• 한국정밀공학회지
• /
• 제29권5호
• /
• pp.554-562
• /
• 2012

Processing of Scanning electron microscope imaging has been analyzed in both secondary electron (SE) imaging and backscattered electron (BSE) image. Because of unique characteristics of both secondary electron and backscattered electron image, mechanism of imaging process and image quality are quite different each other. For the sake of characterize imaging process, Monte Carlo simulation code have been developed. It simulates electron penetration and depth profile in certain material. In addition, secondary electron and backscattered electron generation process as well as their spatial distribution and energy characteristics can be simulated. Geometries that has fundamental feature have been imaged using the developed Monte Carlo code. Two, SE and BSE images generation process will be discussed. BSE imaging process can be readily used to discriminate in both material and geometry by simply changing position and direction of BSE detector. The developed MC code could be useful to design BSE detector and their position. Furthermore, surface reconstruction technique is possibly developed at the further research efforts. Basics of Monte Carlo simulation method will be discussed as well as characteristics of SE and BSE images.

• 제어로봇시스템학회논문지
• /
• 제20권9호
• /
• pp.930-935
• /
• 2014

A Gabor filter is a linear filter used for edge detectionas frequency and orientation representations of Gabor filters are similar to those of the human visual system. In this thesis, we propose a pedestrian detection algorithm using a Gabor filter bank. In order to extract the features of the pedestrian, we use various image processing algorithms and data structure algorithms. First, color image segmentation is performed to consider the information of the RGB color space. Second, histogram equalization is performed to enhance the brightness of the input images. Third, convolution is performed between a Gabor filter bank and the enhanced images. Fourth, statistical values are calculated by using the integral image (summed area table) method. The calculated statistical values are used for the feature matrix of the pedestrian area. To evaluate the proposed algorithm, the INRIA pedestrian database and SVM (Support Vector Machine) are used, and we compare the proposed algorithm and the HOG (Histogram of Oriented Gradient) pedestrian detector, presentlyreferred to as the methodology of pedestrian detection algorithm. The experimental results show that the proposed algorithm is more accurate compared to the HOG pedestrian detector.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권8호
• /
• pp.4285-4299
• /
• 2019

Adversarial attack is a technique that causes a malfunction of classification models by adding noise that cannot be distinguished by humans, which poses a threat to a deep learning model. In this paper, we propose an efficient method to detect adversarial images using Gaussian process regression. Existing deep learning-based adversarial detection methods require numerous adversarial images for their training. The proposed method overcomes this problem by performing classification based on the statistical features of adversarial images and clean images that are extracted by Gaussian process regression with a small number of images. This technique can determine whether the input image is an adversarial image by applying Gaussian process regression based on the intermediate output value of the classification model. Experimental results show that the proposed method achieves higher detection performance than the other deep learning-based adversarial detection methods for powerful attacks. In particular, the Gaussian process regression-based detector shows better detection performance than the baseline models for most attacks in the case with fewer adversarial examples.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제4권5호
• /
• pp.225-230
• /
• 2015

본 논문에서는 키 프레임 영상과 SURF 특징점 기반의 시각 단어들을 이용한 효과적인 실시간 시각 루프 결합 탐지기 VILODE를 제안한다. 시각 루프 결합 탐지기는 과거에 지나온 위치들 중 하나를 다시 재방문하였는지를 판단하기 위해, 새로운 입력 영상을 이미 지나온 위치들에서 수집한 과거 영상들과 모두 비교해보아야 한다. 따라서 새로운 위치나 장소를 방문할수록 비교 대상 영상들이 계속해서 증가하기 때문에, 일반적으로 루프 결합 탐지기는 실시간 제약과 높은 탐지 정확도를 동시에 만족하기 어렵다. 이러한 문제점을 극복하기 위해, 본 시스템에서는 입력 영상들 중에서 의미 있는 것들만을 선택해 이들만을 비교하는 효과적인 키 프레임 선택 방법을 채택하였다. 따라서 루프 탐지에 필요한 영상 비교를 대폭 줄일 수 있다. 또한 본 시스템에서는 루프 결합 탐지의 정확도와 효율성을 높이기 위해, 키 프레임 영상들을 시각 단어들로 표현하고, DBoW 데이터베이스 시스템을 이용해 키 프레임 영상들에 대한 색인을 구성하였다. TUM 대학의 벤치마크 데이터들을 이용한 실험을 통해, 본 논문에서 제안한 시각 루프 결합 탐지기의 높은 성능을 확인할 수 있었다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1996년도 한국자동제어학술회의논문집(국내학술편); 포항공과대학교, 포항; 24-26 Oct. 1996
• /
• pp.117-120
• /
• 1996

In this paper, a new approach to acquire the cross-sentional image for automatic solder joint inspection of double-sided PCB using X-ray source is presented. We designed the apparatus with fixed X-ray source to realize the cross-sectional image by tunning object and detector synchronously. The cross-sectional images are captured at several view angle of X-ray source, the geometric image distortions caused by view angle and the shape of image intensifier are compensated. The precision variation of cross-sectional image by the change of view angle was investigated. Also we acquired the cross-sectional image to the solder joint of double-sided PCB and reconstructed the shape of solder joint.

• Journal of Information Processing Systems
• /
• 제1권1호
• /
• pp.27-31
• /
• 2005

Vehicle detector system based on image processing technology is a significant domain of ITS (Intelligent Transportation System) applications due to its advantages such as low installation cost and it does not obstruct traffic during the installation of vehicle detection systems on the road[1]. In this paper, we propose architecture for vehicle detection by using image processing. The architecture consists of two main parts such as an image processing part, using high speed FPGA, decision and calculation part using CPU. The CPU part takes care of total system control and synthetic decision of vehicle detection. The FPGA part assumes charge of input and output image using video encoder and decoder, image classification and image memory control.

• Nuclear Engineering and Technology
• /
• 제49권4호
• /
• pp.776-780
• /
• 2017

To avoid imaging artifacts and interpretation mistakes, an improvement of the uniformity in gamma camera systems is a very important point. We can expect excellent uniformity using cadmium zinc telluride (CZT) photon counting detector (PCD) because of the direct conversion of the gamma rays energy into electrons. In addition, the uniformity performance such as integral uniformity (IU), differential uniformity (DU), scatter fraction (SF), and contrast-to-noise ratio (CNR) varies according to the energy window setting. In this study, we compared a PCD and conventional scintillation detector with respect to the energy windows (5%, 10%, 15%, and 20%) using a $^{99m}Tc$ gamma source with a Geant4 Application for Tomography Emission simulation tool. The gamma camera systems used in this work are a CZT PCD and NaI(Tl) conventional scintillation detector with a 1-mm thickness. According to the results, although the IU and DU results were improved with the energy window, the SF and CNR results deteriorated with the energy window. In particular, the uniformity for the PCD was higher than that of the conventional scintillation detector in all cases. In conclusion, our results demonstrated that the uniformity of the CZT PCD was higher than that of the conventional scintillation detector.

• 한국전자파학회논문지
• /
• 제21권3호
• /
• pp.278-284
• /
• 2010

본 논문에서는 저잡음 증폭기와 무바이어스 쇼트키 다이오드를 사용하여 W-Band에서 동작하는 밀리미터파 다이오드 검출기를 설계, 제작하였다. 향상된 감도 특성을 얻기 위해 다이오드 검출기는 검출기와 저잡음 증폭기로 구성된다. 저잡음 증폭기단은 발진을 제거하기 위해, 높은 저지 대역 특성을 갖도록 설계된 하우징에 저잡음 증폭기 MMIC 칩을 사용해 제작한다. 다이오드 검출기단은 단순한 구조를 사용하기 위해 무바이어스 쇼트기 다이오드를 사용하고, 저잡음 증폭기단과의 연결을 용이하도록 하기 위해 평면형으로 설계한다. 설계 및 제작된 W-band 검출기는 입력 전력 -45~-20 dBm의 변화에 대하여 20~500 mV의 출력 전압을 얻었다. 본 검출기는 수동 밀리미터파 영상 시스템에 적용할 수 있다.

• 방사선산업학회지
• /
• 제9권2호
• /
• pp.85-90
• /
• 2015

Purpose: Time of operation has been reduced and accuracy of operation has been improved since C-arm, which offer real-time image of patient, was introduced in operation room. However, because of the contamination of patient, C-arm could not be used more appropriately. Therefore, this study is to know factors of controlling exposure dose, image quality and the exposed dose of health professional in operation room. Materials and methods: Height of Wilson frame (bed for operation) was fixed at 130 cm. Then, Model 76-2 Phantom, which was set by assembling manual of Fluke Company, was set on the bed. Head/Spine Fluoroscopy AEC mode was set for exposure condition. According to detector size of C-arm, the absorbed dose per min was measured in the 7 steps OFD (cm) from 10 cm to 40 cm (10, 15, 20, 25, 30, 35, 40 cm). In each step of OFD, the absorbed dose per min of same diameter of collimation was measured. Moreover, using Nero MAX Model 8000, exposure dose per min was measured according to 3 step of distance from detector (20 cm, 60 cm, 100 cm). Finally, resolution was measured by CDRH Disc Phantom and magnification of each OFD was measured by aluminum stick bar. Result: According to detector size of C-arm, difference of absorbed dose shows that the dose of 20 cm OFD is 1.750 times higher than the dose of 40 cm OFD. It means that the C-arm, which has smaller size of detector, shows the bigger difference of absorbed dose per min (p<0.05). In the difference of absorbed dose in the same step of OFD (from 20 cm to 40 cm), the absorbed dose of 9 inch detect or C-arm was 1.370 times higher than 12 inch' s (p<0.05). When OFD was set to 20 cm OFD, the absorbed dose of non-collimation case was approximately 0.816 times lower than the absorbed dose of collimation cases (p<0.05). When the distance was 20 cm from detector, exposed does includes first-ray and scatter-ray. When the distance was 60 cm and 100 cm from detector, exposed does includes just scatter-ray. So, there was the 2.200 times difference of absorbed does. Finally, when OFD was increased, spatial resolution was 4 to 5 step was increased. However, low contrast resolution was not relative. Moreover, there was 1.363 times difference of magnification (p<0.05). Conclusion: When C-Arm is used, avoiding contamination of patient is more important factor than reducing exposed dose of health professional in operation room. Just controlling exposure time is just way to reduce the exposed does of workers. However, in the case, non-probability influence could be occurred. Therefore, this study proved that the exposed dose will be reduced if the factors such as using small detector size of C-arm, setting OFD from 20 cm to 25 cm and non-collimating. Moreover, dose management of C-arm in the non-interesting area will be considered additionally.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1992년도 하계학술대회 논문집 B
• /
• pp.1250-1254
• /
• 1992

This paper describes the designed and fabricated thermal imaging system with the SPRITE(Signal PRocessing in The Element) detector, operating in the 3-12 micron band. This system consists of an afocal telescope, a scan unit containing the SPRITE detector, an electronic processor unit and a cooler. The optical scan system utilizing rotating polygon and oscillating mirror, is 2-dimensional serial/parallel scan type using five elements of the detector. And the electronic processor unit performs digital scan conversion to reform the parallel data stream into serial analog data compatable with conventional RS-170 video. The scan field of view is 40 ${\times}$ 26.7 and the MRTD(Minium Resolvable Temperature Difference) is 0.6 K at 7.5 cycles/mm. The acquired thermal image indicates that this system has a satisfactory performance.