• Journal of Multimedia Information System
• /
• v.7 no.3
• /
• pp.205-214
• /
• 2020

Crime has become one of the major problems in modern society. Even though visual surveillances through closed-circuit television (CCTV) is extensively used for solving crime, the number of crimes has not decreased. This is because there is insufficient workforce for performing 24-hour surveillance. In addition, CCTV surveillance by humans is not efficient for detecting dangerous situations owing to accuracy issues. In this paper, we propose the autonomous detection of dangerous situations in CCTV scenes using a deep learning model with relational inference. The main feature of the proposed method is that it can simultaneously perform object detection and relational inference to determine the danger of the situations captured by CCTV. This enables us to efficiently classify dangerous situations by inferring the relationship between detected objects (i.e., distance and position). Experimental results demonstrate that the proposed method outperforms existing methods in terms of the accuracy of image classification and the false alarm rate even when object detection accuracy is low.

• The Korean Journal of Pain
• /
• v.27 no.1
• /
• pp.72-76
• /
• 2014

A 57-year-old male patient had myeloma. He had severe pain in the left clavicle that did not respond to radiotherapy; therefore, it was treated with radiofrequency thermal ablation (RFTA). Under fluoroscopic guidance, two RF needles at a distance of 1.5 cm from each other were inserted into the mass and conventional radiofrequency ($90^{\circ}C$ and 60 seconds) at two different depths (1 cm apart) was applied. Then, 2 ml of 0.5% ropivacaine along with triamcinolone 40 mg was injected in each needle. The visual analogue pain score (VAS from 0 to 10) was decreased from 8 to 0. In the next 3 months of follow-up, the patient was very satisfied with the procedure and the mass gradually became smaller. There were no complications. This study shows that RFTA could be a useful method for pain management in painful osteolytic myeloma lesions in the clavicle

• Current Optics and Photonics
• /
• v.4 no.2
• /
• pp.121-126
• /
• 2020

With the increasing demand for head-mounted display applications, the image quality provided by a near-eye display device is a key factor in satisfying the consumer. Among various techniques to realize a near-eye display that has a thinner volume than the working distance of a human eye, a lightfield image-generation method based on a pinhole array is attracting much attention, with its simple and thin structure. In this paper, we propose a numerical analysis of the visual parameters and verifications with computational reconstruction.

• The Bulletin of The Korean Astronomical Society
• /
• v.39 no.2
• /
• pp.64.2-64.2
• /
• 2014

Interacting galaxy pairs are important for study of galaxy evolution. We selected 8,542 interacting galaxy pairs out of 593,514 KIAS-VAGC galaxy sample with 0.02 < z < 0.047 and r_mag <17.6. We then classified by their interacting features into 6 types by visual inspection. We focused on two types whose spiral tidal features extend to the center of early type galaxy (ETG) or to the edge of ETG. We compared galactic parameters of these two types with those of entire 8,542 pairs as well as between the two types. Preliminary result shows both types are very close pairs (projected distance ~ 20 kpc). Spiral galaxies in the center type are more massive but less bright than those in edge type. ETGs in the edge type are brighter but not more massive than those in the center type. The center type has a mass ratio 3.4 times greater than the edge type, but the edge type has a higher angular momentum than the center type.

• Journal of Korea Game Society
• /
• v.13 no.3
• /
• pp.131-140
• /
• 2013

The study analyzes how to create stereoscopic image in 3-dimension animation 'Yggdrasil', produced in Autodesk Maya(3D animation software), using Anaglyph method that is the oldest in 3-dimension stereoscopic video production. The anaglyph method help to produce stereoscopic images easily and to give effective 3D effect encoding each eye's image using filters of different colors. In addition, space direction applied visual perception theory expresses necessary distance, depth and speed effectively in 'Yggdrasil'

• Proceedings of the KIEE Conference
• /
• 2005.10b
• /
• pp.61-63
• /
• 2005

Retreiving indoor scene reference image from database using visual information is important issue in Robot Navigation. Scene matching problem in navigation robot is not easy because input image that is taken in navigation process is affinly distorted. We represent probabilistic framework for the feature matching between features in input image and features in database reference images to guarantee robust scene matching efficiency. By reconstructing probabilistic scene matching framework we get a higher precision than the existing feaure-feature matching scheme. To construct probabilistic framework we represent each image as Gaussian Mixture Model using Expectation Maximization algorithm using SIFT(Scale Invariant Feature Transform).

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.50.6-50
• /
• 2002

In this paper, the relationship between a slave robot and the uncertain remote environment is modeled as the impedance to generate the virtual force to feed back to the operator. For the control of a teleoperated mobile robot equipped with camera, the teleoperated mobile robot take pictures of remote environment and sends the visual information back to the operator over the Internet. Because of the limitation of communication bandwidth and narrow view-angles of camera, it is not possible to watch the environment clearly, especially shadow and curved areas. To overcome this problem, the virtual force is generated according to both the distance between the obstacle and robot and the approachin...

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.108.1-108
• /
• 2002

This paper is willing to propose a method of wall tracing, a moving algorithm between two points, when a Cleaning robot between two points moves. We use the information about obstacles and wall side in calculating different weight vector the each infrared sensors in cleaning robot. Therefore the cleaning robot navigates the wall. In the algorithm of wall tracing, the value of error in angle and distance between starting point and ending point should be zero to navigate the wall safely. The propriety of algorithm of the wall tracing is simulated as this method by using Visual C++. The result simulated proved to the simulation.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1997.10a
• /
• pp.273-276
• /
• 1997

Seeing is an intelligent act. Retina is the most important part among the components of eye which is comprised of iris, pupil, lens, optic disk, and so forth. Because retina acts like a photo receiver to detect light from every object, if damaged, animals have a severe problem to live along with themselves and sometimes they lose their sight. In the ophthalmology, doctors use special instrument to see exact retina image of object and operate the surgery by rotating focus control knob. In this study, a basic test is done to achieve the auto focus control instrument. Specially made sledge and attenuator are installed on the optic bed to change the distance and laser power between the first reflective mirror and the second one. Control panel which is compiled by Visual C t t to control stepping motors ,laser power and photodiode are implemented with ADA-board.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2011.01a
• /
• pp.317-320
• /
• 2011

현재 고층 빌딩의 규모가 대형화, 고층화 되어감에 따라 엘리베이터 시스템의 성능 향상 연구가 활발히 추진되며, 엘리베이터 시스템의 성능향상은 모든 이용자들이 현재 층에서 목적 층까지 최대한 신속하게 이동할 수 있도록 서비스향상을 목적으로 한다. 요즘은 하나의 승장에 여러 개의 엘리베이터가 부속되어 상호협조적인 운행으로 이용자들의 대기시간을 극소화 하고 있으나 제한적인 조건에서 융통성 있는 퍼지 시스템을 적용한다면 에너지 절감과 경우에 따라 대기시간을 줄일 수 있다고 생각한다. 본 논문에서는 현재 층수와 층과의 거리를 검출하여 퍼지기법을 이용하여 손실 여부를 판단 후, 기존의 엘리베이터보다 효율적인 방법을 제안한다.