• The Journal of the Convergence on Culture Technology
• /
• v.10 no.1
• /
• pp.435-441
• /
• 2024

In order for the large amount of collected data sets to be used as deep learning training data, sensitive personal information such as resident registration number and disease information must be changed or encrypted to prevent it from being exposed to hackers, and the data must be reconstructed to match the structure of the built deep learning model. Currently, these tasks are performed manually by experts, which takes a lot of time and money. To solve these problems, this paper proposes a technique that can automatically perform data processing tasks to protect personal information during the deep learning process. In the proposed technique, privacy protection tasks are performed based on data generalization and data reconstruction tasks are performed using circular queues. To verify the validity of the proposed technique, it was directly implemented using C language. As a result of the verification, it was confirmed that data generalization was performed normally and data reconstruction suitable for the deep learning model was performed properly.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.4
• /
• pp.639-644
• /
• 2024

Recent developments in the use of drones on battlefields, extending beyond reconnaissance to firepower support, have greatly increased the importance of technologies for early automatic drone identification. In this study, to identify an effective image classification model that can distinguish drones from other aerial targets of similar size and appearance, such as birds and balloons, we utilized a dataset of 3,600 images collected from the internet. We adopted a transfer learning approach that combines the feature extraction capabilities of three pre-trained convolutional neural network models (VGG16, ResNet50, InceptionV3) with an additional classifier. Specifically, we conducted a comparative analysis of the performance of these three pre-trained models to determine the most effective one. The results showed that the InceptionV3 model achieved the highest accuracy at 99.66%. This research represents a new endeavor in utilizing existing convolutional neural network models and transfer learning for drone identification, which is expected to make a significant contribution to the advancement of drone identification technologies.

• Journal of radiological science and technology
• /
• v.38 no.1
• /
• pp.1-6
• /
• 2015

Whole spine scanography (WSS) is a radiological examination that exposes the whole body of the individual being examined to x-ray radiation. WSS is often repeated during the treatment period, which results in a much greater radiation exposure than that in routine x-ray examinations. The aims of the current study were to evaluate the patient dose of WSS using computer simulation, image magnification and angulation of phantom image using different patient position. We evaluated the effective dose(ED) of 23 consecutive patients (M : F = 13:10) who underwent WSS, based on the automatic image pasting method for multiple exposure digital radiography. The Anterior-Posterior position(AP) and Posterior-Anterior position( PA) projection EDs were evaluated based on the PC based Monte Carlo simulation. We measured spine transverse process distance and angulation using DICOM measurement. For all patient, the average ED was 0.069 mSv for AP position and 0.0361 mSv for PA position. AP position calculated double exposure then PA position. For male patient, the average ED was 0.089 mSv(AP) and 0.050 mSv(PA). For female patient, the average ED was 0.0431 mSv(AP) and 0.026 mSv(PA). The transverse process of PA spine image measured 5% higher than AP but angulation of transverse process was no significant differences. In clinical practice, just by change the patient position was conformed to reduce the ED of patient. Therefor we need to redefine of protocol for digital radiography such as WSS. whole spine scanography, effective dose, patient exposure dose, exposure direction. protocol optimization.

• Journal of Animal Science and Technology
• /
• v.62 no.4
• /
• pp.485-494
• /
• 2020

Limiting amino acids (AAs) for milk production in dairy cows fed on a concentrate diet of corn grain and soybean meal was evaluated in this study. Four lactating and multiparous Holstein cows (in third or fourth parities, with an average body weight of 633 ± 49.2 kg), 8 to 9 weeks into their lactation period, were used in a 4 × 4 Latin square design. The experiment comprised four dietary treatments: (1) no intravenous infusion (control); (2) control plus intravenous infusion of an AA mixture of 6 g/d methionine, 19.1 g/d lysine, 13.8 g/d isoleucine, and 15.4 g/d valine (4AA); (3) control plus intravenous infusion of the AA mixture without methionine (no-Met); and (4) control plus intravenous infusion of the AA mixture without lysine (no-Lys). All animals were fed on a controlled diet (1 kg/d alfalfa hay, 10 kg/d silage, 14 kg/d concentrate mixture, ad libitum timothy hay). The AA composition of the diet and blood were determined using an automatic AA analyzer. Milk composition (protein, fat, lactose, urea nitrogen, and somatic cell counts) was determined using a MilkoScan. The results showed that feed intake for milk production did not differ from that of intravenous infusion using a limiting AA mixture. The 4AA treatment numerically had the highest milk yield (32.4 kg/d), although there was no difference when compared with the control (31.2 kg/d), no-Met (31.3 kg/d), and no-Lys (31.7 kg/d) treatments. The concentration of AAs in blood plasma of cows in all treatments, mainly isoleucine and valine, increased significantly compared with that of control. The no-Met treatment increased (p < 0.05) the concentration of lysine in the blood relative to the control and no-Lys treatments, whereas the no-Lys treatment increased (p < 0.05) the concentration of methionine relative to the control and no-Met treatments. In conclusion, milk production increased when feeding 10 g/d methionine to the cows, together with their concentrate diet of corn grain and soybean meal.

• Journal of the Korean Society of Industry Convergence
• /
• v.21 no.3
• /
• pp.141-147
• /
• 2018

In a time when product development cycles are getting shorter and shorter, many companies are making efforts to develop products with high reliability in a short period of time, accelerated life test is widely used as a method to quickly evaluate reliability. Accelerated life test reduces the test life or the life of the product from the observed data by shortening the lifetime of the product or abruptly lowering the performance under the worse condition than the actual condition in order to shorten the test cost or the test time. In this paper, BL3640A-06P+RB35, DC12V model, which is used in the support device of an automatic rotation type digital signage, which display various information such as textures and images on a display screen in a public place or a commercial space, BLDC motors were subjected to a constant stress test and at the rotational speed of 1rpm, $180^{\circ}$ rotation and reverse rotation under actual use conditions, the stress was imposed on the rotating speed of 2rpm and the weight of the actual installed product from 22.2kgf to 10kgf were installed. The lifetime of the actual use environment condition is 23,545 hours and the rotation speed is accelerated. The life time of the acceleration condition with the additional weight is 1,380 hours. The acceleration factor is calculated as 17.06, the one year guarantee test day is 235 days to 14 days, of the period from 470 days to 28 days, and the third year from 704 days to 42 days. The test date of the BLDC motor was tested on the shortened test date, and the rotational speed and the current value were measured. It is found that there is no defect even if it operates as the test date corresponding to the specified one year warranty period and the 3 year accelerated life test which is experimented. Using the statistical technique of the regression analysis the expected time for the motor to defect to #4 samples was 20 years.

• Journal of the Institute of Convergence Signal Processing
• /
• v.10 no.1
• /
• pp.40-48
• /
• 2009

Leading of the computer, IT technology has make great strides. As a information-industry-community was highly developed, user's needs to convenience about intelligence and humanization of interface is being increase today. Nowadays, researches with are related to BCI are progress put the application-technology development first in importance eliminating research about fountainhead technology with DB construction. These problems are due to a BCI-related research studies have not overcome the initial level, and not toward a systematic study. Brain wave are collected from subjects is a signal that the signal is appropriate and necessary in the experiment is difficult to distinguish. In addition, brain wave that it's not necessary to collect the experiment, serious eyes flicker, facial and body movements of an EMG and electrodes attached to the state, noise, vibration, etc. It is hard to collect accurate brain wave was caused by mixing disturbance wave in experiment on the environment. This movement, and the experiment of subject impact on the environment due to the mixing disturbance wave can cause that lowering cognitive and decline of efficiency when embodied BCI system. Therefore, in this paper, we propose an accurate and efficient brain-wave DB building system that more exactness and cognitive basis studies when embodied BCI system with brain-wave. For the minimize about brain wave DB with mixing disturbance, we propose a DB building method using an automatic control and prevent unnecessary action, put to use the subjects face tracking.

• Journal of Convergence for Information Technology
• /
• v.9 no.3
• /
• pp.56-67
• /
• 2019

The aim of this study was to investigate the influence of rosemary oil inhalation on memory, attention and autonomic nervous system according to the concentration difference in the aged. The research design was non-equivalent control group non-synchronized design. Participants were 89 individuals aged 65 or older who live in the community. Participants inhaled almond carrier oil(control group), 10%(experimental group A) and 100%(experimental group B) rosemary oil. Memory, attention, and autonomic nervous system responses were measured. Data were analyzed by SPSS win 24.0. The differences of the group and time were analyzed through repeated measure ANOVA. There were no significant differences in immediate recall (F=.42, p =.656), delayed recall (F=.45, p=.639), recognition (F=1.45, p=.242), digit span-forward (F=1.53, p=.223), digit span-backward (F=.46, p=.636), activities of sympathetic nerve system (LF)(F=.19, p=.828), activities of parasympathetic nerve system (HF)(F=.37, p=.694), LH/HF(F=1.39, p=.256), systolic blood pressure (F=.37, p=.694), diastolic blood pressure (F=1.25, p=.291). The inhalation of 10% and 100% rosemary oil for five minutes showed no significant effects on memory, attention and automatic nervous system in the aged.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.4
• /
• pp.200-208
• /
• 2022

In this study, an automatic information collection system and geographic information construction algorithm for the transportation disadvantaged using electric mobility aids are implemented using an object recognition model. Recognizes objects that the disabled person encounters while moving, and acquires coordinate information. It provides an improved route selection map compared to the existing geographic information for the disabled. Data collection consists of a total of four layers including the HW layer. It collects image information and location information, transmits them to the server, recognizes, and extracts data necessary for geographic information generation through the process of classification. A driving experiment is conducted in an actual barrier-free zone, and during this process, it is confirmed how efficiently the algorithm for collecting actual data and generating geographic information is generated.The geographic information processing performance was confirmed to be 70.92 EA/s in the first round, 70.69 EA/s in the second round, and 70.98 EA/s in the third round, with an average of 70.86 EA/s in three experiments, and it took about 4 seconds to be reflected in the actual geographic information. From the experimental results, it was confirmed that the walking weak using electric mobility aids can drive safely using new geographic information provided faster than now.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.2
• /
• pp.517-524
• /
• 2024

This study aims to train and implement a deep learning model for the fusion of website creation and artificial intelligence, in the era known as the AI revolution following the launch of the ChatGPT service. The deep learning model was trained using 3,000 collected web page images, processed based on a system of component and layout classification. This process was divided into three stages. First, prior research on AI models was reviewed to select the most appropriate algorithm for the model we intended to implement. Second, suitable web page and paragraph images were collected, categorized, and processed. Third, the deep learning model was trained, and a serving interface was integrated to verify the actual outcomes of the model. This implemented model will be used to detect multiple paragraphs on a web page, analyzing the number of lines, elements, and features in each paragraph, and deriving meaningful data based on the classification system. This process is expected to evolve, enabling more precise analysis of web pages. Furthermore, it is anticipated that the development of precise analysis techniques will lay the groundwork for research into AI's capability to automatically generate perfect web pages.

• Journal of Korean Society of Disaster and Security
• /
• v.15 no.1
• /
• pp.1-12
• /
• 2022

An automatic optimization technique is used to estimate the optimal parameters of the hydrologic model, and different hydrologic response results can be provided depending on objective functions. In this study, the parameters of the event-based rainfall-runoff model were estimated using various objective functions, the reproducibility of the hydrograph according to the objective functions was evaluated, and appropriate objective functions were proposed. As the rainfall-runoff model, the storage function model(SFM), which is a lumped hydrologic model used for runoff simulation in the current Korean flood forecasting system, was selected. In order to evaluate the reproducibility of the hydrograph for each objective function, 9 rainfall events were selected for the Cheoncheon basin, which is the upstream basin of Yongdam Dam, and widely-used 7 objective functions were selected for parameter estimation of the SFM for each rainfall event. Then, the reproducibility of the simulated hydrograph using the optimal parameter sets based on the different objective functions was analyzed. As a result, RMSE, NSE, and RSR, which include the error square term in the objective function, showed the highest accuracy for all rainfall events except for Event 7. In addition, in the case of PBIAS and VE, which include an error term compared to the observed flow, it also showed relatively stable reproducibility of the hydrograph. However, in the case of MIA, which adjusts parameters sensitive to high flow and low flow simultaneously, the hydrograph reproducibility performance was found to be very low.