• Journal of the Korean Association of Geographic Information Studies
• /
• v.21 no.4
• /
• pp.64-80
• /
• 2018

In Korea, citizens can only know general information about crime. Thus it is difficult to know how much they are exposed to crime. If the police can predict the crime risky area, it will be possible to cope with the crime efficiently even though insufficient police and enforcement resources. However, there is no prediction system in Korea and the related researches are very much poor. From these backgrounds, the final goal of this study is to develop an automated crime prediction system. However, for the first step, we build a big data set which consists of local real crime information and urban physical or non-physical data. Then, we developed a crime prediction model through machine learning method. Finally, we assumed several possible scenarios and calculated the probability of crime and visualized the results in a map so as to increase the people's understanding. Among the factors affecting the crime occurrence revealed in previous and case studies, data was processed in the form of a big data for machine learning: real crime information, weather information (temperature, rainfall, wind speed, humidity, sunshine, insolation, snowfall, cloud cover) and local information (average building coverage, average floor area ratio, average building height, number of buildings, average appraised land value, average area of residential building, average number of ground floor). Among the supervised machine learning algorithms, the decision tree model, the random forest model, and the SVM model, which are known to be powerful and accurate in various fields were utilized to construct crime prevention model. As a result, decision tree model with the lowest RMSE was selected as an optimal prediction model. Based on this model, several scenarios were set for theft and violence cases which are the most frequent in the case city J, and the probability of crime was estimated by $250{\times}250m$ grid. As a result, we could find that the high crime risky area is occurring in three patterns in case city J. The probability of crime was divided into three classes and visualized in map by $250{\times}250m$ grid. Finally, we could develop a crime prediction model using machine learning algorithm and visualized the crime risky areas in a map which can recalculate the model and visualize the result simultaneously as time and urban conditions change.

• Journal of Nutrition and Health
• /
• v.51 no.6
• /
• pp.567-579
• /
• 2018

Purpose: This study examined the mediation effects of dietary habits, nutrient intake, daily life in the relationship between the working hours of Korean shift workers and metabolic syndrome. Methods: Data were collected from the sixth (2013-2015) Korea National Health and Nutrition Examination Survey (KNHANES). The stochastic regression imputation was used to fill missing data. Statistical analysis was performed in Korean shift workers with metabolic syndrome using the SPSS 24 program for Windows and a structural equation model (SEM) using an analysis of moment structure (AMOS) 21.0 package. Results: The model fitted the data well in terms of the goodness of fit index (GFI) = 0.939, root mean square error of approximation (RMSEA) = 0.025, normed fit index (NFI) = 0.917, Tucker-Lewis index (TLI) = 0.984, comparative fit index (CFI) = 0.987, and adjusted goodness of fit index (AGFI) = 0.915. Specific mediation effect of dietary habits (p = 0.023) was statistically significant in the impact of the working hours of shift workers on nutrient intake, and specific mediation effect of daily life (p = 0.019) was statistically significant in the impact of the working hours of shift workers on metabolic syndrome. On the other hand, the dietary habits, nutrient intake and daily life had no significant multiple mediator effects on the working hours of shift workers with metabolic syndrome. Conclusion: The appropriate model suggests that working hours have direct effect on the daily life, which has the mediation effect on the risk of metabolic syndrome in shift workers.

• The Journal of Korean Society for Radiation Therapy
• /
• v.31 no.1
• /
• pp.7-15
• /
• 2019

Purpose: The purpose of this study is to evaluate beam delivery accuracy for small sized lung SBRT through experiment. In order to assess the accuracy, Eclipse TPS(Treatment planning system) equipped Acuros XB and radiochromic film were used for the dose distribution. Comparing calculated and measured dose distribution, evaluated the margin for PTV(Planning target volume) in lung tissue. Materials and Methods : Acquiring CT images for Rando phantom, planned virtual target volume by size(diameter 2, 3, 4, 5 cm) in right lung. All plans were normalized to the target Volume=prescribed 95 % with 6MV FFF VMAT 2 Arc. To compare with calculated and measured dose distribution, film was inserted in rando phantom and irradiated in axial direction. The indexes of evaluation are percentage difference(%Diff) for absolute dose, RMSE(Root-mean-square-error) value for relative dose, coverage ratio and average dose in PTV. Results: The maximum difference at center point was -4.65 % in diameter 2 cm size. And the RMSE value between the calculated and measured off-axis dose distribution indicated that the measured dose distribution in diameter 2 cm was different from calculated and inaccurate compare to diameter 5 cm. In addition, Distance prescribed 95 % dose($D_{95}$) in diameter 2 cm was not covered in PTV and average dose value was lowest in all sizes. Conclusion: This study demonstrated that small sized PTV was not enough covered with prescribed dose in low density lung tissue. All indexes of experimental results in diameter 2 cm were much different from other sizes. It is showed that minimized PTV is not accurate and affects the results of radiation therapy. It is considered that extended margin at small PTV in low density lung tissue for enhancing target center dose is necessary and don't need to constraint Maximum dose in optimization.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.21 no.1
• /
• pp.29-41
• /
• 2019

Terrestrial Gross Primary Production (GPP) is the largest global carbon flux, and forest ecosystems are important because of the ability to store much more significant amounts of carbon than other terrestrial ecosystems. There have been several attempts to estimate GPP using mechanism-based models. However, mechanism-based models including biological, chemical, and physical processes are limited due to a lack of flexibility in predicting non-stationary ecological processes, which are caused by a local and global change. Instead mechanism-free methods are strongly recommended to estimate nonlinear dynamics that occur in nature like GPP. Therefore, we used the mechanism-free machine learning techniques to estimate the daily GPP. In this study, support vector machine (SVM), random forest (RF) and artificial neural network (ANN) were used and compared with the traditional multiple linear regression model (LM). MODIS products and meteorological parameters from eddy covariance data were employed to train the machine learning and LM models from 2006 to 2013. GPP prediction models were compared with daily GPP from eddy covariance measurement in a deciduous forest in South Korea in 2014 and 2015. Statistical analysis including correlation coefficient (R), root mean square error (RMSE) and mean squared error (MSE) were used to evaluate the performance of models. In general, the models from machine-learning algorithms (R = 0.85 - 0.93, MSE = 1.00 - 2.05, p < 0.001) showed better performance than linear regression model (R = 0.82 - 0.92, MSE = 1.24 - 2.45, p < 0.001). These results provide insight into high predictability and the possibility of expansion through the use of the mechanism-free machine-learning models and remote sensing for predicting non-stationary ecological processes such as seasonal GPP.

• Korean Journal of Remote Sensing
• /
• v.37 no.3
• /
• pp.515-530
• /
• 2021

This study is to estimate soil moisture (SM) using Sentinel-1A/B C-band SAR (synthetic aperture radar) images and Multiple Linear Regression Model(MLRM) in the Yongdam-Dam watershed of South Korea. Both the Sentinel-1A and -1B images (6 days interval and 10 m resolution) were collected for 5 years from 2015 to 2019. The geometric, radiometric, and noise corrections were performed using the SNAP (SentiNel Application Platform) software and converted to backscattering coefficient of VV and VH polarization. The in-situ SM data measured at 6 locations using TDR were used to validate the estimated SM results. The 5 days antecedent precipitation data were also collected to overcome the estimation difficulty for the vegetated area not reaching the ground. The MLRM modeling was performed using yearly data and seasonal data set, and correlation analysis was performed according to the number of the independent variable. The estimated SM was verified with observed SM using the coefficient of determination (R²) and the root mean square error (RMSE). As a result of SM modeling using only BSC in the grass area, R² was 0.13 and RMSE was 4.83%. When 5 days of antecedent precipitation data was used, R² was 0.37 and RMSE was 4.11%. With the use of dry days and seasonal regression equation to reflect the decrease pattern and seasonal variability of SM, the correlation increased significantly with R² of 0.69 and RMSE of 2.88%.

• Journal of Korea Water Resources Association
• /
• v.53 no.12
• /
• pp.1081-1095
• /
• 2020

This study is to evaluate stream flow and water quality changes of Yeongsan river basin (3,371.4 km2) by inter-basin water transfer (IBWT) from Juam dam of Seomjin river basin using SWAT (Soil and Water Assessment Tool). The SWAT was established using inlet function for IBWT between donor and receiving basins. The SWAT was calibrated and validated with 14 years (2005 ~ 2018) data of 1 stream (MR) and 2 multi-functional weir (SCW, JSW) water level gauging stations, and 3 water quality stations (GJ2, NJ, and HP) including data of IBWT and effluent from wastewater treatment plants of Yeongsan river basin. For streamflow and weir inflows (MR, SCW, and JSW), the coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE), root mean square error (RMSE), and percent bias (PBIAS) were 0.69 ~ 0.81, 0.61 ~ 0.70, 1.34 ~ 2.60 mm/day, and -8.3% ~ +7.6% respectively. In case of water quality, the R2 of SS, T-N, and T-P were 0.69 ~ 0.81, 0.61 ~ 0.70, and 0.54 ~ 0.63 respectively. The Yeongsan river basin average streamflow was 12.0 m3/sec and the average SS, T-N, and T-P were 110.5 mg/L, 4.4 mg/L, 0.18 mg/L respectively. Under the 130% scenario of IBWT amount, the streamflow, SS increased to 12.94 m3/sec (+7.8%), 111.26 mg/L (+0.7%) and the T-N, T-P decreased to 4.17 mg/L (-5.2%), 0.165 mg/L (-8.3%) respectively. Under the 70% scenario of IBWT amount, the streamflow, SS decreased to 11.07 m3/sec (-7.8%), 109.74 mg/L (-0.7%) and the T-N, T-P increased to 4.68 mg/L (+6.4%), 0.199 mg/L (+10.6%) respectively.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.66 no.2
• /
• pp.105-111
• /
• 2021

Rice lodging is an annual occurrence caused by typhoons accompanied by strong winds and strong rainfall, resulting in damage relating to pre-harvest sprouting during the ripening period. Thus, rapid estimations of the area of lodged rice are necessary to enable timely responses to damage. To this end, we obtained images related to rice lodging using a drone in Gimje, Buan, and Gunsan, which were converted to 128 × 128 pixels images. A convolutional neural network (CNN) model, a deep learning model based on these images, was used to predict rice lodging, which was classified into two types (lodging and non-lodging), and the images were divided in a 8:2 ratio into a training set and a validation set. The CNN model was layered and trained using three optimizers (Adam, Rmsprop, and SGD). The area of rice lodging was evaluated for the three fields using the obtained data, with the exception of the training set and validation set. The images were combined to give composites images of the entire fields using Metashape, and these images were divided into 128 × 128 pixels. Lodging in the divided images was predicted using the trained CNN model, and the extent of lodging was calculated by multiplying the ratio of the total number of field images by the number of lodging images by the area of the entire field. The results for the training and validation sets showed that accuracy increased with a progression in learning and eventually reached a level greater than 0.919. The results obtained for each of the three fields showed high accuracy with respect to all optimizers, among which, Adam showed the highest accuracy (normalized root mean square error: 2.73%). On the basis of the findings of this study, it is anticipated that the area of lodged rice can be rapidly predicted using deep learning.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_1
• /
• pp.1125-1139
• /
• 2022

In an unmanned aerial vehicles (UAVs) system, a physical offset can be existed between the global positioning system/inertial measurement unit (GPS/IMU) sensor and the observation sensor such as a hyperspectral sensor, and a lidar sensor. As a result of the physical offset, a misalignment between each image can be occurred along with a flight direction. In particular, in a case of multi-sensor system, an observation sensor has to be replaced regularly to equip another observation sensor, and then, a high cost should be paid to acquire a calibration parameter. In this study, we establish a precise sensor model equation to apply for a multiple sensor in common and propose an independent physical offset estimation method. The proposed method consists of 3 steps. Firstly, we define an appropriate rotation matrix for our system, and an initial sensor model equation for direct-georeferencing. Next, an observation equation for the physical offset estimation is established by extracting a corresponding point between a ground control point and the observed data from a sensor. Finally, the physical offset is estimated based on the observed data, and the precise sensor model equation is established by applying the estimated parameters to the initial sensor model equation. 4 region's datasets(Jeon-ju, Incheon, Alaska, Norway) with a different latitude, longitude were compared to analyze the effects of the calibration parameter. We confirmed that a misalignment between images were adjusted after applying for the physical offset in the sensor model equation. An absolute position accuracy was analyzed in the Incheon dataset, compared to a ground control point. For the hyperspectral image, root mean square error (RMSE) for X, Y direction was calculated for 0.12 m, and for the point cloud, RMSE was calculated for 0.03 m. Furthermore, a relative position accuracy for a specific point between the adjusted point cloud and the hyperspectral images were also analyzed for 0.07 m, so we confirmed that a precise data mapping is available for an observation without a ground control point through the proposed estimation method, and we also confirmed a possibility of multi-sensor fusion. From this study, we expect that a flexible multi-sensor platform system can be operated through the independent parameter estimation method with an economic cost saving.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_2
• /
• pp.1723-1735
• /
• 2022

Landslides are one of the most prevalent natural disasters, threating both humans and property. Also landslides can cause damage at the national level, so effective prediction and prevention are essential. Research to produce a landslide susceptibility map with high accuracy is steadily being conducted, and various models have been applied to landslide susceptibility analysis. Pixel-based machine learning models such as frequency ratio models, logistic regression models, ensembles models, and Artificial Neural Networks have been mainly applied. Recent studies have shown that the kernel-based convolutional neural network (CNN) technique is effective and that the spatial characteristics of input data have a significant effect on the accuracy of landslide susceptibility mapping. For this reason, the purpose of this study is to analyze landslide vulnerability using a pixel-based deep neural network model and a patch-based convolutional neural network model. The research area was set up in Gangwon-do, including Inje, Gangneung, and Pyeongchang, where landslides occurred frequently and damaged. Landslide-related factors include slope, curvature, stream power index (SPI), topographic wetness index (TWI), topographic position index (TPI), timber diameter, timber age, lithology, land use, soil depth, soil parent material, lineament density, fault density, normalized difference vegetation index (NDVI) and normalized difference water index (NDWI) were used. Landslide-related factors were built into a spatial database through data preprocessing, and landslide susceptibility map was predicted using deep neural network (DNN) and CNN models. The model and landslide susceptibility map were verified through average precision (AP) and root mean square errors (RMSE), and as a result of the verification, the patch-based CNN model showed 3.4% improved performance compared to the pixel-based DNN model. The results of this study can be used to predict landslides and are expected to serve as a scientific basis for establishing land use policies and landslide management policies.

• Journal of the Korean earth science society
• /
• v.43 no.5
• /
• pp.604-617
• /
• 2022

Sea surface temperature (SST) is a key variable that can be used to understand ocean-atmosphere phenomena and predict climate change. Satellite microwave remote sensing enables the measurement of SST despite the presence of clouds and precipitation in the sensor path. Therefore, considering the high utilization of microwave SST, it is necessary to continuously verify its accuracy and analyze its error characteristics. In this study, the validation of the microwave global precision measurement (GPM)/GPM microwave imager (GMI) SST around the Northwest Pacific and Korean Peninsula was conducted using surface drifter temperature data for approximately eight years from March 2014 to December 2021. The GMI SST showed a bias of 0.09K and an average root mean square error of 0.97K compared to the actual SST, which was slightly higher than that observed in previous studies. In addition, the error characteristics of the GMI SST were related to environmental factors, such as latitude, distance from the coast, sea wind, and water vapor volume. Errors tended to increase in areas close to coastal areas within 300 km of land and in high-latitude areas. In addition, relatively high errors were found in the range of weak wind speeds (<6 m s^-1) during the day and strong wind speeds (>10 m s^-1) at night. Atmospheric water vapor contributed to high SST differences in very low ranges of <30 mm and in very high ranges of >60 mm. These errors are consistent with those observed in previous studies, in which GMI data were less accurate at low SST and were estimated to be due to differences in land and ocean radiation, wind-induced changes in sea surface roughness, and absorption of water vapor into the microwave atmosphere. These results suggest that the characteristics of the GMI SST differences should be clarified for more extensive use of microwave satellite SST calculations in the seas around the Korean Peninsula, including a part of the Northwest Pacific.