• Journal of the Korean Association of Geographic Information Studies
• /
• v.25 no.4
• /
• pp.63-80
• /
• 2022

As a technology to analyze and predict an issue has been developed by constructing real space into virtual space, it is becoming more important to acquire precise spatial information in complex cities. In this study, images were acquired using an unmanned aerial vehicle for urban area with complex landscapes, and land cover classification was performed object-based image analysis and semantic segmentation techniques, which were image classification technique suitable for high-resolution imagery. In addition, based on the imagery collected at the same time, the replicability of land cover classification of each artificial intelligence (AI) model was examined for areas that AI model did not learn. When the AI models are trained on the training site, the land cover classification accuracy is analyzed to be 89.3% for OBIA-RF, 85.0% for OBIA-DNN, and 95.3% for U-Net. When the AI models are applied to the replicability assessment site to evaluate replicability, the accuracy of OBIA-RF decreased by 7%, OBIA-DNN by 2.1% and U-Net by 2.3%. It is found that U-Net, which considers both morphological and spectroscopic characteristics, performs well in land cover classification accuracy and replicability evaluation. As precise spatial information becomes important, the results of this study are expected to contribute to urban environment research as a basic data generation method.

• Journal of the Korean Geotechnical Society
• /
• v.39 no.7
• /
• pp.31-37
• /
• 2023

Since tunnels are built underground, it is impossible to check visually the location and degree of deterioration of steel ribs. Therefore, in tunnel maintenance, GPR images are generally used to detect steel ribs. While research on GPR image analysis employing artificial neural networks has primarily focused on detecting underground pipes and road damage, there have been limited applications for analyzing tunnel GPR data, specifically for steel rib detection, both internationally and domestically. In this study, a one-step object detection algorithm called YOLO, based on a convolutional neural network, was utilized to automate the localization of steel ribs using GPR data. The performance of the algorithm is then analyzed. Two datasets were employed for the analysis. A dataset comprising 512 original images and another dataset consisting of 2,048 augmented images. The omission rate, which represents the ratio of undetected steel ribs to the total number of steel ribs, was 0.38% for the model using the augmented data, whereas the omission rate for the model using only the original data was 7.18%. Thus, from an automation standpoint, it is more practical to employ an augmented dataset.

• Korean Journal of Remote Sensing
• /
• v.39 no.6_3
• /
• pp.1679-1692
• /
• 2023

South Korea is a country that emits a large amount of pollutants as a result of population growth and industrial development and is also severely affected by transboundary air pollution due to its geographical location. As pollutants from both domestic and foreign sources contribute to air pollution in Korea, the location of air pollutant emission sources is crucial for understanding the movement and distribution of pollutants in the atmosphere and establishing national-level air pollution management and response strategies. Based on this background, this study aims to effectively acquire spatial information on domestic and international air pollutant emission sources, which is essential for analyzing air pollution status, by utilizing high-resolution optical satellite images and deep learning-based image segmentation models. In particular, industrial parks and quarries, which have been evaluated as contributing significantly to transboundary air pollution, were selected as the main research subjects, and images of these areas from multi-purpose satellites 3 and 3A were collected, preprocessed, and converted into input and label data for model training. As a result of training the U-Net model using this data, the overall accuracy of 0.8484 and mean Intersection over Union (mIoU) of 0.6490 were achieved, and the predicted maps showed significant results in extracting object boundaries more accurately than the label data created by course annotations.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.6
• /
• pp.1-8
• /
• 2005

In this paper, we propose a two-stage head tracking algorithm adequate for real-time active camera system having pan-tilt-zoom functions. In the color convergence stage, we first assume that the shape of a head is an ellipse and its model color histogram is acquired in advance. Then, the min-shift method is applied to roughly estimate a target position by examining the histogram similarity of the model and a candidate ellipse. To reflect the temporal change of object color and enhance the reliability of mean-shift based tracking, the target histogram obtained in the previous frame is considered to update the model histogram. In the updating process, to alleviate error-accumulation due to outliers in the target ellipse of the previous frame, the target histogram in the previous frame is obtained within an ellipse adaptively shrunken on the basis of the model histogram. In addition, to enhance tracking reliability further, we set the initial position closer to the true position by compensating the global motion, which is rapidly estimated on the basis of two 1-D projection datasets. In the subsequent stage, we refine the position and size of the ellipse obtained in the first stage by using shape information. Here, we define a robust shape-similarity function based on the gradient direction. Extensive experimental results proved that the proposed algorithm performs head hacking well, even when a person moves fast, the head size changes drastically, or the background has many clusters and distracting colors. Also, the propose algorithm can perform tracking with the processing speed of about 30 fps on a standard PC.

• Journal of Distribution Research
• /
• v.16 no.5
• /
• pp.43-61
• /
• 2011

In this research, we analyze about the effect of the traditional market supporting policy about 16 large unit self-governing bodies and effeciency through DEA analysis. The plan for supports of the traditional market were shown. Now, the object of the traditional market supporting policy was any more not improvement of facilities and the thing which is the attraction of customer activity for the sales increase of the traditional market could be confirmed. For the sales increase, supporting of the field like the client information center, source indicator, autonomic packing stand, and the broadcasting facility are effective more. In addition, for the visiting customer inducement activation, we could know that supporting of the field like client information center, broadcasting facility, broadcasting advertisement, and the premium event for gift certificate were required. The method including the customer distribution service operation, which is the various product development and cross sellings and client friendly based on data which it investigates the actual conditions the market merchant on the Incheon area and consumer with the concrete plan for support gift certificate activation, the market information system construction, and etc. was shown.

• Smart Media Journal
• /
• v.9 no.3
• /
• pp.107-115
• /
• 2020

Recently, BIM has become mandatory in the construction field, research on various use cases is increasing. In particular, when virtual reality technology, one of the core technologies of the 4th industrial revolution, and BIM are combined, it can be used in various fields such as preliminary design review and construction simulation. Until now, however, virtual reality grafting technology is only used as a simple prototype or as a model house. Also, it is difficult to activate virtual construction because it is expensive to produce high-quality virtual reality contents. Therefore, in this paper, in order to increase the utilization and quality of the virtual construction field, a study was conducted to shorten the material mapping time, which takes a lot of time when producing virtual reality contents using BIM. To this end, object properties were assigned to enable material mapping in the BIM model, and materials most used in the construction field were configured, and automated material function development and final tests were conducted that automatically map properties and materials. For the test, 10 models were used and the test was repeated three times, and the productivity improvement of about 50.16% was finally achieved. In the future, we plan to conduct research on physical data weight reduction based on the advanced material mapping automation function and the large-capacity BIM model.

• Spatial Information Research
• /
• v.23 no.2
• /
• pp.49-57
• /
• 2015

Recently, Through the linkage and opening, the fusion of the spatial information, it is necessary for productive ecosystem to provide a variety of information and to increase the civil use. Depending on the economic growth, demand for quality of life and well-being has been on the increase. Spatial information service contents for the public convenience has emerged to solve the problem such as health, safety, welfare and discomfort of daily life This study aims to implement search services for a tailored residence space through the three-dimensional data modeling on cadastral information. To achieve this goal, we established the requirements for deriving a registered object by investigating recent trend with respect to existing cadastral data model and defined property and relationship. Focusing on Songpa-gu, Jamsil station in Seoul, we implemented search services for a tailored residence space for three-dimensional right analysis in conjunction with residential and commercial complex building. As a result, we derived a way to supply 3D cadastre information through open platforms (VWorld) and to represent efficiently, which is able to improve the quality of spatial information service contents for the public convenience as well as to widen utilization of information.

• Journal of the Korean Society of Radiology
• /
• v.12 no.7
• /
• pp.879-886
• /
• 2018

A 3-dimensional (D) printer is a device capable of outputting a three-dimensional solid object based on data modeled in a computer. These features are utilized in the bone model X - ray phantom production etc using CT data by fusing with the radiation science field. A bone model phantom was made using data obtained by CT scan of an existing Pelvis phantom, using PLA, Wood, XT-CF20, Glow fill, Steel filaments which are materials of Fused Filament Fabrication (FFF) 3D printer.Measure Hounsfield Unit (HU) with images obtained by CT scan of the existing Pelvis phantom and five material phantoms made with 3D printer under the same conditions,SI and SNR were measured using a diagnostic X-ray generator, and each phantom was compared and analyzed.As a result, the X - ray phantom in the X - ray examination condition of the limb was found to be most suitable for the glow fill filament.The characteristics of the filament can be known to the base of this research and the practicality of X - ray phantom fabrication was confirmed.

• Journal of Practical Engineering Education
• /
• v.16 no.4
• /
• pp.449-457
• /
• 2024

AIoT, the intelligent Internet of Things, refers to a technology that collects data measured by IoT devices and applies machine learning technology to create and utilize predictive models. Existing research on AIoT technology education focused on building an educational AIoT platform and teaching how to use it. However, there was a lack of case studies that taught the process of automatically creating and utilizing machine learning models from data measured by IoT devices. In this paper, we developed a case study using a machine learning model for AIoT technology education. The case developed in this paper consists of the following steps: data collection from AIoT devices, data preprocessing, automatic creation of machine learning models, calculation of accuracy for each model, determination of valid models, and data prediction using the valid models. In this paper, we considered that sensors in AIoT devices measure different ranges of values, and presented an example of data preprocessing accordingly. In addition, we developed a case where AIoT devices automatically determine what information they can predict by automatically generating several machine learning models and determining effective models with high accuracy among these models. By applying the developed cases, a variety of educational contents using AIoT, such as prediction-based object control using AIoT, can be developed.

• Journal of Computational Design and Engineering
• /
• v.3 no.4
• /
• pp.385-397
• /
• 2016

Scaffolds are essential in bone tissue engineering, as they provide support to cells and growth factors necessary to regenerate tissue. In addition, they meet the mechanical function of the bone while it regenerates. Currently, the multiple methods for designing and manufacturing scaffolds are based on regular structures from a unit cell that repeats in a given domain. However, these methods do not resemble the actual structure of the trabecular bone which may work against osseous tissue regeneration. To explore the design of porous structures with similar mechanical properties to native bone, a geometric generation scheme from a reaction-diffusion model and its manufacturing via a material jetting system is proposed. This article presents the methodology used, the geometric characteristics and the modulus of elasticity of the scaffolds designed and manufactured. The method proposed shows its potential to generate structures that allow to control the basic scaffold properties for bone tissue engineering such as the width of the channels and porosity. The mechanical properties of our scaffolds are similar to trabecular tissue present in vertebrae and tibia bones. Tests on the manufactured scaffolds show that it is necessary to consider the orientation of the object relative to the printing system because the channel geometry, mechanical properties and roughness are heavily influenced by the position of the surface analyzed with respect to the printing axis. A possible line for future work may be the establishment of a set of guidelines to consider the effects of manufacturing processes in designing stages.