• Korean Journal of Remote Sensing
• /
• v.38 no.6_1
• /
• pp.991-1005
• /
• 2022

Satellite imaging is an effective supplementary data source for detecting and verifying nuclear activity. It is also highly beneficial in regions with limited access and information, such as nuclear installations. Time series analysis, in particular, can identify the process of preparing for the conduction of a nuclear experiment, such as relocating equipment or changing facilities. Differences in the semantic segmentation findings of time series photos were employed in this work to detect changes in meaningful items connected to nuclear activity. Building, road, and small object datasets made of KOMPSAT 3/3A photos given by AIHub were used to train deep learning models such as U-Net, PSPNet, and Attention U-Net. To pick relevant models for targets, many model parameters were adjusted. The final change detection was carried out by including object information into the first change detection, which was obtained as the difference in semantic segmentation findings. The experiment findings demonstrated that the suggested approach could effectively identify altered pixels. Although the suggested approach is dependent on the accuracy of semantic segmentation findings, it is envisaged that as the dataset for the region of interest grows in the future, so will the relevant scope of the proposed method.

• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.73-80
• /
• 2023

The recent steep increase in the minimum hourly wage has increased the burden of labor costs, and the share of unmanned stores is increasing in the aftermath of COVID-19. As a result, theft crimes targeting unmanned stores are also increasing, and the "Just Walk Out" system is introduced to prevent such thefts, and LiDAR sensors, weight sensors, etc. are used or manually checked through continuous CCTV monitoring. However, the more expensive sensors are used, the higher the initial cost of operating the store and the higher the cost in many ways, and CCTV verification is difficult for managers to monitor around the clock and is limited in use. In this paper, we would like to propose an AI image processing fusion algorithm that can solve these sensors or human-dependent parts and detect customers who perform abnormal behaviors such as theft at low costs that can be used in unmanned stores and provide cloud-based notifications. In addition, this paper verifies the accuracy of each algorithm based on behavior pattern data collected from unmanned stores through motion capture using mediapipe, object detection using YOLO, and fusion algorithm and proves the performance of the convergence algorithm through various scenario designs.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.57-73
• /
• 2021

Maintenance and prevention of failure through anomaly detection of ICT infrastructure is becoming important. System monitoring data is multidimensional time series data. When we deal with multidimensional time series data, we have difficulty in considering both characteristics of multidimensional data and characteristics of time series data. When dealing with multidimensional data, correlation between variables should be considered. Existing methods such as probability and linear base, distance base, etc. are degraded due to limitations called the curse of dimensions. In addition, time series data is preprocessed by applying sliding window technique and time series decomposition for self-correlation analysis. These techniques are the cause of increasing the dimension of data, so it is necessary to supplement them. The anomaly detection field is an old research field, and statistical methods and regression analysis were used in the early days. Currently, there are active studies to apply machine learning and artificial neural network technology to this field. Statistically based methods are difficult to apply when data is non-homogeneous, and do not detect local outliers well. The regression analysis method compares the predictive value and the actual value after learning the regression formula based on the parametric statistics and it detects abnormality. Anomaly detection using regression analysis has the disadvantage that the performance is lowered when the model is not solid and the noise or outliers of the data are included. There is a restriction that learning data with noise or outliers should be used. The autoencoder using artificial neural networks is learned to output as similar as possible to input data. It has many advantages compared to existing probability and linear model, cluster analysis, and map learning. It can be applied to data that does not satisfy probability distribution or linear assumption. In addition, it is possible to learn non-mapping without label data for teaching. However, there is a limitation of local outlier identification of multidimensional data in anomaly detection, and there is a problem that the dimension of data is greatly increased due to the characteristics of time series data. In this study, we propose a CMAE (Conditional Multimodal Autoencoder) that enhances the performance of anomaly detection by considering local outliers and time series characteristics. First, we applied Multimodal Autoencoder (MAE) to improve the limitations of local outlier identification of multidimensional data. Multimodals are commonly used to learn different types of inputs, such as voice and image. The different modal shares the bottleneck effect of Autoencoder and it learns correlation. In addition, CAE (Conditional Autoencoder) was used to learn the characteristics of time series data effectively without increasing the dimension of data. In general, conditional input mainly uses category variables, but in this study, time was used as a condition to learn periodicity. The CMAE model proposed in this paper was verified by comparing with the Unimodal Autoencoder (UAE) and Multi-modal Autoencoder (MAE). The restoration performance of Autoencoder for 41 variables was confirmed in the proposed model and the comparison model. The restoration performance is different by variables, and the restoration is normally well operated because the loss value is small for Memory, Disk, and Network modals in all three Autoencoder models. The process modal did not show a significant difference in all three models, and the CPU modal showed excellent performance in CMAE. ROC curve was prepared for the evaluation of anomaly detection performance in the proposed model and the comparison model, and AUC, accuracy, precision, recall, and F1-score were compared. In all indicators, the performance was shown in the order of CMAE, MAE, and AE. Especially, the reproduction rate was 0.9828 for CMAE, which can be confirmed to detect almost most of the abnormalities. The accuracy of the model was also improved and 87.12%, and the F1-score was 0.8883, which is considered to be suitable for anomaly detection. In practical aspect, the proposed model has an additional advantage in addition to performance improvement. The use of techniques such as time series decomposition and sliding windows has the disadvantage of managing unnecessary procedures; and their dimensional increase can cause a decrease in the computational speed in inference.The proposed model has characteristics that are easy to apply to practical tasks such as inference speed and model management.

• Journal of Intelligence and Information Systems
• /
• v.28 no.1
• /
• pp.89-106
• /
• 2022

Recently, the amount of video data collected from smartphones, CCTVs, black boxes, and high-definition cameras has increased rapidly. According to the increasing video data, the requirements for analysis and utilization are increasing. Due to the lack of skilled manpower to analyze videos in many industries, machine learning and artificial intelligence are actively used to assist manpower. In this situation, the demand for various computer vision technologies such as object detection and tracking, action detection, emotion detection, and Re-ID also increased rapidly. However, the object detection and tracking technology has many difficulties that degrade performance, such as re-appearance after the object's departure from the video recording location, and occlusion. Accordingly, action and emotion detection models based on object detection and tracking models also have difficulties in extracting data for each object. In addition, deep learning architectures consist of various models suffer from performance degradation due to bottlenects and lack of optimization. In this study, we propose an video analysis system consists of YOLOv5 based DeepSORT object tracking model, SlowFast based action recognition model, Torchreid based Re-ID model, and AWS Rekognition which is emotion recognition service. Proposed model uses single-linkage hierarchical clustering based Re-ID and some processing method which maximize hardware throughput. It has higher accuracy than the performance of the re-identification model using simple metrics, near real-time processing performance, and prevents tracking failure due to object departure and re-emergence, occlusion, etc. By continuously linking the action and facial emotion detection results of each object to the same object, it is possible to efficiently analyze videos. The re-identification model extracts a feature vector from the bounding box of object image detected by the object tracking model for each frame, and applies the single-linkage hierarchical clustering from the past frame using the extracted feature vectors to identify the same object that failed to track. Through the above process, it is possible to re-track the same object that has failed to tracking in the case of re-appearance or occlusion after leaving the video location. As a result, action and facial emotion detection results of the newly recognized object due to the tracking fails can be linked to those of the object that appeared in the past. On the other hand, as a way to improve processing performance, we introduce Bounding Box Queue by Object and Feature Queue method that can reduce RAM memory requirements while maximizing GPU memory throughput. Also we introduce the IoF(Intersection over Face) algorithm that allows facial emotion recognized through AWS Rekognition to be linked with object tracking information. The academic significance of this study is that the two-stage re-identification model can have real-time performance even in a high-cost environment that performs action and facial emotion detection according to processing techniques without reducing the accuracy by using simple metrics to achieve real-time performance. The practical implication of this study is that in various industrial fields that require action and facial emotion detection but have many difficulties due to the fails in object tracking can analyze videos effectively through proposed model. Proposed model which has high accuracy of retrace and processing performance can be used in various fields such as intelligent monitoring, observation services and behavioral or psychological analysis services where the integration of tracking information and extracted metadata creates greate industrial and business value. In the future, in order to measure the object tracking performance more precisely, there is a need to conduct an experiment using the MOT Challenge dataset, which is data used by many international conferences. We will investigate the problem that the IoF algorithm cannot solve to develop an additional complementary algorithm. In addition, we plan to conduct additional research to apply this model to various fields' dataset related to intelligent video analysis.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.34 no.2
• /
• pp.202-208
• /
• 2005

This study was conducted to examine the food habits, knowledge of nutrition and actual conditions of food ingestion of adolescent middle school students according to questionnaire answers. Questionnaires were completed by 524 students, divided into a healthy group (n=289) and an unhealthy group (n=235) according to clinical signs. Further questions were asked of the two groups in the areas of food habits, knowledge of nutrition and nutritional attitude. The results were as follows: Mean age of all subjects was 14, heights for male and female students were 162.0 em, and 157.2 cm, weights were 53.4 kg, and 49.4, respectively. Heights and weights of male students were greater than those of female students. The body mass index (BMI) for male and female students was 20.3 kg/$m^2$ and 20.0 kg/$m^2$, respectively, and all data were within normal ranges. There were no significant differences in mean age, height, weight, and BMI between the healthy and unhealthy groups. There was no significant difference in body image recognition between the two groups, although the ratio of dissatisfaction with their own body shape was significantly higher in the female unhealthy group (46.1%), than in the female healthy group (33.0%) (p<0.05). In the area of the struggle to control body weight during the previous year, the female unhealthy group (59.4%) was higher than the female healthy group (38.4%) (p<0.01). There was no significant difference in the scores between the two groups in the areas of knowledge of nutrition and the nutritional attitude. Meal frequency and meal patterns were showed that having breakfast less than 4x/week was significantly higher in the female unhealthy group (44.0%), than in the female healthy group (30.7%) (p<0.01). Meal frequency for suppers<4x/week showed that the female unhealthy group (18.8%) was also higher than the female healthy group (10.7%). Therefore, the unhealthy group exhibited a higher pattern of missing both breakfast and supper. The male unhealthy group (16.7%) dined out more frequently than the male healthy group (12.3%) (p<0.01), and female unhealthy group also indulged in snacking significantly more frequently than the female healthy group. The unhealthy group also ate only 1 item for meals more frequently than the healthy group and no significant difference. The conclusion of this study is that adolescent Korean middle school students, who showed a higher incidence of clinical symptoms, representing an unhealthy status, missed breakfast and supper, and dined out and indulged in snacking more frequently. Their quality of breakfast and satisfaction of body image were also lower than the healthy group. These results indicated that there is a high correlation between a Korean adolescent's health status, food habits and body image satisfaction. It is recommended that a more intense program of nutritional education and monitoring be introduce into the current Korean middle-school system in order to optimally support and maximize the health potential of the current population of Korean student.

• Progress in Medical Physics
• /
• v.22 no.3
• /
• pp.140-147
• /
• 2011

The purpose of this study was to estimate internal motion using molecular sieve for quantitative improvement of lung tumor and to localize lung tumor in the small animal PET image by evaluated data. Internal motion has been demonstrated in small animal lung region by molecular sieve contained radioactive substance. Molecular sieve for internal lung motion target was contained approximately 37 kBq Cu-64. The small animal PET images were obtained from Siemens Inveon scanner using external trigger system (BioVet). SD-Rat PET images were obtained at 60 min post injection of FDG 37 MBq/0.2 mL via tail vein for 20 min. Each line of response in the list-mode data was converted to sinogram gated frames (2~16 bin) by trigger signal obtained from BioVet. The sinogram data was reconstructed using OSEM 2D with 4 iterations. PET images were evaluated with count, SNR, FWHM from ROI drawn in the target region for quantitative tumor analysis. The size of molecular sieve motion target was $1.59{\times}2.50mm$. The reference motion target FWHM of vertical and horizontal was 2.91 mm and 1.43 mm, respectively. The vertical FWHM of static, 4 bin and 8 bin was 3.90 mm, 3.74 mm, and 3.16 mm, respectively. The horizontal FWHM of static, 4 bin and 8 bin was 2.21 mm, 2.06 mm, and 1.60 mm, respectively. Count of static, 4 bin, 8 bin, 12 bin and 16 bin was 4.10, 4.83, 5.59, 5.38, and 5.31, respectively. The SNR of static, 4 bin, 8 bin, 12 bin and 16 bin was 4.18, 4.05, 4.22, 3.89, and 3.58, respectively. The FWHM were improved in accordance with gate number increase. The count and SNR were not proportionately improve with gate number, but shown the highest value in specific bin number. We measured the optimal gate number what minimize the SNR loss and gain improved count when imaging lung tumor in small animal. The internal motion estimation provide localized tumor image and will be a useful method for organ motion prediction modeling without external motion monitoring system.

• Nuclear Medicine and Molecular Imaging
• /
• v.40 no.4
• /
• pp.218-227
• /
• 2006

Purpose: The HSV1-tk reporter gene system is the most widely used system because of its advantage that direct monitoring is possible without the introduction of a separate reporter gene in case of HSV1-tk suicide gene therapy. In this study, we investigate the usefulness of the reporter probe (substrate), $9-(4-[^{18}F]Fluoro-3-hydroxymethylbutyl)$guanine ($[^{18}F]FHBG$) for non-invasive reporter gene imaging using PET in HSV1-tk expressing hepatoma model. Materials and Methods: Radiolabeled FHBG was prepared in 8 steps from a commercially available triester. The labeling reaction was carried out by NCA nucleophilic substitution with $K[^{18}F]/K2.2.2.$ in acetonitrile using N2-monomethoxytrityl-9-14-(tosyl)-3-monomethoxytritylmethylbutyl]guanine as a precursor, followed by deprotection with 1 N HCl. Preliminary biological properties of the probe were evaluated with MCA cells and MCA-tk cells transduced with HSV1-tk reporter gene. In vitro uptake and release-out studies of $[^{18}F]FHBG$ were performed, and was analyzed correlation between $[^{18}F]FHBG$ uptake ratio according to increasing numeric count of MCA-tk cells and degree of gene expression. MicroPET scan image was obtained with MCA and MCA-tk tumor bearing Balb/c-nude mouse model. Results: $[^{18}F]FHBG$ was purified by reverse phase semi-HPLC system and collected at around 16-18 min. Radiothemical yield was about 20-25%) (corrected for decay), radiochemical purity was >95% and specific activity was around >55.5 $GBq/{\mu}\;mol$. Specific accumulation of $[^{18}F]FHBG$ was observed in HSV1-tk gene transduced MCA-tk cells but not in MCA cells, and consecutive 1 hour release-out results showed more than 86% of uptaked $[^{18}F]FHBG$ was retained inside of cells. The uptake of $[^{18}F]FHBG$ was showed a highly significant linear correlation ($R^2=0.995$) with increasing percentage of MCA-tk numeric cell count. In microPET scan images, remarkable difference of accumulation was observed for the two type of tumors. Conclusion: $[^{18}F]FHBG$ appears to be a useful as non-invasive PET imaging substrate in HSV1-tk expressing hepatoma model.

• Economic and Environmental Geology
• /
• v.49 no.5
• /
• pp.381-388
• /
• 2016

The purpose of this study is to integrate and visualize using mapping techniques based on precise seabed geomorphology, seafloor backscattering images and high-resolution underwater images of the nearshore area around the Dokdo, in the East Sea. We have been obtained the precise topography map using multibeam echosounder system around the nearshore area(~50 m) of the southern part of the Seodo. Side scan sonar survey for analysis seafloor backscattering images was carried out in the same area of topography data. High-resolution underwater images(zone(a), zone(b), zone(c)) were taken in significant habitat scope of the nearshore area of the southern part of the Seodo. Using the results of bathymetry, seafloor backscattering images, high-resolution underwater images, we performed an integrated visualization about the nearshore area of the Dokdo. The integrated visualizing techniques are possible to make the seabed characteristic mapping results of the nearshore area of the Dokdo. The integrated visualization results present more complex and reliable information than separate geological products for seabed environmental mapping study and it is useful to understand the relation between seafloor characteristics and topographic environments of the study area. The integrated visualizing techniques and mapping analysis need to study sustainably and periodically, for effective monitoring of the nearshore ecosystem of the Dokdo.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.25 no.4
• /
• pp.258-266
• /
• 2023

This study analyzed methodologies for estimating carbon stocks of perennial woody crops and the research cases in overseas countries. As a result, we found that Australia, Bulgaria, Canada, and Japan are using the stock-difference method, while Austria, Denmark, and Germany are estimating the change in the carbon stock based on the gain-loss method. In some overseas countries, the researches were conducted on estimating the carbon stock change using image data as tier 3 phase beyond the research developing country-specific factors as tier 2 phase. In South Korea, convergence studies as the third stage were conducted in forestry field, but advanced research in the agricultural field is at the beginning stage. Based on these results, we suggest directions for the following four future researches: 1) securing national-specific factors related to emissions and removals in the agricultural field through the development of allometric equation and carbon conversion factors for perennial woody crops to improve the completeness of emission and removals statistics, 2) implementing policy studies on the cultivation area calculation refinement with fruit tree-biomass-based maturity, 3) developing a more advanced estimation technique for perennial woody crops in the agricultural sector using allometric equation and remote sensing techniques based on the agricultural and forestry satellite scheduled to be launched in 2025, and to establish a matrix and monitoring system for perennial woody crop cultivation areas in the agricultural sector, Lastly, 4) estimating soil carbon stocks change, which is currently estimated by treating all agricultural areas as one, by sub-land classification to implement a dynamic carbon cycle model. This study suggests a detailed guideline and advanced methods of carbon stock change calculation for perennial woody crops, which supports 2050 Carbon Neutral Strategy of Ministry of Agriculture, Food, and Rural Affairs and activate related research in agricultural sector.

• Journal of Distribution Science
• /
• v.8 no.3
• /
• pp.49-56
• /
• 2010

At the initial stage of Internet advertising, banner advertising came into fashion. As the Internet developed into a central part of daily lives and the competition in the on-line advertising market was getting fierce, there was not enough space for banner advertising, which rushed to portal sites only. All these factors was responsible for an upsurge in advertising prices. Consequently, the high-cost and low-efficiency problems with banner advertising were raised, which led to an emergence of keyword advertising as a new type of Internet advertising to replace its predecessor. In the beginning of 2000s, when Internet advertising came to be activated, display advertisement including banner advertising dominated the Net. However, display advertising showed signs of gradual decline, and registered minus growth in the year 2009, whereas keyword advertising showed rapid growth and started to outdo display advertising as of the year 2005. Keyword advertising refers to the advertising technique that exposes relevant advertisements on the top of research sites when one searches for a keyword. Instead of exposing advertisements to unspecified individuals like banner advertising, keyword advertising, or targeted advertising technique, shows advertisements only when customers search for a desired keyword so that only highly prospective customers are given a chance to see them. In this context, it is also referred to as search advertising. It is regarded as more aggressive advertising with a high hit rate than previous advertising in that, instead of the seller discovering customers and running an advertisement for them like TV, radios or banner advertising, it exposes advertisements to visiting customers. Keyword advertising makes it possible for a company to seek publicity on line simply by making use of a single word and to achieve a maximum of efficiency at a minimum cost. The strong point of keyword advertising is that customers are allowed to directly contact the products in question through its more efficient advertising when compared to the advertisements of mass media such as TV and radio, etc. The weak point of keyword advertising is that a company should have its advertisement registered on each and every portal site and finds it hard to exercise substantial supervision over its advertisement, there being a possibility of its advertising expenses exceeding its profits. Keyword advertising severs as the most appropriate methods of advertising for the sales and publicity of small and medium enterprises which are in need of a maximum of advertising effect at a low advertising cost. At present, keyword advertising is divided into CPC advertising and CPM advertising. The former is known as the most efficient technique, which is also referred to as advertising based on the meter rate system; A company is supposed to pay for the number of clicks on a searched keyword which users have searched. This is representatively adopted by Overture, Google's Adwords, Naver's Clickchoice, and Daum's Clicks, etc. CPM advertising is dependent upon the flat rate payment system, making a company pay for its advertisement on the basis of the number of exposure, not on the basis of the number of clicks. This method fixes a price for advertisement on the basis of 1,000-time exposure, and is mainly adopted by Naver's Timechoice, Daum's Speciallink, and Nate's Speedup, etc, At present, the CPC method is most frequently adopted. The weak point of the CPC method is that advertising cost can rise through constant clicks from the same IP. If a company makes good use of strategies for maximizing the strong points of keyword advertising and complementing its weak points, it is highly likely to turn its visitors into prospective customers. Accordingly, an advertiser should make an analysis of customers' behavior and approach them in a variety of ways, trying hard to find out what they want. With this in mind, her or she has to put multiple keywords into use when running for ads. When he or she first runs an ad, he or she should first give priority to which keyword to select. The advertiser should consider how many individuals using a search engine will click the keyword in question and how much money he or she has to pay for the advertisement. As the popular keywords that the users of search engines are frequently using are expensive in terms of a unit cost per click, the advertisers without much money for advertising at the initial phrase should pay attention to detailed keywords suitable to their budget. Detailed keywords are also referred to as peripheral keywords or extension keywords, which can be called a combination of major keywords. Most keywords are in the form of texts. The biggest strong point of text-based advertising is that it looks like search results, causing little antipathy to it. But it fails to attract much attention because of the fact that most keyword advertising is in the form of texts. Image-embedded advertising is easy to notice due to images, but it is exposed on the lower part of a web page and regarded as an advertisement, which leads to a low click through rate. However, its strong point is that its prices are lower than those of text-based advertising. If a company owns a logo or a product that is easy enough for people to recognize, the company is well advised to make good use of image-embedded advertising so as to attract Internet users' attention. Advertisers should make an analysis of their logos and examine customers' responses based on the events of sites in question and the composition of products as a vehicle for monitoring their behavior in detail. Besides, keyword advertising allows them to analyze the advertising effects of exposed keywords through the analysis of logos. The logo analysis refers to a close analysis of the current situation of a site by making an analysis of information about visitors on the basis of the analysis of the number of visitors and page view, and that of cookie values. It is in the log files generated through each Web server that a user's IP, used pages, the time when he or she uses it, and cookie values are stored. The log files contain a huge amount of data. As it is almost impossible to make a direct analysis of these log files, one is supposed to make an analysis of them by using solutions for a log analysis. The generic information that can be extracted from tools for each logo analysis includes the number of viewing the total pages, the number of average page view per day, the number of basic page view, the number of page view per visit, the total number of hits, the number of average hits per day, the number of hits per visit, the number of visits, the number of average visits per day, the net number of visitors, average visitors per day, one-time visitors, visitors who have come more than twice, and average using hours, etc. These sites are deemed to be useful for utilizing data for the analysis of the situation and current status of rival companies as well as benchmarking. As keyword advertising exposes advertisements exclusively on search-result pages, competition among advertisers attempting to preoccupy popular keywords is very fierce. Some portal sites keep on giving priority to the existing advertisers, whereas others provide chances to purchase keywords in question to all the advertisers after the advertising contract is over. If an advertiser tries to rely on keywords sensitive to seasons and timeliness in case of sites providing priority to the established advertisers, he or she may as well make a purchase of a vacant place for advertising lest he or she should miss appropriate timing for advertising. However, Naver doesn't provide priority to the existing advertisers as far as all the keyword advertisements are concerned. In this case, one can preoccupy keywords if he or she enters into a contract after confirming the contract period for advertising. This study is designed to take a look at marketing for keyword advertising and to present effective strategies for keyword advertising marketing. At present, the Korean CPC advertising market is virtually monopolized by Overture. Its strong points are that Overture is based on the CPC charging model and that advertisements are registered on the top of the most representative portal sites in Korea. These advantages serve as the most appropriate medium for small and medium enterprises to use. However, the CPC method of Overture has its weak points, too. That is, the CPC method is not the only perfect advertising model among the search advertisements in the on-line market. So it is absolutely necessary that small and medium enterprises including independent shopping malls should complement the weaknesses of the CPC method and make good use of strategies for maximizing its strengths so as to increase their sales and to create a point of contact with customers.