• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.1199-1206
• /
• 2021

Accordingly, attention is also being paid to the agricultural use of remote sensing technique that non-destructively and continuously detects the growth and physiological status of crops. However, when remote sensing techniques are used for crop monitoring, it is possible to continuously monitor the abnormality of crops in real time. For this, standard growth information of crops is required and relative growth considering the cultivation environment must be identified. With the relationship between GDD (Growing Degree Days), which is the cumulative temperature related to crop growth obtained from ideal cultivation management, and the vegetation index as standard growth information, compared with the vegetation index observed with the spectralreflectance sensor(SRS_NDVI & SRS_PRI) in each rice paddy treated with standard cultivation management and non-fertilized, it was quantitatively identified as a time series. In the future, it is necessary to accumulate a database targeting various climatic conditions and varieties in the standard cultivation management area to establish a more reliable standard growth information.

• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.93-99
• /
• 2021

Recently, the miniaturization and digitalization for the inspection devices of point-of-care testing (POCT) are rapidly evolving. In the urine test, a lot of researches on index paper technology are being conducted because people can be self-diagnosed through visual color comparison using a urine test paper, Dipsick. The purpose of this study is to analyze the RGB values from the color changes on Dipstick Pad, which isused for urine test, using a smartphone camera. To this end, the primary, analytes in urine wasdiabetes-related parameters such as glucose, ketone body and pH, which is the most frequently tested elements, and we pursuited to quantify the changes in dipstick color caused from artificial urine containing different ranges of sugar, ketone body, and pH. In this experiment, changes in RGB values under bright and dark illuminances were compared, and changes in RGB value were monitored as a function of concentration of analytes under the ambient illumination of laboratory. As a result, color separation at the bright luminance region was good, but it did not appearat the low luminance region, and the changed profiles in RGB value under different illuminances was suggested to correct the problem of the color separation algorithm.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.2
• /
• pp.371-378
• /
• 2021

In this study, in order to analyze the spatial distribution of cold water occurred in the Southeast Sea of Korea, the K-means clustering method was used to analyze the ocean observatory buoy of Gori and Yangpo and GHTSST Level 4 from 2016 to 2018. The buoy data was used to identify the change in sea water temperature and the cold water occurrence at Gori and Yangpo in the Southeast Sea. As a result, the sea water temperature of Gori and Yangpo decreased equally at the cold water occurrence. Therefore, the reciprocal of the sea water temperature and the variance of SST were compared to see the changes of SST when the cold water occurs. When the reciprocal of the sea water temperature increases, the dispersion of SST also increases. Through this, it can be seen that there is a change in the water temperature distribution of SST in the sea when the cold water occurs. After that, K-means clustering was used to classify the cold water. After analyzing the optimal K value for clustering by using the Elbow method, it was possible to classify a region with cold water. Through this, it is estimated that the spatial distribution and diffusion range of the cold water, and it can be estimated and used in future studies to identify damage caused by the cold water and predict spatial spread.

• Journal of Broadcast Engineering
• /
• v.26 no.2
• /
• pp.125-131
• /
• 2021

In recent years, Convolutional Neural Networks (CNNs) have achieved outstanding performance in the fields of computer vision such as image classification, object detection, visual quality enhancement, etc. However, as huge amount of computation and memory are required in CNN models, there is a limitation in the application of CNN to low-power environments such as mobile or IoT devices. Therefore, the need for neural network compression to reduce the model size while keeping the task performance as much as possible has been emerging. In this paper, we propose a method to compress CNN models by combining matrix decomposition methods of LR (Low-Rank) approximation and CP (Canonical Polyadic) decomposition. Unlike conventional methods that apply one matrix decomposition method to CNN models, we selectively apply two decomposition methods depending on the layer types of CNN to enhance the compression performance. To evaluate the performance of the proposed method, we use the models for image classification such as VGG-16, RestNet50 and MobileNetV2 models. The experimental results show that the proposed method gives improved classification performance at the same range of 1.5 to 12.1 times compression ratio than the existing method that applies only the LR approximation.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.25 no.4
• /
• pp.75-82
• /
• 2021

The objective of this study is to examine the loss of prestressing stress in the developed prestressing reinforcing units using steel bar and pipe (SP). The main parameters were the reinforcing bar type, the magnitude of prestressed force, and prestressing method. The test results showed that the loss of prestressing stress for SP was highest in the initial prestressing step, which was higher for the compression introduction typed specimens than tension introduction typed specimens. The loss of prestressing stress of SP made with P800 was 1.6% for the compression introduction typed specimen with 0.8f_y, which was lowest than the other specimens. Meanwhile, the relaxation of SP with the respect to the time ranged between 0.4 and 1.9%, irrespective of SP material type, the magnitude of prestressed force, and prestressing method. These values were less than 2.5%, which is the maximum value for the relaxation of prestressed reinforcing steel bars in design codes. Consequently, considering the loss of stress developed in the initial prestressing step, the developed SP material type, prestressing introduction method, and magnitude are recommended to be P800, compression introduction type, and 0.8f_y.

• Journal of vocational education research
• /
• v.35 no.4
• /
• pp.63-88
• /
• 2016

This study surveyed and analyzed the labor market result when comparing the case that junior college students found a job after transferring to 4-year university with the case that they found the job directly after graduating the junior college. The difference of the occupational structure following the transfer was surveyed in detail and this study examined what effect the transfer had on the occupation status. The graduate occupation movement survey(GOMS) was used and the samples of 544 persons were extracted to use through propensity score matching(PSM) to raise the choice convenience of the sample. The occupation status index to use as a dependent variable was developed to apply. The study result is as follows. First, as a result of confirming the occupational structure of the transfer, the transfer students were distributed in more various industrial fields evenly compared to the non-transfer students in the horizontal industrial structure, and non-transfer students were intensively distributed in the certain field. In the vertical occupation status structure, transfer students were distributed in the high occupation status more than non-transfer students who were broadly distributed in the low occupation status. Second, it was revealed the college transfer was highly effective for the occupation status, which appeared statistically significantly. The explanation variable that appeared significantly outside of that included the parents' academic background, major affiliation, employment pattern, major job coincidence, and job coincidence. The higher students' academic background was, the higher the occupation status was and in case of the engineering natural science affiliation, the occupation status was higher than the humanities social science. In case of full-time workers, their occupation status was higher than one of part time workers and it was revealed the occupation status was high when the their major coincided with the job and their educational/technical level coincided with their job.

• Journal of Wetlands Research
• /
• v.21 no.1
• /
• pp.77-83
• /
• 2019

Stream water qualities have been predicted in the year 2002 and 2014 through providing the Hwangguji Stream Rectification Plan. However, the reliability of result for predicted water quality was relatively lower by applying conventional values of the parameters in model. In this study deoxygenation coefficients between Sema bridge(HGJ2) and Sujik bridge(HGJ3) have been evaluated based on the observed data of water quality and travelling time to compare with the applied value of coefficients in predicting water quality model. The values of deoxygenation coefficient $0.078day^{-1}{\sim}0.748day^{-1}$ for normal period and $0.053day^{-1}{\sim}0.505day^{-1}$ for drought period have been calculated based of observed data between Sema bridge and Sujik bridge. The values of coefficients $0.02day^{-1}{\sim}3.4day^{-1}$ have been applied in predicting water quality model in the year 2002 and $0.043day^{-1}$ 2014. Thus, the simulated results of stream water quality were better than the observed data in 2002, and worse in 2014. It has shown that values of deoxygenation coefficient should be properly estimated based on observed data to predict proper stream water quality by model.

• Journal of the Korean Wood Science and Technology
• /
• v.47 no.2
• /
• pp.229-238
• /
• 2019

Various wood defects occur during tree growing or wood processing. Thus, to use wood practically, it is necessary to objectively assess their quality based on the usage requirement by accurately classifying their defects. However, manual visual grading and species classification may result in differences due to subjective decisions; therefore, computer-vision-based image analysis is required for the objective evaluation of wood quality and the speeding up of wood production. In this study, the SIFT+k-NN and CNN models were used to implement a model that automatically classifies knots and analyze its accuracy. Toward this end, a total of 1,172 knot images in various shapes from five domestic conifers were used for learning and validation. For the SIFT+k-NN model, SIFT technology was used to extract properties from the knot images and k-NN was used for the classification, resulting in the classification with an accuracy of up to 60.53% when k-index was 17. The CNN model comprised 8 convolution layers and 3 hidden layers, and its maximum accuracy was 88.09% after 1205 epoch, which was higher than that of the SIFT+k-NN model. Moreover, if there is a large difference in the number of images by knot types, the SIFT+k-NN tended to show a learning biased toward the knot type with a higher number of images, whereas the CNN model did not show a drastic bias regardless of the difference in the number of images. Therefore, the CNN model showed better performance in knot classification. It is determined that the wood knot classification by the CNN model will show a sufficient accuracy in its practical applicability.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.22 no.2
• /
• pp.1-14
• /
• 2019

In rural areas, composting is a source of non-point pollutants. However, as the quantitative distribution and loading have not been estimated, it is difficult to determine the effect of composting on stream water quality. In this study, composting datum acquired by unmanned aerial vehicle(UAV) was verified by using terrestrial LiDAR, and the management status and load change of the composting was investigated by UAV with manual control flight, thereby obtaining the basic data to determine the effect on the water system. As a result of the comparative accuracy assessment based on terrestrial LiDAR, the difference in the digital surface model(DSM) was within 0.21m and the accuracy of the volume was 93.24%. We expect that the accuracy is sufficient to calculate and utilize the composting load acquired by UAV. Thus, the management status of composting can be investigated by UAV. As the total load change of composting were determined to be $1,172.16m^3$, $1,461.66m^3$, and $1,350.53m^3$, respectively, the load change of composting could be confirmed. We expect that the results of this study can contribute to efficient management of non-point source pollution by UAV.

• Journal of Digital Convergence
• /
• v.17 no.6
• /
• pp.219-227
• /
• 2019

This study aims to examine Blockchain research trend using bibliometrics-based network analysis. The data were collected from WoS, Scopus, Korea Citation Index and National science & Technology Information Service, from 2009 to 2018. As results, the number of publications has started increasing rapidly from 2017 and it showed the initial stage of formation of coauthor network. Words often used in the title of the publications were related to application development, controversy and technology development. In addition, the majority of domestic papers are in the subject of social science, while international papers tend to focus on engineering issues. The results of the temporal analysis show that Korean researchers' block chain 3.0 started in 2017 and are rapidly increasing in 2018. The number of citations was associated with publication year in a statistically signifiant way. By examining these research trends, we hope that this paper can be a useful basis for the development of blockchain. Future research is expected to reveal more clearly the knowledge structure and characteristics of blockchain around the world.