• The Journal of Korean Institute of Information Technology
• /
• v.17 no.5
• /
• pp.13-20
• /
• 2019

Recently, object detection technology based on CNN has been actively studied. Object detection technology is used as an important technology in autonomous vehicles, intelligent image analysis, and so on. In this paper, we propose a rotation change robust object detection system by applying MI-FL (Moment Invariant-Feature Layer) to SSD (Single Shot Multibox Detector) which is one of CNN-based object detectors. First, the features of the input image are extracted based on the VGG network. Then, a total of six feature layers are applied to generate bounding boxes by predicting the location and type of object. We then use the NMS algorithm to get the bounding box that is the most likely object. Once an object bounding box has been determined, the invariant moment feature of the corresponding region is extracted using MI-FL, and stored and learned in advance. In the detection process, it is possible to detect the rotated image more robust than the conventional method by using the previously stored moment invariant feature information. The performance improvement of about 4 ~ 5% was confirmed by comparing SSD with existing SSD and MI-FL.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4439-4455
• /
• 2021

Compared with two-stage object detection algorithms, one-stage algorithms provide a better trade-off between real-time performance and accuracy. However, these methods treat the intermediate features equally, which lacks the flexibility to emphasize meaningful information for classification and location. Besides, they ignore the interaction of contextual information from different scales, which is important for medium and small objects detection. To tackle these problems, we propose an image pyramid network based on dual attention mechanism (DAIPNet), which builds an image pyramid to enrich the spatial information while emphasizing multi-scale informative features based on dual attention mechanisms for one-stage object detection. Our framework utilizes a pre-trained backbone as standard detection network, where the designed image pyramid network (IPN) is used as auxiliary network to provide complementary information. Here, the dual attention mechanism is composed of the adaptive feature fusion module (AFFM) and the progressive attention fusion module (PAFM). AFFM is designed to automatically pay attention to the feature maps with different importance from the backbone and auxiliary network, while PAFM is utilized to adaptively learn the channel attentive information in the context transfer process. Furthermore, in the IPN, we build an image pyramid to extract scale-wise features from downsampled images of different scales, where the features are further fused at different states to enrich scale-wise information and learn more comprehensive feature representations. Experimental results are shown on MS COCO dataset. Our proposed detector with a 300 × 300 input achieves superior performance of 32.6% mAP on the MS COCO test-dev compared with state-of-the-art methods.

• Journal of the Korean Society of Radiology
• /
• v.17 no.6
• /
• pp.897-902
• /
• 2023

The purpose of this paper is to analyze the characteristics of Silicon Photomultiplier (SiPM) for the realization of high-sensitivity radiation detection in portable detectors. Portable X-ray detectors offer the advantage of quickly accessing the patient's location and obtaining real-time images, allowing physicians to perform rapid diagnoses. However, this mobility comes with challenges in achieving accurate radiation detection. In existing detectors, SiPM is used for a simple purpose of detecting X-ray triggers. To verify the feasibility of high-sensitivity X-ray detection through SiPM, seven types of SiPM sensors were compared and selected, and their characteristics were analyzed. The SiPM used in the final test demonstrated the ability to distinguish signals at the ultra-low radiation level of 10 nGy, and it was observed that the slope of the signal rise curve varies with the X-ray tube voltage. Utilizing the characteristics of SiPM, which exhibits changes in signal level and duration with X-ray dose, it appears possible to achieve high-sensitivity measurements for X-ray detection.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2007.04a
• /
• pp.243-248
• /
• 2007

This study proposed a procedure to improve the error defection of attached cadastral maps using digital map data. In addition, this study also provided the direction for the accuracy improvement of attached cadastral maps by comparing analysis methods. - such as centroid, Lee Sallee shape index, and area index. The analysis is performed as follows. First, by using centroid measurement, the center point of cadastral maps and attached cadastral maps are compared. Secondly by using Lee Sallee shape measurement, the location accuracy of range area is investigated. Thirdly, by using area measurement, the range area within allowable error scope is verified. Based on analysis, the discrepancy between cadastral maps and the attacked cadastral maps are detected as follows; 98.2% from Lee Sallee shape index, 41.8% from centroid, 15.4% from area index in the whole error.

• Proceedings of the KIEE Conference
• /
• 1995.07c
• /
• pp.1293-1295
• /
• 1995

This paper describes the detection of the ultrasonic signals reduced by barrier in test transformer. The ultrasonic signals are generated by partial discharges which cause the insulation failure of transformer. The ultrasonic signals are reduced by barrier. But it was shown that the reduced ultrasonic signals can be measured as the location of the ultrasonic signal detectors is selected properly.

• Proceedings of the KSR Conference
• /
• 2011.10a
• /
• pp.2082-2089
• /
• 2011

This study is examination of application on site for performance and stability of real-time broken rail detection equipment development system about connection, welded part occurring data errors of broken rail monitoring system. As a result of analysis about data collected in Seoul Metro No.2 Subway from Mar. 2010 to Jan. 2011, we found it is possible to detect crack location(Thermit welding) within the margin of error of ${\pm}1m$ accurately as the first attenuation was -1.2dB and the second was -1.3dB.

• The Bulletin of The Korean Astronomical Society
• /
• v.40 no.1
• /
• pp.66.2-66.2
• /
• 2015

In today's age when telecommunications using satellite has become part of our daily lives, one has to be employ preventive measures to avert any possible danger, of which solar activity is the major cause. Coronal mass ejections (CMEs) heading towards the Earth can lead to disturbances in the Earth's magnetosphere, if their magnetic field is oriented southward. Monitoring of solar filaments in this case becomes very very crucial, as their eruption is associated with most of the CMEs. Monitoring of solar filaments in this case becomes very very crucial, as their eruption is associated with most of the CMEs. Also, filaments show activation up to a few hours prior to launch of a CME and thus can provide advance warning. In this study, we present an algorithm for the detection of solar filaments seen in the extreme ultraviolet (EUV) from Atmospheric Imaging Assembly (AIA) on board the Solar Dynamics Observatory (SDO). Various morphological operations are employed to identify and extract the filaments. These filaments are then tracked in order to determine their size and location continuously.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.11
• /
• pp.2500-2504
• /
• 2011

Living organisms are composed of cells that can replicate themselves through growth, division packed with tons capacity. On DNA mutations, ie mutations in the offspring's survival and reproduction can be held against you, and packed with tons ambivalence that could benefit. In this study, the DNA double helix is used as a template for replication, we first separated into single strands of the double helix must be opened Combining the double helix portion of the location of errors in the bond provides a way to find and repair.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2012.05a
• /
• pp.202-206
• /
• 2012

In this paper, we proposed on the direction of self-repairing mimicking the cell on the digital system design. Three-dimensional array of cells rather than using the original structure of FPGA, an array of blocks for efficient error detection methods were investigated. With a certain regularity, so the design method in detail by dividing the full array. The digital circuits can be detected fault location easily and quickly.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.660-665
• /
• 2006

본 논문에서는 유비쿼터스 컴퓨팅 구현에 있어서 중요 핵심 요소 기술인 실내에서의 사용자의 위치를 인식하면서 동시에 사용자를 식별하는 방법에 대해 제안하고자 한다. 제안된 사용자 식별 방법은 초음파 센서의 물체까지의 거리를 측정할 수 있는 특성에 기초하여 구성원의 키 높이의 다름을 주요한 식별 특성으로 이용하는 방법이다. 또한 사용자 식별을 수행하는 것과 동시에 식별된 사용자가 센서가 장착된 문을 통과하여 들어왔는지 나갔는지를 감지하는 출입 감지 방법도 제안한다. 기본적인 출입 감지 방법의 아이디어는 2개 초음파 센서의 물체 감지 순서를 이용하여 개발되었다. 제안된 방법의 성능 검증을 위해 1단계의 프로토타입을 만들었으며, 개발된 프로토타입을 이용하여 3명의 가족 구성원으로 이루어진 일반 가정집의 화장실에 장착하여 성능 실험을 실시하였다. 실험 결과 100%의 사용자 식별 성능 및 평균 98%의 출입 감지 성능을 얻었다.