self-compatible variety, while 2n = 39 (x = 13, triploid) in native lily variety. The progenies of the self-compatible diploid lily variety showed the same characteristics as those of its mother plant in morphology, seed germination, and polyploidy. The mother plant of the self-compatible diploid lily variety showed 58% pollen germination and the 2-year-old and 3-year-old progenies showed similar germination percentages. The pollen grains of Korean native triploid lily, however, never germinated.

• Korean Journal of Remote Sensing
• /
• v.35 no.5_1
• /
• pp.681-692
• /
• 2019

In this paper, a hybrid deep learning model, called 2D convolution with bidirectional long short-term memory (2DCBLSTM), is presented that can effectively combine both spatial and temporal features for crop classification. In the proposed model, 2D convolution operators are first applied to extract spatial features of crops and the extracted spatial features are then used as inputs for a bidirectional LSTM model that can effectively process temporal features. To evaluate the classification performance of the proposed model, a case study of crop classification was carried out using multi-temporal unmanned aerial vehicle images acquired in Anbandegi, Korea. For comparison purposes, we applied conventional deep learning models including two-dimensional convolutional neural network (CNN) using spatial features, LSTM using temporal features, and three-dimensional CNN using spatio-temporal features. Through the impact analysis of hyper-parameters on the classification performance, the use of both spatial and temporal features greatly reduced misclassification patterns of crops and the proposed hybrid model showed the best classification accuracy, compared to the conventional deep learning models that considered either spatial features or temporal features. Therefore, it is expected that the proposed model can be effectively applied to crop classification owing to its ability to consider spatio-temporal features of crops.

• Journal of Convergence for Information Technology
• /
• v.11 no.11
• /
• pp.57-65
• /
• 2021

We propose a novel method to detect abnormal data of specific symptoms using deep learning in air pollution measurement system. Existing methods generally detect abnomal data by classifying data showing unusual patterns different from the existing time series data. However, these approaches have limitations in detecting specific symptoms. In this paper, we use DeepLab V3+ model mainly used for foreground segmentation of images, whose structure has been changed to handle one-dimensional data. Instead of images, the model receives time-series data from multiple sensors and can detect data showing specific symptoms. In addition, we improve model's performance by reducing the complexity of noisy form time series data by using 'piecewise aggregation approximation'. Through the experimental results, it can be confirmed that anomaly data detection can be performed successfully.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_3
• /
• pp.1405-1423
• /
• 2021

Precipitation is one of the main factors that affect water and energy cycles, and its estimation plays a very important role in securing water resources and timely responding to water disasters. Satellite-based quantitative precipitation estimation (QPE) has the advantage of covering large areas at high spatiotemporal resolution. In this study, machine learning-based rainfall intensity models were developed using Himawari-8 Advanced Himawari Imager (AHI) water vapor channel (6.7 ㎛), infrared channel (10.8 ㎛), and weather radar Column Max (CMAX) composite data based on random forest (RF). The target variables were weather radar reflectivity (dBZ) and rainfall intensity (mm/hr) converted by the Z-R relationship. The results showed that the model which learned CMAX reflectivity produced the Critical Success Index (CSI) of 0.34 and the Mean-Absolute-Error (MAE) of 4.82 mm/hr. When compared to the GeoKompsat-2 and Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks (PERSIANN)-Cloud Classification System (CCS) rainfall intensity products, the accuracies improved by 21.73% and 10.81% for CSI, and 31.33% and 23.49% for MAE, respectively. The spatial distribution of the estimated rainfall intensity was much more similar to the radar data than the existing products.

• Korean Journal of Environment and Ecology
• /
• v.35 no.5
• /
• pp.503-524
• /
• 2021

A largemouth bass (Micropterus salmoides) is an ecosystem disturbance fish species at the highest rank in the aquatic ecosystem, causing a serious imbalance in freshwater ecosystems. Although various attempts have been made to eradicate and control largemouth bass, no effective measures were found. Therefore, it is necessary to find an approach to maximize the effective population reduction based on the unique characteristics of largemouth bass. This study used the transcriptome analysis to derive 182,887 unigene contigs and select 12 types of final target sequences for applying the CRISPR/Cas9 system in the genes of IZUMO1 and Zona pellucida sperm-binding protein, which are proteins involved in sperm-egg recognition. After synthesizing 12 types of sgRNA capable of recognizing each target sequence, 12 types of Cas9-sgRNA ribonucleoprotein (RNP) complexes to be used in subsequent studies were prepared. This study searched the protein-coding gene of sperm-egg through the Next Generation Sequencing (NGS) and edited genes through the CRISPR/Cas9 system to induce infertile individuals that produced reproductive cells but could not form fertilized eggs. Through such a series of processes, it successfully established a composition development process for largemouth bass. It is judged that this study contributed to securing the valuable basic data for follow-up studies to verify its effect for the management of ecological disturbances without affecting the habitat of other endemic species in the same water system with the largemouth bass.

• Korean Journal of Remote Sensing
• /
• v.37 no.6_3
• /
• pp.1965-1974
• /
• 2021

Satellite remote sensing approach can be actively used for forest monitoring. Especially, it is much meaningful to utilize Korea multi-purpose satellites, an independently operated satellite in Korea, for forest monitoring of Korea, Recently, several studies have been performed to exploit meaningful information from satellite remote sensed data via machine learning approaches. The forest information produced through machine learning approaches can be used to support the efficiency of traditional forest monitoring methods, such as in-situ survey or qualitative analysis of aerial image. The performance of machine learning approaches is greatly depending on the characteristics of study area and data. Thus, it is very important to survey the best model among the various machine learning models. In this study, the performance of deep neural network to classify artificial or natural forests was analyzed in Samcheok, Korea. As a result, the pixel accuracy was about 0.857. F1 scores for natural and artificial forests were about 0.917 and 0.433 respectively. The F1 score of artificial forest was low. However, we can find that the artificial and natural forest classification performance improvement of about 0.06 and 0.10 in F1 scores, compared to the results from single layered sigmoid artificial neural network. Based on these results, it is necessary to find a more appropriate model for the forest type classification by applying additional models based on a convolutional neural network.

• Korean Journal of Remote Sensing
• /
• v.34 no.6_4
• /
• pp.1555-1565
• /
• 2018

Detection of subtle ground deformation of volcanoes plays an important role in evaluating the risk and possibility of volcanic eruptions. Ground-fixed observation equipment is difficult to maintain and cost-inefficient. In contrast, satellite remote sensing can regularly monitor at low cost. In this paper, following the study of Chadwick et al. (2006), which applied the interferometric SAR (InSAR) technique to the Sierra Negra volcano, Galapagos. In order to investigate the deformation of the volcano before 2005 eruption, the recent activities of this volcano were analyzed using Sentinel-1, the latest SAR satellite. We obtained the descending mode Sentinel-1A SAR data from January 2017 to January 2018, applied the Persistent Scatter InSAR, and estimated the depth and expansion quantity of magma in recent years through the Mogi model. As a result, it was confirmed that the activity pattern of volcano prior to the eruption in June 2018 was similar to the pattern before the eruption in 2005 and was successful in estimating the depth and expansion amount. The results of this study suggest that satellite SAR can characterize the activity patterns of volcano and can be possibly used for early monitoring of volcanic eruption.

• Korean Journal of Food Science and Technology
• /
• v.51 no.3
• /
• pp.221-226
• /
• 2019

This study investigated the optimal amount of shell powder (0.1, 0.2, 0.3, and 0.4%) to add to rice noodles containing 20% rice flour and compared their quality characteristics to those of wheat noodles containing a commercial alkaline reagent (added at 0.4%). As the amount of shell powder was increased, the L and b values (Hunter's color) increased. The pH, turbidity, and water absorption also increased as the amount of shell powder was increased. However, when the shell powder content exceeded 0.3%, the hardness, chewiness, springiness, and tension tended to decrease below acceptable levels. This might be because the shell powder inhibited network formation. The textural properties and pH value of rice noodles containing 0.2% shell powder were similar to those of the wheat noodles. This suggests that 0.2% shell powder may be the optimal amount to add to rice noodles when used as the alkaline reagent.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.10
• /
• pp.397-402
• /
• 2019

Accurately predicting cancer prognosis to provide appropriate treatment strategies for patients is one of the critical challenges in bioinformatics. Many researches have suggested machine learning models to predict patients' outcomes based on their gene expression data. Gene expression data is high-dimensional numerical data containing about 17,000 genes, so traditional researches used feature selection or dimensionality reduction approaches to elevate the performance of prognostic prediction models. These approaches, however, have an issue of making it difficult for the predictive models to grasp any biological interaction between the selected genes because feature selection and model training stages are performed independently. In this paper, we propose a novel two-dimensional image formatting approach for gene expression data to achieve feature selection and prognostic prediction effectively. Node2Vec is exploited to integrate biological interaction network and gene expression data and a convolutional neural network learns the integrated two-dimensional gene expression image data and predicts cancer prognosis. We evaluated our proposed model through double cross-validation and confirmed superior prognostic prediction accuracy to traditional machine learning models based on raw gene expression data. As our proposed approach is able to improve prediction models without loss of information caused by feature selection steps, we expect this will contribute to development of personalized medicine.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.26 no.7
• /
• pp.904-914
• /
• 2020

Through the Ministry of Trade, Industry, and Energy, South Korea is trying to support the "Building Project for Liquified Natural Gas (LNG) Bunkering Ship," centered on the Korea Gas Corporation, while the Ministry of Maritime Af airs and Fisheries is pushing to construct an LNG bunkering terminal at Busan New Port. LNG bunkering ships are essential for supplying LNG fuel from the terminal to the ships, resulting in the need for safety operation procedures. Therefore, in this study, the stability of a coastal LNG bunkering ship operating from Busan New Port to the anchorage in Busan Port was assessed to investigate the need for operational procedures for coastal LNG bunkering ships. Seakeeping analysis of the LNG bunkering ship was performed for each significant wave height by combining the response amplitude operator from the ship motion analysis under the potential flow theory with the actual observed sea data for five years and Texel, Marsen, and Arsloe (TMA) spectrum suitable for the Busan coast. The results showed that the roll and horizontal acceleration were the main risks that affected the navigation seakeeping performance above a significance wave height of 2 m. The operational periods of the LNG bunkering ship ranged from 83.3% to 99.9% of the total observation period.