Earticle

연구목적: 철도 안전에 영향을 미치는 콘크리트 궤도는 이미지분석 기술을 사용하여 균열 을 감지 할 수 있으나 균열을 검출하기 위한 콘크리트 궤도 및 표면 오염의 조건이 균열 검측에 방해되므로 이를 효과적으로 제거하기 위한 방법이 필요하다. 연구방법: 본 연구에서는 한국 철도의 균열을 효과적으로 감지하기 위한 이미지 분석 기법 을 적용한 프로세스를 제안하고 실험 모듈을 통해 취득된 이미지를 분석하여 성능을 검증 하였다. 또한, 우리는 제안된 Gabor Filter Bank 기법을 사용하여 철도 콘크리트 도상 이미 지를 획득한 데이터 중 무작위로 선택된 2000개의 이미지를 개발된 프로세스를 통해 자동 균열 검측을 수행하여 타당성을 검토하였다. 연구결과: 연구에서 제안된 시스템으로 균열 검측 결과 탐지율이 약 94% 성능으로 검토되 었으며 취득된 철도콘크리트도상이미지의 균열이 동일한 크기와 형식으로 일치하였다. 결론: Gabor Filter Bank를 사용한 균열 검측법은 한국 철도의 콘크리트 궤도도상에 노이즈 를 포함한 균열 이미지에 효과적으로 분석되는 것을 확인 할 수 있었다. 이 시스템은 기존 의 인간 위주의 철도 산업에서 자동화 된 유지 관리 시스템이 될 수 있을 것으로 기대된다.

Purpose: Concrete track that affects on railway safety can detect cracks using image processing technique. However, since a condition of concrete track and surface noisy are obstructed to detect cracks, there is a need for a way to remove them effectively. Method: In this study, we proposed an image processing to detect cracks effectively for Korean railway and verified its performance through experiment. We developed image acquisition system for capture a railway concrete track and acquired railway concrete track images, randomly selected 2000 images and detected cracks in the image process using proposed Gabor Filter Bank methods. Results: As a result, 94% of detection rate are matched to the actual cracks in same quality and format railway concrete track image. Conclution: The crack detection method using Garbor Filter Bank was confirmed to be effective for crack image including noise in the Korean railway concrete track. This system is expected to become an automated maintenance system in the existing human- centered railway industry

본 개발은 중간 크기 미만의 외형 결함을 검출함에 있어 X-Ray 광원을 사용하는 비파괴 장비이다. 본 연구개발물의 차별성은 X-Ray광원을 이용 중소형 생산제품들의 외부결함 유무를 실시간 검사할 수 있으며 표면 방사선 누설선량에 있으면서 검사 속도제어가 가능하다. 이를 소프트웨어로 해결하였다. 또한 작업 중 작업자 안전 을 위해 도어 차단이 가능하고 동시에 장비 내부 구동 영상의 획인이 가능하다는 점이다. 이런 원리로 현재는 중소 형 부품에서부터 크게는 컨테이너와 같은 검사시스템의 기반 기술로 활용이 가능하다. 본 연구는 자동차 산업 외에 다른 다양한 비파괴 시장 산업에 큰 영향을 미칠 것이다. 마지막으로 개발된 비파괴 장비는 0.2 μSv 하에서 제어된 다는 사실은 아주 중요한 사실이다.

This development is for non-destructive machine using X-Ray source about detecting outline faults of below middle size products. The differentiation is product of research and development unspecialized small and medium-sized products using X-Ray light sources can check real time if the surface of an external fault of radiation dose reference, within the leakage. The speed control is possible by software solution. In addition, we're working on possibly block doors for worker safety and equipment at the same time that inner drive can be identified in the image. These principles, as a key enabler of the current inspection system such as the container is small to medium-sized parts - a long way from utilization level is possible. This research will give rise to major effects for other various non-destructive market industries except car-industry. The most important fact is that this developed non-destructive machine is controlled below 0.2 micro-μSv.

In this study, we propose a system that enables beverage placement on a table using plane recognition and robot control. RGB-D sensors are used to extract 3D depth images around the robot, estimate the relative movement distance from the table, and simultaneously predict the height required when delivering objects through plane recognition. Considering the workspace of the robot, the moving distance and placement position of the robot were calculated, and the drive command of the robot was generated. The final task was performed using mobile manipulation based on the given object placement coordinates. To verify this, the sensor recognition and positioning accuracy of the robot and the success rate of the service performance through the action part were evaluated on a table with an open front and a step, without considering the workspace of the robot. The achievement of a 95% success rate in the given environment confirmed the feasibility of the proposed process for performing beverage delivery services for robots.

본 논문에서는 선분분류를 이용하여 실내영상의 소실점을 검출하는 방법을 제안한다. 실내영상을 효율적으로 검출하기 위하여 2 단계로 소실점을 추출한다. 1 단계에서는 이미지가 1 점 투시인지 2 점 투시인지 판별한다. 만일 이미지가 2 점 투시이면, 선분분류를 위하여 검출된 소실점을 지나는 수평선을 구한다. 2 단계에서는 선분분류를 이용하여 2 개의 소실점을 정확히 검출한다. 또 본 논문에서는 인공영상과 이미지 DB를 이용하여 제안한 방법을 평가하였다. 노이즈를 첨가한 인공 영상에 대해서는 노이즈가 60%를 차지할 때까지 검출한 소실점과 실제 소실점과의 차이가 16 픽셀 이하였다. A. Quattoni 와 A. Torralba가 제안한 이미지 DB를 이용한 평가에서는 87% 이상의 이미지에 대하여 소실점을 검출하였다.

This paper proposes a method to detect vanishing points of an indoor scene using line segment classification. Two-stage vanishing points detection is carried out to detect vanishing point in indoor scene efficiently. In the first stage, the method examines whether the image composition is a one-point perspective projection or a two-point one. If it is a two-point perspective projection, a horizontal line through the detected vanishing point is found for line segment classification. In the second stage, the method detects two vanishing points exactly using line segment classification. The method is evaluated by synthetic images and an image DB. In the synthetic image which some noise is added in, vanishing point detection error is under 16 pixels until the percent of the noise to the image becomes 60%. Vanishing points detection ratio by A.Quattoni and A.Torralba's image DB is over 87%.

전력선통신 시스템에서는 전력망을 활용한 통신 방식을 사용하기 때문에 일반적 통신선로를 활용한 통신 방식에 비해 잡음이 많고, 이것으로 인한 성능 저하가 문제가 되고 있다. 이러한 잡음으로 인한 성능 저하를 완화시키기 위해, 본 논문에서는 전력선통신 시스템에서의 임펄스 잡음 환경에서 신호를 검출하는 인공지능 알고리즘을 제안한다. 다항식 회귀법을 이용하여 임펄스 잡음 신호의 원신호를 예측하고, 시뮬레이션 결과를 통해 본 논문에서 제안한 인공지능 알고리즘을 적용한 전력선통신 시스템에서 임펄스 잡음 환경내 신호 검출 성능 향상을 입증한다.

It is known that power line communication systems have more noise than general wired communication systems due to the high voltage that flows in power line cables, and the noise causes a serious performance degradation. In order to mitigate performance degradation due to such noise, this paper proposes an artificial intelligence algorithm based on polynomial regression, which detects signals in the impulse noise environment in the power line communication system. The polynomial regression method is used to predict the original transmitted signal from the impulse noise signal. Simulation results show that the signal detection performance in the impulse noise environment of the power line communication is improved through the artificial intelligence algorithm proposed in this paper.

We propose a practical camera calibration system on mobile platforms to calibrate the camera’s intrinsic parameters, based on the geometrical property of the vanishing points, which are determined by two perpendicular groups of parallel lines. In this system, first, we detect and cluster the lines of the input images to compute the vanishing points. Then, we establish the intrinsic parameter constraint equations based on the perpendicular property of the vectors that connect the optical center and the vanishing points, and solve them using the least squares method. This system only requires the camera to observe a rectangular card shown in a few (at least four) different orientations. The experimental results of real images show that the proposed calibration system is robust, simple, and practicable.

P2P botnet has become a significant threats in security network. In this paper, we propose a new method to detection the P2P botnet through the analysis of the P2P botnet host’s life cycle, use the method of off-line detection to find the suspected botnet hosts, and determine the P2P botnet host through online monitoring method. In this way, the efficiency and the accurate rate of P2P botnet detection have raised then only use one method, and reduce the harm of the P2P botnet.

Trip-line detection is one of the very important research topics of intelligent monitoring category. This paper will present a trip-line detection system. In our trip line detection algorithm, Gauss model is used to model background, background differencing is used to detect object. Scanning the set of points on the line, if object is cross the line, for double-direction trip-line detection system directly alerts, but for single-direction trip-line detection, the moving direction of the object must be analyzed first. If the moving direction of the object violate passing rule, system directly alerts.

New voltage-based on-line fault detection and faulty switch identification algorithms under multiple open-switches in a grid-connected wind power converter are proposed in this paper. The proposed algorithms are based on the three-phase voltages which are calculated by using the DC link voltage and switching times determined from the space vector PWM scheme. From the calculated three-phase voltages, the absolute of average phase voltage and normalized three-phase RMS voltages are obtained to be used for the fault detection and faulty switch identification algorithms in three-phase AC/DC PWM converter. To determine faulty switches, three-phase voltages and RMS currents are taken as faulty switch localization variables. The complete diagnosis is carried out by a simple method during operation, which does not require any additional sensors. Therefore, the proposed methods are cost-effective and easy to use. To verify the validity and performance of the proposed algorithms, the simulation is carried out using the PSIM software for twenty-one cases of faults including multiple open-switches. As a result, the proposed scheme can effectively detect the occurrence of fault as well as faulty switches for these conditions independent of operating conditions.

Shape template matching is an important approach in object detection and recognition. In this paper, we propose a fast and novel method to represent edge maps by using salient line segments, which is used to detect objects based on shape template matching. Firstly, image edges are computed and, on these image edge fragments, corner points are extracted. Then, a parabola model is proposed to represent image edges. Secondly, based on salient points, a directional chamfer matching framework is used to compute the similarity between template image and corresponding locations in target image. Our method has two main contributions. One is we apply the way to represent contours with salient points to the framework, which is used to detect the target rapidly and accurately. The other is that our method can directly operate on real images, which improves its practicability. A series of experiments are performed on two benchmark data sets. Comparing with previous work, our method is much more time-saving and memory-saving.

In order to ensure the flow of goods available and more flexible, reduce labor costs, many factories and industrial zones around the world are gradually moving to use automated solutions. One of them is to use Automated guided vehicles (AGV). Currently, there are a line tracing method as an AGV operating method, and a method of estimating the current position of the AGV and matching with a factory map and knowing the moving direction of the AGV. In this paper, we propose ceiling Light and guided line based moving direction estimation algorithm using multi-camera on the AGV in smart factory that can operate stable AGV by compensating the disadvantages of existing AGV operation method. The proposed algorithm is able to estimate its position and direction using a general - purpose camera instead of a sensor. Based on this, it can correct its movement error and estimate its own movement path.

Recently, Microsoft unveiled a new operating system called Windows 10. As it is highly expected that Windows 10 will play a significant role in the Internet of Things era, the Portable Executable (PE) format is drawing attention even more widely than before. PE is a standard file format for executables and object code used in MS Windows operating systems. Since a number of various malwares have widely spread by exploiting vulnerabilities of PE structure, the need of automatic tools for PE-malware detection is being magnified. In this paper, we designed and developed a command-line PE file analysis tool using Python language for automatic detection of Windows malware.

An optical sensing membrane has been fabricated to measure the concentration of dissolved oxygen(DO) and pH value simultaneously. It has employed HPTS as a pH sensitive dye and a ruthenium(Ⅱ) complex as a DO sensitive dye. The sensing membrane has been applied to wells in a 24-well microtiter plate. Using the 24-well microtiter plate the concentrations of dissolved oxygen and pH values have been on-line monitored during the cultivations of E.coli DH5α, B.cereus 318 and P.pastoris X-33. On-line monitoring of DO and pH in microorganism cultivation processes showed good performance of the sensing membrane containing 5 mM HPTS and 2 or 5 mg/mL Rudpp.

Image Registration is used many works in image processing widely. But It is difficult to find the accuracy informations such as translation, rotation, and scaling between images. This paper proposes an algorithm that Fourier based image registration using the pyramid edge detection and line fitting. It can be estimated the informations by each sub-pixels. The proposed algorithm can be used for image registrations which high efficiency is required such as GIS, or MRI, CT, image mosaicing, weather forecasting, etc.

A method for faulted line detection of four parallel lines on the same tower is presented, based on four-summing and double-differential sequences of one terminal current. Four-summing and double-differential sequences of fault current can be calculated using a certain transformation matrix for parameter decoupling of four parallel transmission lines. According to fault boundary conditions, the amplitude and phase characteristics of four-summing and double-differential sequences of fault current is studied under conditions of different types of fault. Through the analysis of the relationship of terminal current and fault current, a novel methodology for fault line detection of four parallel transmission line on the same tower is put forward, which can pick out the fault lines no matter the fault occurs in single line or cross double lines. Simulation results validate that the methodology is correct and reliable under conditions of different load currents, transient resistances and fault locations.

In this paper, we propose the horizontal line detection using the Haar-like features and linear regression in infrared images. In the marine environment horizon image is very useful information on a variety of systems. In the proposed method Haar-like features it was noted that the standard deviation be calculated in real time on a static area. Based on the pixel position, calculating the standard deviation of the around area in real time and, if the reaction is to filter out the largest pixel can get the energy map of the area containing the straight horizontal line. In order to select a horizontal line of pixels from the energy map, we applied the linear regression, calculating a linear fit to the transverse horizontal line across the image to select the candidate optimal horizontal. The proposed method was carried out in a horizontal line detecting real infrared image experiment for day and night, it was confirmed the excellent detection results than the legacy methods.

Monitoring of the drill wear and hole quality change is conducted during the drilling process. Cutting force measured by tool dynamometer is a evident feature estimating abnormal state of drilling. One major difficulty in using tool dynamometer is that the work piece must be mounted on the dynamometer, and thus the machining process is disturbed and discontinuous. Acoustic transducer do not disturb the normal machining process, and provide a relatively easy way to monitor a machining process for industrial application. For this advantage, AE signal is used to estimate the abnormal state. In this study vision system is used to detect flank wear tendency and hole quality, there are many formal factors in hole quality decision circularity, cylindricity, straightness, and so on, but these are difficult to measure in on-line monitoring. The movement of hole center and increasement of hole diameter is presented to determine hole quality As the results of this experiment, AE RMS signal and measurements by vision system are shown the similar tendency as abnormal state of drilling. And detection of the abnormal states using BPNs was achieved 96.4% reliability.

The appearance of retinal blood vessels is an important diagnostic indicator of serious disease, such as hypertension, diabetes, cardiovascular disease, and stroke. Automatic segmentation of the retinal vasculature is a primary step towards automatic assessment of the retinal blood vessel features. This paper presents an automated method for the enhancement and segmentation of blood vessels in fundus images. To decrease the influence of the optic disk, and emphasize the vessels for each retinal image, a multidirectional morphological top-hat transform with rotating structuring elements is first applied to the background homogenized retinal image. Then, an improved multiscale line detector is presented to produce a vessel response image, and yield the retinal blood vessel tree for each retinal image. Since different line detectors at varying scales have different line responses in the multiscale detector, the line detectors with longer length produce more vessel responses than the ones with shorter length; the improved multiscale detector combines all the responses at different scales by setting different weights for each scale. The methodology is evaluated on two publicly available databases, DRIVE and STARE. Experimental results demonstrate an excellent performance that approximates the average accuracy of a human observer. Moreover, the method is simple, fast, and robust to noise, so it is suitable for being integrated into a computer-assisted diagnostic system for ophthalmic disorders.

Background: Resistance of Mycobacterium tuberculosis to anti-tuberculosis (TB) drugs is almost exclusively due to spontaneous chromosomal mutations in target genes. Rapid detection of drug resistance to both first- and second-line anti-TB drugs has become a key component of TB control programs. Technologies that allow rapid, cost-effective, and high-throughput detection of specific nucleic acid sequences are needed. This study was to develop a high-throughput assay based on allele-specific primer extension (ASPE) and MagPlex-TAG microspheres to detect anti-TB drug resistance mutations. Methods: DNA samples from 357 M. tuberculosis clinical isolates and H37Rv were amplified by multiplex PCR using four primer sets, followed by multiplex ASPE using 23 TAG-ASPE primers. The products were sorted on the TAG-ASPE array and detected by using the Luminex xMAP system. Genotypes were also determined by sequencing. Results: Genetic drug susceptibility typing by the TAG-ASPE method was 100% concordant with those obtained by sequencing. Compared with phenotypic drug susceptibility testing (DST) as a reference method, the sensitivity and specificity of the TAG-ASPE method were 83% (95% confidence interval [CI], 79-88%) and 97% (95% CI, 90-100%) for isoniazid. For rifampin testing, the sensitivity and specificity were 90% (95% CI, 86-93%) and 100% (95% CI, 99-100%). Also, the sensitivity and specificity were 58% (95% CI, 51-65%) and 86% (95% CI, 79-93%) for ethambutol. Conclusions: This study demonstrated the TAG-ASPE method is suitable for highly reproducible, cost-effective, and high-throughput clinical genotyping applications.

In this paper, we propose the augmented reality(AR) based driving navigation based on robust lane detection method to dynamic environment changes. The proposed technique uses the detected lane position as a marker which is a key element for enhancing driving information. We propose Symmetrical Local Threshold(SLT) algorithm which is able to robustly detect lane to dynamic illumination environment change such as shadows. In addition, by using Morphology operation and Connected Component Analysis(CCA) algorithm, it is possible to minimize noises in the image, and Region Of Interest(ROI) is defined through region division using a straight line passing through several vanishing points We also propose the augmented reality aided visualization method for Interchange(IC) and driving navigation using reference point detection based on the detected lane coordinates inside and outside the ROI. Validation experiments were carried out to assess the accuracy and robustness of the proposed system in vairous environment changes. The average accuracy of the proposed system in daytime, nighttime, rainy day, and cloudy day is 79.3% on 4600 images. The results of the proposed system for AR based IC and driving navigation were also presented. We are hopeful that the proposed research will open a new discussion on AR based driving navigation platforms, and thus, that such efforts will enrich the autonomous vehicle services in the near future.