• Journal of Drive and Control
• /
• v.20 no.2
• /
• pp.15-23
• /
• 2023

In this study, two-dimensional location of crops for auto weeding was detected using deep learning. To construct a dataset for soybean detection, an image-capturing system was developed using a mono camera and single-board computer and the system was mounted on a weeding robot to collect soybean images. A dataset was constructed by extracting RoI (region of interest) from the raw image and each sample was labeled with soybean and the background for classification learning. The deep learning model consisted of four convolutional layers and was trained with a weakly supervised learning method that can provide object localization only using image-level labeling. Localization of the soybean area can be visualized via CAM and the two-dimensional position of the soybean was estimated by clustering the pixels associated with the soybean area and transforming the pixel coordinates to world coordinates. The actual position, which is determined manually as pixel coordinates in the image was evaluated and performances were 6.6(X-axis), 5.1(Y-axis) and 1.2(X-axis), 2.2(Y-axis) for MSE and RMSE about world coordinates, respectively. From the results, we confirmed that the center position of the soybean area derived through deep learning was sufficient for use in automatic weeding systems.

• Convergence Security Journal
• /
• v.22 no.2
• /
• pp.105-114
• /
• 2022

This paper is presents an intelligent control system that visualizes the direction of arrival for hearing impaired using personal mobility, and aims to recognize and prevent dangerous situations caused by sound such as alarm sounds and crack sounds on roads. The position estimation method of sound source uses a machine learning classification model characterized by generalized correlated phase transformation based on time difference of arrival. In the experimental environment reproducing the road situations, four classification models learned after extracting learning data according to wind speeds 0km/h, 5.8km/h, 14.2km/h, and 26.4km/h were compared with grid search cross validation, and the Muti-Layer Perceptron(MLP) model with the best performance was applied as the optimal algorithm. When wind occurred, the proposed algorithm showed an average performance improvement of 7.6-11.5% compared to the previous studies.

• The KIPS Transactions:PartB
• /
• v.9B no.6
• /
• pp.853-862
• /
• 2002

Face Detection can be defined as follows : Given a digitalized arbitrary or image sequence, the goal of face detection is to determine whether or not there is any human face in the image, and if present, return its location, direction, size, and so on. This technique is based on many applications such face recognition facial expression, head gesture and so on, and is one of important qualify factors. But face in an given image is considerably difficult because facial expression, pose, facial size, light conditions and so on change the overall appearance of faces, thereby making it difficult to detect them rapidly and exactly. Therefore, this paper proposes fast and exact face detection which overcomes some restrictions by using neural network. The proposed system can be face detection irrelevant to facial expression, background and pose rapidily. For this. face detection is performed by neural network and detection response time is shortened by reducing search region and decreasing calculation time of neural network. Reduced search region is accomplished by using skin color segment and frame difference. And neural network calculation time is decreased by reducing input vector sire of neural network. Principle Component Analysis (PCA) can reduce the dimension of data. Also, pose estimates in extracted facial image and eye region is located. This result enables to us more informations about face. The experiment measured success rate and process time using the Squared Mahalanobis distance. Both of still images and sequence images was experimented and in case of skin color segment, the result shows different success rate whether or not camera setting. Pose estimation experiments was carried out under same conditions and existence or nonexistence glasses shows different result in eye region detection. The experiment results show satisfactory detection rate and process time for real time system.

• Korean Journal of Clinical Laboratory Science
• /
• v.49 no.3
• /
• pp.300-307
• /
• 2017

Electroencephalography (EEG) is distinct from other medical imaging tests in that it is a functional test that helps to diagnosis disorders related to the brain, such as epilepsy. The most important abilities for a medical technologist when performing an EEG are knowing the exact location of the electrode and recording the EEG wave clearly, except for artifacts. Although theoretical education and practical training are both included in the curriculum for improving these abilities, sufficient practical training has been lacking due to problems like expensive equipment and insufficient practical training time. We try to solve these issues by manufacturing the training 10-20 electrode system and by estimating the efficiency and satisfaction of the training 10-20 electrode system through a questionnaire. The time required for practical training using this system was $43.58{\pm}9.647min$, which proved to be efficient. The satisfaction score of participants who experienced curriculum practical training was improved from $7.21{\pm}2.285$ to $9.46{\pm}1.166$. Based on these findings, it is considered that practical training via the use of the training 10-20 electrode system will solve the problems, such as lack of equipment and insufficient practical training time. Nonetheless, to further improve the training 10-20 electrode system, it must overcome the limitations of developing a device capable of checking the actual brain waves and validating the exact location of electrode attachment.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.15 no.3
• /
• pp.45-51
• /
• 2016

Existing real-time signal control system was brought up typical problems which are supersaturated condition, point detection system and loop detection system. For that reason, the next generation signal control system of advanced form is required. Following thesis aimed at calculating queue length for the next generation signal control system to utilize basic parameter of signal control in crossing queue instead of the volume of real-time through traffic. Overflow saturated condition which was appeared as limit of existing system was focused to set-up range. Real-time location information of individual vehicle which is collected by GPS data. It converted into the coordinate to apply shock wave model with an linear equation that is extracted by regression model applied by a least square. Through the calculated queue length and link length by contrast, If queue length exceed the link, queue of downstream intersection is included as queue length that upstream queue vehicle is judeged as affecting downstream intersection. In result of operating correlation analysis among link travel time to judge confidence of extracted queue length, Both of links were shown over 0.9 values. It is appeared that both of links are highly correlated. Following research is significant using real-time data to calculate queue length and contributing to signal control system.

• Journal of Korea Water Resources Association
• /
• v.55 no.6
• /
• pp.447-459
• /
• 2022

This study estimated the grid-type precipitation quantile for the Korean Peninsula using PERSIANN-CCS-CDR (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Cloud Classification System-Climate Data Record), a satellite based re-analysis precipitation data. The period considered is a total of 38 years from 1983 to 2020. The spatial resolution of the data is 0.04° and the temporal resolution is 3 hours. For the probability distribution, the Gumbel distribution which is generally used for frequency analysis was used, and the probability weighted moment method was applied to estimate parameters. The duration ranged from 3 hours to 144 hours, and the return period from 2 years to 500 years was considered. The results were compared and reviewed with the estimated precipitation quantile using precipitation data from the Automated Synoptic Observing System (ASOS) weather station. As a result, the parameter estimates of the Gumbel distribution from the PERSIANN-CCS-CDR showed a similar pattern to the results of the ASOS as the duration increased, and the estimates of precipitation quantiles showed a rather large difference when the duration was short. However, when the duration was 18 h or longer, the difference decreased to less than about 20%. In addition, the difference between results of the South and North Korea was examined, it was confirmed that the location parameters among parameters of the Gumbel distribution was markedly different. As the duration increased, the precipitation quantile in North Korea was relatively smaller than those in South Korea, and it was 84% of that of South Korea for a duration of 3 h, and 70-75% of that of South Korea for a duration of 144 h.

• 한국방재학회:학술대회논문집
• /
• 2011.02a
• /
• pp.172-172
• /
• 2011

The interaction between land surface and atmosphere is essentially affected by hydrometeorological variables including soil moisture. Accurate estimation of soil moisture at spatial and temporal scales is crucial to better understand its roles to the weather systems. The KLDAS(Korea Land Data Assimilation System) is a regional, specifically Korea peninsula land surface information systems. As other prior land data assimilation systems, this can provide initial soil field information which can be used in atmospheric simulations. For this study, as an enabling high-resolution tool, weather research and forecasting(WRF-ARW) model is applied to produce precipitation data using GFS(Global Forecast System) with GFS embedded and KLDAS soil moisture information as initialization data. WRF-ARW generates precipitation data for a specific region using different parameters in physics options. The produced precipitation data will be employed for simulations of Hydrological Models such as HEC(Hydrologic Engineering Center) - HMS(Hydrologic Modeling System) as predefined input data for selected regional water responses. The purpose of this study is to show the impact of a hydrometeorological variable such as soil moisture in KLDAS on hydrological consequences in Korea peninsula. The study region, Chongmi River Basin, is located in the center of Korea Peninsular. This has 60.8Km river length and 17.01% slope. This region mostly consists of farming field however the chosen study area placed in mountainous area. The length of river basin perimeter is 185Km and the average width of river is 9.53 meter with 676 meter highest elevation in this region. We have four different observation locations : Sulsung, Taepyung, Samjook, and Sangkeug observatoriesn, This watershed is selected as a tentative research location and continuously studied for getting hydrological effects from land surface information. Simulations for a real regional storm case(June 17~ June 25, 2006) are executed. WRF-ARW for this case study used WSM6 as a micro physics, Kain-Fritcsch Scheme for cumulus scheme, and YSU scheme for planetary boundary layer. The results of WRF simulations generate excellent precipitation data in terms of peak precipitation and date, and the pattern of daily precipitation for four locations. For Sankeug observatory, WRF overestimated precipitation approximately 100 mm/day on July 17, 2006. Taepyung and Samjook display that WRF produced either with KLDAS or with GFS embedded initial soil moisture data higher precipitation amounts compared to observation. Results and discussions in detail on accuracy of prediction using formerly mentioned manners are going to be presented in 2011 Annual Conference of the Korean Society of Hazard Mitigation.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.25 no.6
• /
• pp.541-548
• /
• 2012

The hollow slab has advantages that its self-weight does not greatly increase notwithstanding the increase of its thickness and its flexural performance does not significantly degrade in comparison with general reinforced concrete slab. However, the utilization of the hollow slab is currently being underestimated in spite of structural system that enables economic design of building and construction of eco-friendly structure. the significant reasons for this situation is that the method of structural analysis and design for hollow slab is not generalized. In this study, to consider practical compressive zone of hollow slab, the equation for its flexural strength is proposed by the volume of compressive stress block according to neutral axis location in hollow section assumed. Existing estimation method of flexural strength of hollow slab considering only compressive zone above hollow part is evaluated as the most conservative method and the method estimating flexural strength by two alternative cross-section of hollow slab is evaluated as more practical method.

• Spatial Information Research
• /
• v.15 no.1
• /
• pp.81-94
• /
• 2007

Even though the data of digital cadastral map, stored in the Korea Land Information System(hereinafter referred to as "KLIS") has a wide variety of errors, it has been used in a cadastral surveying field without any compensating work. As a result, it has given rise to a lot of problems when we use this digital cadastral map in the cadastral surveying field. That is why we should analyse the accuracy of digital cadastral map which has been stored in the KLIS. In this context, this paper has been intended to provide acceptable proposals with regards to present operation of KLIS or the cadastral re-survey project in the future by analysing the data of digital cadastral map which is one of the layers of the KLIS. Our study results showed that the accuracy of both the location and area in the present KLIS datum was not satisfactory Therefore, in our own judgment, it is necessary to revise cadastral maps more accurately in order to use in the field of cadastral surveying or in carrying out the cadastral re-survey project.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.25 no.12
• /
• pp.2048-2055
• /
• 2001

This paper develops a computational model for estimating the ride quality of a cabover type large-sized truck in a double wheel bumpy ride test. The computational model is developed using ADAMS. To verify the developed model, an actual vehicle double wheel bumpy ride test is performed. In the test, the vehicle maintains a straight course with a constant velocity such that the front two wheels are passed the bump at the same time. The bump has the height of 60mm, and the width of 550mm. In the test, four velocities are used. They are 10kph, 20kph, 30kph and 40kph. Since the large-sized truck's center of gravity location is high, and its weight is heavy, it is a quite severe test condition to perform the test with more than 30kph velocity. In the test, vertical accelerations on the floor of the cab are measured. The measured accelerations are compared to the simulation results. From the comparison, it is shown that the developed model can predict not only the measured acceleration's tendency but also peak accelerations quite well. In this paper, the validated model is utilized to compare the ride quality between a vehicle with a multi-leaf spring and a vehicle with a tapered leaf spring in the front suspension system in a double bumpy ride test.