• Journal of Korea Multimedia Society
• /
• v.20 no.8
• /
• pp.1271-1281
• /
• 2017

It is essential to collect and analyze target information around the vehicle for autonomous driving of the vehicle. Based on the analysis, environmental information such as location and direction should be analyzed in real time to control the vehicle. In particular, obstruction or cutting of objects in the image must be handled to provide accurate information about the vehicle environment and to facilitate safe operation. In this paper, we propose a method to simultaneously generate 2D and 3D bounding box proposals using LiDAR Edge generated by filtering LiDAR sensor information. We classify the classes of each proposal by connecting them with Region-based Fully-Covolutional Networks (R-FCN), which is an object classifier based on Deep Learning, which uses two-dimensional images as inputs. Each 3D box is rearranged by using the class label and the subcategory information of each class to finally complete the 3D bounding box corresponding to the object. Because 3D bounding boxes are created in 3D space, object information such as space coordinates and object size can be obtained at once, and 2D bounding boxes associated with 3D boxes do not have problems such as occlusion.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.29 no.6
• /
• pp.591-602
• /
• 2009

The Compton backscatter technique has been applied to lap-joint in aircraft structure in order to determine mass loss due to exfoliative corrosion of the aluminum alloy sheet skin. The mass loss of each layer has been estimated from Compton backscatter A-scan including the aluminum sheet, the corrosion layer, and the sealant. A Compton backscattering imaging system has been also developed to obtain a cross-sectional profile of corroded lap-splices of aging aircraft using a specially designed slit-type camera. The camera is to focus on a small scattering volume inside the material from which the backscattered photons are collected by a collimated scintillator detector for interpretation of material characteristics. The cross section of the layered structure is scanned by moving the scattering volume through the thickness direction of the specimen. The theoretical model of the Compton scattering based on Boltzmann transport theory is presented for quantitative characterization of exfoliative corrosion through deconvolution procedure using a nonlinear least-square error minimization method. It produces practical information such as location and width of planar corrosion in layered structures of aircraft, which generally cannot be detected by conventional NDE techniques such as the ultrasonic method.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.28 no.10
• /
• pp.1467-1474
• /
• 2004

In this study, a new damage identification method for beam-like structures with a fatigue crack is proposed. which does not require comparative measurement on an intact structure but require several measurements at different level of excitation forces on the cracked structure. The idea comes from the fact that dynamic behavior of a structure with a fatigue crack changes with the level of the excitation force. The 2$^{nd}$ spatial derivatives of frequency response functions along the longitudinal direction of a beam are used as the sensitive indicator of crack existence. Then, weighting function is employed in the averaging process in frequency domain to account for the modal participation of the differences between the dynamic behavior of a beam with a fatigue crack at the low excitation and one at the high excitation. Subsequently, a damage index is defined such that the location and level of the crack may be identified. It is shown from the analysis of vibration measurements in this study that comparison of frequency response characteristics of a beam with a single fatigue crack at different level of excitation forces enables an effective detection of the crack.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.54 no.7
• /
• pp.345-351
• /
• 2005

An out-of-step condition results from the loss of the synchronism of the generators. A disturbance in a power system causes the generator angle to oscillate. When there is a severe disturbance such as a heavy current fault loss of major generation or loss of a large block of load, the oscillation can be severe and even increase largely and finally the out-of-step condition may un. During the power swing and out-of-step conditions, the apparent impedance at a relay location changes, and the power flow also changes as the angle difference is varied. This paper presents a method to analyze the trajectory of complex power during a power swing and out-of-step condition. The trajectory of the complex power is analyzed when a power swings and a fault occurs. Moreover, the complex power is analyzed when the ratios between the voltages at both sides and the line impedances are changed. These methods are verified through simulation using the ATP/EMTP MODELS.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.54 no.9
• /
• pp.449-456
• /
• 2005

The preventive diagnostic technique prevents transformers from power failure through giving alarm and observing transformers in service. And it helps to establish the plan for optimum maintenance of the transformer as well as to find location or cause of fault using accumulated data. Data detection and experience of the preventive diagnostic system need to establish the preventive diagnostic algorithm regarding interrelationship between detected data and deterioration of equipment. Therefore in-depth analysis about the preventive diagnosis system is required. KEPCO has adopted the preventive diagnostic system at nine 345kV substations since 1997. Techniques for component sensors of the preventive diagnosis system were settled but diagnosis algorithm, diagnostic criteria and practical use of accumulated data are not yet established. This paper, to build up the base of preventive diagnostic algorithm for the Power transformer. investigated the preventive diagnostic criteria for the power transformer.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.1
• /
• pp.101-109
• /
• 2011

This paper describes the attempts to improve the performance of an image matching method for multiple image. Typically, matching between two images is performed by using correlation between a reference and corresponding images. The proposed multiple image matching algorithm performs matching in an object space, chooses the image closest to the true vertical image as a reference image, calculates the correlation based on the chosen reference image. The algorithm also detects occluded regions automatically and keep them from matching. We could find that it is possible to create high quality DEM by this method, regardless of the location of image. From the performance improvement experiments through the occlusion detection, we could confirm the possibility of a more accurate representation of 3D information.

• Journal of Power System Engineering
• /
• v.15 no.1
• /
• pp.18-23
• /
• 2011

Image processing has provided powerful techniques to extract from the acoustic signals the desired information on evaluation for leakage existence, leakage rate, and searching for leakage location, etc. The imagery NDE data available can add additional and significant dimension in nondestructive evaluation(NDE) information and thus for exploiting in applications. To extract such information the use of advanced image processing techniques is much needed. In recent years, there has been much increased use of acoustic signal image processing techniques in acoustic NDE. This approach will increase the efficiency of inspection procedures and reduce inspection time. In this paper we are concerned only with This paper is concerned mainly with the use of advanced image processing techniques in valve leakage detection and advanced image restoration and enhancement methods, which attempt to evaluate promptly by a visualization method the acoustic sources while detecting the valve leakage.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.542-546
• /
• 2009

This paper introduces a Mixed-Reality based Software technology in Smart-Phone Environment. The field of Mixed-Reality in mobile environment is relatively young. but Cause to develop Mobile infra and improvement of mobile device, open-platform mobile OS, the request extended This paper suggest the method for Marker Detection and Marker Tracking method. This method is the one of some kind of a base-technology in Mixed Reality. this method is to effect to location and registration. This paper suggest the method in low CPU computing power. Using a deformable searching area, the method improve computing power. and Using a Cam-shift algorithm, we suggest a calibration free method.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.28 no.2
• /
• pp.70-75
• /
• 2014

A track circuit is a critical piece of equipment that allows the existence of the trains on the track to be detected. Train detection is very important for the signal safety equipment. It allows for the confirmation of a train's location, the locking of switch points, the coordination of safe distance between trains, and the advance notice of a train's arrival. Track circuits can be powered by AC or DC. The DC track circuits are usually used on non-electrified tracks. On these tracks, many signal errors can be caused by lightning or problems with the power source. These problems can also cause damages to the rectifiers which must be repaired or replaced, promptly. This issue is especially problematic in the summer because of the higher frequency of thunderstorms. Issues with track circuits also cause problems for other equipment, such as railroad crossings and switch points. This further disrupts the safe operation of trains. This study aims to enhance maintenance efficiency and improve safety by utilizing parallel operation and multiplexing of a DC track circuit as well as installing an external surge protector. The experimental results on the operation of the proposed method is also presented.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.19 no.4
• /
• pp.443-452
• /
• 2016

Current low-altitude radar system often fails to detect small unmanned aerial vehicles (UAV) because of their small radar cross section (RCS) compared with larger targets. As a potential alternative, a passive bistatic radar system has been considered. We study an optimal deployment problem for the passive bistatic radar system. We model this problem as a covering problem, and develop an integer programming model. The objective of the model is to maximize coverage of a passive bistatic radar system. Our model takes into account factors specific to a bistatic radar system, including bistatic RCS and transmitter-receiver pair coverage. Considering bistatic RCS instead of constant RCS is important because the slight difference of RCS value for small UAVs could significantly influence the detection probability. The paired radar coverage is defined by using the concept of gradual coverage and cooperative coverage to represent a realistic environment.