• Tunnel and Underground Space
• /
• v.31 no.5
• /
• pp.374-384
• /
• 2021

In this study, we propose a new approach for automatic fracture detection in CT scan images of rock specimens. This approach is built on top of two-stage object detection deep learning algorithm called Faster R-CNN with a major modification of using rotated bounding box. The use of rotated bounding box plays a key role in the future work to overcome several inherent difficulties of fracture segmentation relating to the heterogeneity of uninterested background (i.e., minerals) and the variation in size and shape of fracture. Comparing to the commonly used bounding box (i.e., axis-align bounding box), rotated bounding box shows a greater adaptability to fit with the elongated shape of fracture, such that minimizing the ratio of background within the bounding box. Besides, an additional benefit of rotated bounding box is that it can provide relative information on the orientation and length of fracture without the further segmentation and measurement step. To validate the applicability of the proposed approach, we train and test our approach with a number of CT image sets of fractured granite specimens with highly heterogeneous background and other rocks such as sandstone and shale. The result demonstrates that our approach can lead to the encouraging results on fracture detection with the mean average precision (mAP) up to 0.89 and also outperform the conventional approach in terms of background-to-object ratio within the bounding box.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.26 no.2
• /
• pp.129-152
• /
• 2024

In the application of deep learning object detection via CCTV in tunnels, a large number of false positive detections occur due to the poor environmental conditions of tunnels, such as low illumination and severe perspective effect. This problem directly impacts the reliability of the tunnel CCTV-based accident detection system reliant on object detection performance. Hence, it is necessary to reduce the number of false positive detections while also enhancing the number of true positive detections. Based on a deep learning object detection model, this paper proposes a false positive data training method that not only reduces false positives but also improves true positive detection performance through retraining of false positive data. This paper's false positive data training method is based on the following steps: initial training of a training dataset - inference of a validation dataset - correction of false positive data and dataset composition - addition to the training dataset and retraining. In this paper, experiments were conducted to verify the performance of this method. First, the optimal hyperparameters of the deep learning object detection model to be applied in this experiment were determined through previous experiments. Then, in this experiment, training image format was determined, and experiments were conducted sequentially to check the long-term performance improvement through retraining of repeated false detection datasets. As a result, in the first experiment, it was found that the inclusion of the background in the inferred image was more advantageous for object detection performance than the removal of the background excluding the object. In the second experiment, it was found that retraining by accumulating false positives from each level of retraining was more advantageous than retraining independently for each level of retraining in terms of continuous improvement of object detection performance. After retraining the false positive data with the method determined in the two experiments, the car object class showed excellent inference performance with an AP value of 0.95 or higher after the first retraining, and by the fifth retraining, the inference performance was improved by about 1.06 times compared to the initial inference. And the person object class continued to improve its inference performance as retraining progressed, and by the 18th retraining, it showed that it could self-improve its inference performance by more than 2.3 times compared to the initial inference.

• Journal of the HCI Society of Korea
• /
• v.10 no.2
• /
• pp.13-19
• /
• 2015

Display technology has recently made enormous progress. In particular, display companies are competing each other to develop flexible display. Curved display, as a precursor of flexible display, are now used for smart phones and TVs. Curved monitors have been just introduced in the market, and are used for office work or entertainment. The aim of the current study was to investigate whether the curvature of a 42" multi-monitor affects postural control when it is used for entertainment purpose. The current study used two curvature levels (flat and 600mm). Ten college students [mean(SD) age = 20.9 (1.5)] with at least 20/25 visual acuity, and without color blindness and musculoskeletal disorders participated in this study. In a typical VDT environment, each participant played a car racing video game using a steering wheel and pedals for 30 minutes at each curvature level. During the video game, a pressure mat on the seat pan measured the participant's COP (Center of Pressure), and from which four measures (Mean Velocity, Median Power Frequency, Root-Mean-Square Distance, and 95% Confidence Ellipse Area) were derived. A larger AP (Anterior-Posterior) RMS distance was observed in the flat condition, indicating more forward-backward upper body movements. It can be partly due to more variability in visual distance across display, and hence longer ocular accommodation time in the case of the flat display. In addition, a different level of presence or attention between two curvature conditions can lead to such a difference. Any potential effect of such a behavioral change by display curvature on musculoskeletal disorders should be further investigated.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.3
• /
• pp.95-105
• /
• 2019

In recent years, image processing techniques for detecting road surface damaged spot have been actively researched. Especially, it is mainly used to acquire images through a smart phone or a black box that can be mounted in a vehicle and recognize the road surface damaged region in the image using several algorithms. In addition, in conjunction with the GPS module, the exact damaged location can be obtained. The most important technology is image processing algorithm. Recently, algorithms based on artificial intelligence have been attracting attention as research topics. In this paper, we will also discuss artificial intelligence image processing algorithms. Among them, an object detection method based on an region-based convolution neural networks method is used. To improve the recognition performance of road surface damage objects, 600 road surface damaged images and 1500 general road driving images are added to the learning database. Also, supervised learning using background object recognition method is performed to reduce false alarm and missing rate in road surface damage detection. As a result, we introduce a new method that improves the recognition performance of the algorithm to 8.66% based on average value of mAP through the same test database.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.7
• /
• pp.1106-1115
• /
• 2021

In this study, a fin that controls ship stern flow was attached on stern hull of a 80k bulk carrier to improve resistance performance. The rectangular cross-sectional fin was attached at several locations on the hull, and angle to streamline was changed with constant length, breadth, and thickness. The resistance performance and wake on propeller plane of the hull with and without the fin were analyzed using model-scale computational fluid dynamics simulation. The analysis results were extrapolated to full-scale to compare the performance and wake of the full-scale ship. First, the fin changed path of bilge vortex that flowed into the propeller along the stern hull without the fin to transom stern. This change increased pressure of the stern hull and upper region of the propeller, so pressure resistance and total resistance of the hull were reduced - the nearer the fin location to after perpendicular (AP) and base line of the hull, the larger the reduction of the resistances. Second, nominal wake fraction of the hull with the fin was lower than that without the fin. This dif erence was in proportion to the angle of the fin, but the total resistance reduction was in proportion until a certain angle at which the reduction was maximum. The largest total resistance reduction was approximately 2.1% at 12.5% of length between perpendiculars from the AP, 10% of draft from the base line, and 14° with respect to the streamline.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.5
• /
• pp.211-220
• /
• 2022

Recently, research on smart factories triggered by the 4th industrial revolution is being actively conducted. Accordingly, the manufacturing industry is conducting various studies to improve productivity and quality based on deep learning technology with robust performance. This paper is a study on the method of detecting tire surface defects in the visual inspection stage of the tire manufacturing process, and introduces a tire surface defect detection method using a depth image acquired through a 3D camera. The tire surface depth image dealt with in this study has the problem of low contrast caused by the shallow depth of the tire surface and the difference in the reference depth value due to the data acquisition environment. And due to the nature of the manufacturing industry, algorithms with performance that can be processed in real time along with detection performance is required. Therefore, in this paper, we studied a method to normalize the depth image through relatively simple methods so that the tire surface defect detection algorithm does not consist of a complex algorithm pipeline. and conducted a comparative experiment between the general normalization method and the normalization method suggested in this paper using YOLO V3, which could satisfy both detection performance and speed. As a result of the experiment, it is confirmed that the normalization method proposed in this paper improved performance by about 7% based on mAP 0.5, and the method proposed in this paper is effective.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.65 no.1
• /
• pp.128-134
• /
• 2016

IoT devices including smart devices are connected with internet, thus they have security threats everytime. Particularly, IoT devices are composed of low performance MCU and small-capacity memory because they are miniaturized, so they are likely to be exposed to various security threats like DoS attacks. In addition, in case of IoT devices installed for a remote place, it's not easy for users to control continuously them and to install immediately security patch for them. For most of IoT devices connected directly with internet under user's intention, devices exposed to outside by setting IoT gateway, and devices exposed to outside by the DMZ function or Port Forwarding function of router, specific protocol for IoT services was used and the devices show a response when services about related protocol are required from outside. From internet search engine for IoT devices, IP addresses are inspected on the basis of protocol mainly used for IoT devices and then IP addresses showing a response are maintained as database, so that users can utilize related information. Specially, IoT devices using HTTP and HTTPS protocol, which are used at usual web server, are easily searched at usual search engines like Google as well as search engine for the sole IoT devices. Ill-intentioned attackers get the IP addresses of vulnerable devices from search engine and try to attack the devices. The purpose of this study is to find the problems arisen when HTTP, HTTPS, CoAP, SOAP, and RestFUL protocols used for IoT devices are detected by search engine and are maintained as database, and to seek the solution for the problems. In particular, when the user ID and password of IoT devices set by manufacturing factory are still same or the already known vulnerabilities of IoT devices are not patched, the dangerousness of the IoT devices and its related solution were found in this study.

• KEPCO Journal on Electric Power and Energy
• /
• v.6 no.1
• /
• pp.29-33
• /
• 2020

Despite the same system architecture and operation principle, Advanced Metering Infrastructure (AMI) and Internet of Things (IoT) are recognized as a heterogeneous system. This is due to the different object modeling and communication protocols used in smart meters and sensors. However, data interworking between AMI and IoT is expected to be inevitable in the future. In this paper, we propose Device Language Message Specification (DLMS) to Lightweight Machine to Machine (LwM2M) conversion model. The proposed interworking model can reduce the packet size by 46.5% compared to that of the encapsulation method.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.2
• /
• pp.466-474
• /
• 2014

According to wireless network technology and mobile processors performance were improved, the small wireless mobile device such as smart phones has been widely utilized. The mobile devices can be used GPS information, thereby the services based on location information was increased. GPS was impossible to provide location information in indoor and signal shading environment, and the tracking systems based on short distance wireless communication are required infrastructure. The IEEE 802.11 based tracking system is possible estimation using APs, however the tracking device is exhausted battery power seriously. In this paper, we propose IEEE 802.11 based low-power tracking system. We reduced power consumption from channel scanning and network connection. For performance evaluation, we designed and implemented the tracking tag device, and measured power consumption of the device. As the simulation result, we confirmed that the power consumption was reduced 46% compare to the standard execution.

• Nutrition Research and Practice
• /
• v.13 no.6
• /
• pp.473-479
• /
• 2019

BACKGROUND/OBJECTIVES: Anti-inflammatory and antioxidative activities of luteolin and luteolin-7-O-glucoside were compared in galactosamine (GalN)/lipopolysaccharide (LPS)-induced hepatitic ICR mice. MATERIALS/METHODS: Male ICR mice (6 weeks old) were divided into 4 groups: normal control, GalN/LPS, luteolin, and luteolin-7-O-glucoside groups. The latter two groups were administered luteolin or luteolin-7-O-glucoside (50 mg/kg BW) daily by gavage for 3 weeks after which hepatitis was induced by intraperitoneal injection of GalN and LPS (1 g/kg BW and $10{\mu}g/kg\;BW$, respectively). RESULTS: GalN/LPS produced acute hepatic injury by a sharp increase in serum AST, ALT, and $TNF-{\alpha}$ levels, increases that were ameliorated in the experimental groups. In addition, markedly increased expressions of cyclooxygenase (COX)-2 and its transcription factors, nuclear factor $(NF)-{\kappa}B$ and activator protein (AP)-1, were also significantly attenuated in the experimental groups. Compared to luteolin-7-O-glucoside, luteolin more potently ameliorated the levels of inflammatory mediators. Phase II enzymes levels and NF-E2 p45-related factor (Nrf)-2 activation that were decreased by GalN/LPS were increased by luteolin and luteolin-7-O-glucoside administration. In addition, compared to luteolin, luteolin-7-O-glucoside acted as a more potent inducer of changes in phase II enzymes. Liver histopathology results were consistent with the mediator and enzyme results. CONCLUSION: Luteolin and luteolin-7-O-glucoside protect against GalN/LPS-induced hepatotoxicity through the regulation of inflammatory mediators and phase II enzymes.