• ETRI Journal
• /
• 제29권5호
• /
• pp.700-702
• /
• 2007

We present a relevance feedback approach based on multi-class support vector machine (SVM) learning and cluster-merging which can significantly improve the retrieval performance in region-based image retrieval. Semantically relevant images may exhibit various visual characteristics and may be scattered in several classes in the feature space due to the semantic gap between low-level features and high-level semantics in the user's mind. To find the semantic classes through relevance feedback, the proposed method reduces the burden of completely re-clustering the classes at iterations and classifies multiple classes. Experimental results show that the proposed method is more effective and efficient than the two-class SVM and multi-class relevance feedback methods.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제9권2호
• /
• pp.199-208
• /
• 2017

We perform an experiment on lacquer formation with simple test device. The anti-lacquer is one of important issues to increase durability, and to improve performance in the engines because the lacquer formation cause sticking of fuel injection pump, scuffing of cylinder liners, and increase of lubricant oil consumption in the marine diesel engines. We suggest this simple test in order to save enormous experimental cost in marine diesel engines, and in order to have ease in performing the various tests. The influences of the Base Number (BN) of lubricant oils and the sulfur content of fuel oils in the formation of lacquer are investigated. In order to investigate physical and chemical properties of lacquer, we perform a variety of tests such as, visual inspection, EDS. In addition, we investigate adhesion of lacquer by pull-off test quantitatively, and perform dissolution test with dilute sulfuric acid.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2006년도 International Symposium on GPS/GNSS Vol.1
• /
• pp.247-250
• /
• 2006

Integrating a GPS based vehicle navigation system and the latest optimal algorithms, this research aims to develop a real-time decision support platform for concrete plant to provide the optimal solutions for ready mixed concrete delivery. The platform includes fleet tracking system, simulation and optimization tools, and visual interface which is useful to monitor delivery progress, to obtain crucial historical and real-time data for simulation, and to improve the efficiency of the plant operation. This paper presents configuration of the system and performance evaluation based on operational data.

• 한국멀티미디어학회논문지
• /
• 제24권6호
• /
• pp.796-804
• /
• 2021

In computer vision, visual tracking method addresses the problem of localizing an specific object in video sequence according to the bounding box. In this paper, we propose a tracking method by introducing the feature correlation comparison into the siamese network to increase its matching identification. We propose a way to compute location of object to improve matching performance by a correlation operation, which locates parts for solving the searching problem. The higher layer in the network can extract a lot of object information. The lower layer has many location information. To reduce error rate of the object center point, we built a siamese network that extracts the distribution and location information of target objects. As a result of the experiment, the average center error rate was less than 25%.

• 한국간호교육학회지
• /
• 제18권1호
• /
• pp.111-119
• /
• 2012

Purpose: The purpose of this study was to identify the effects on knowledge and performance confidence of nursing students in the emergency care of patients with dyspnea after simulation education using a human simulator. Method: The research design was a nonequivalent control group pretest-posttest design. For the experimental group the human simulator was used to provide simulation. Also included were base learning with audio-visual material, explanations about simulation, using SimMan for human simulation, and debriefing. Pre and post-tests were conducted to compare differences in knowledge and performance confidence. Result: The (t=3.83, p<.000) than the control group. For the experimental group, the differences in pretest-posttest scores for knowledge (t=2.30, p=.025) and performance confidence (t=4.28, p<.000) were significantly higher than the experimental group had significantly higher scores for knowledge (t=3.03, p=.004) and performance confidence (t=3.83, p<.001knowledge (t=2.30, p=.025) and performance confidence (t=4.28, p<.000) were significantly higher than the control group. Conclusion: The results of this study indicate that for student nurses, knowledge and performance confidence in emergency care of patients with dyspnea improve with human simulator simulation education. Further study is suggested to develop other scenarios for emergency care and identify the effects of critical thinking and satisfaction when using human simulator simulation education.

• KIEAE Journal
• /
• 제11권5호
• /
• pp.137-143
• /
• 2011

An atrium space, unlike ordinary office rooms, accommodates variety of activities such as moving and resting of people and usually houses variety of vegitation to improve amenity and indoor environment. Many atrium buildings in Korea have been designed by considering the environmental criteria for human beings, not for the vegitation in the atrium space. Especially the daylighting designs are mostly focused on the required illuminances for various visual tasks of the occupants and glare controls. As a result, some atrium spaces do not provide sufficient light to the interior plants. Consequently, these atrium spaces require a high level of electric lighting to compensate the deficit of natural light for the photosynthesis of the vegitation. The purpose of this study was to suggest design guidelines for 4-sided atrium spaces having different well indices (WI), plan aspect ratio (PAR), and cardinal orientation. The findings from this study might be referenced by building designers when designing or selecting canopy systems by considering the daylight performances of the uncovered atrium spaces. In the study, the daylight performance was evaluated in terms of daylight autonomy (DA).

• Corrosion Science and Technology
• /
• 제18권6호
• /
• pp.253-257
• /
• 2019

A deep learning based real-time painting surface inspection algorithm is proposed herein, designed for developing an autonomous inspection drone. The painting surface inspection is usually conducted manually. However, the manual inspection has a limitation in obtaining accurate data for correct judgement on the surface because of human error and deviation of individual inspection experiences. The best method to replace manual surface inspection is the vision-based inspection method with a camera, using various image processing algorithms. Nevertheless, the visual inspection is difficult to apply to surface inspection due to diverse appearances of material, hue, and lightning effects. To overcome technical limitations, a deep learning-based pattern recognition algorithm is proposed, which is specialized for painting surface inspections. The proposed algorithm functions in real time on the embedded board mounted on an autonomous inspection drone. The inspection results data are stored in the database and used for training the deep learning algorithm to improve performance. The various experiments for pre-inspection of painting processes are performed to verify real-time performance of the proposed deep learning algorithm.

• 대한전기학회논문지:전력기술부문A
• /
• 제53권10호
• /
• pp.565-572
• /
• 2004

This paper proposes an expert system which can enhance the accuracy of real-time bus reconfiguration strategy by adopting local minimum tree search method and minimize the spreading effect of the fault by considering totally the operating condition when a main transformer fault occurs in the automated substation. The local minimum tree search method to expand the best-first search method. This method has an advantage which can improve the performance of solution within the limits of the real-time condition. The inference strategy proposed expert system consists of two stages. The first stage determines the switching candidate set by searching possible switching candidates starting from the main transformer or busbar related to the event. And, second stage determines the rational real-time bus reconfiguration strategy based on heuristic rules for the obtained switching candidate set. Also, this paper studies the generalized distribution substation modelling using graph theory and a substation database is designed based on the study result. The inference engine of the expert system and the substation database is implemented in MFC function of Visual C++. Finally, the performance and effectiveness of the proposed expert system is verified by comparing the best-first search solution and local minimum tree search solution based on diversity event simulations for typical distribution substation.

• 한국멀티미디어학회논문지
• /
• 제18권12호
• /
• pp.1432-1438
• /
• 2015

We present a methodology to detect moving shadows in video sequences, which is considered as a challenging and critical problem in the most visual surveillance systems since 1980s. While most previous moving shadow detection methods used hand-crafted features such as chromaticity, physical properties, geometry, or combination thereof, our method can automatically learn features to classify whether image segments are shadow or foreground by using a deep learning architecture. Furthermore, applying Markov Random Field enables our system to refine our shadow detection results to improve its performance. Our algorithm is applied to five different challenging datasets of moving shadow detection, and its performance is comparable to that of state-of-the-art approaches.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권10호
• /
• pp.5161-5178
• /
• 2019

Correlation Filter was recently demonstrated to have good characteristics in the field of video object tracking. The advantages of Correlation Filter based trackers are reflected in the high accuracy and robustness it provides while maintaining a high speed. However, there are still some necessary improvements that should be made. First, most trackers cannot handle multi-scale problems. To solve this problem, our algorithm combines position estimation with scale estimation. The difference from the traditional method in regard to the scale estimation is that, the proposed method can track the scale of the object more quickly and effective. Additionally, in the feature extraction module, the feature representation of traditional algorithms is relatively simple, and furthermore, the tracking performance is easily affected in complex scenarios. In this paper, we design a novel and powerful feature that can significantly improve the tracking performance. Finally, traditional trackers often suffer from model drift, which is caused by occlusion and other complex scenarios. We introduce a relocation component to detect object at other locations such as the secondary peak of the response map. It partly alleviates the model drift problem.