• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.885-888
• /
• 2019

다양한 요소에 의해 영향을 받는 주식 시장에서 정확한 분석과 예측은 막대한 수익과 최소 손실을 보장한다. 본 논문은 주가 움직임 예측을 위하여 다양한 기술적 지표로부터 적합한 특징을 선택하고 세 가지 분류 알고리즘 LSTM, SVM, MLP 을 통해 향후 1, 3, 5, 7, 10, 15, 20, 25, 30 일 후의 주가 움직임을 예측하는 실험을 진행하였다. LSTM 에서 30 일 후를 예측할 때 74.4%의 가장 높은 분류 정확도를 보였으며 전반적으로 LSTM 을 통한 분류가 우수한 결과를 나타냈다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.3
• /
• pp.11-22
• /
• 2010

This paper is a study on implementation of intelligent image surveillance system using context information and supplements temporal-spatial constraint, the weak point in which it is hard to process it in real time. In this paper, we propose scene analysis algorithm which can be processed in real time in various environments at low resolution video(320*240) comprised of 30 frames per second. The proposed algorithm gets rid of background and meaningless frame among continuous frames. And, this paper uses wavelet transform and edge histogram to detect shot boundary. Next, representative key-frame in shot boundary is selected by key-frame selection parameter and edge histogram, mathematical morphology are used to detect only motion region. We define each four basic contexts in accordance with angles of feature points by applying vertical and horizontal ratio for the motion region of detected object. These are standing, laying, seating and walking. Finally, we carry out scene analysis by defining simple context model composed with general context and emergency context through estimating each context's connection status and configure a system in order to check real time processing possibility. The proposed system shows the performance of 92.5% in terms of recognition rate for a video of low resolution and processing speed is 0.74 second in average per frame, so that we can check real time processing is possible.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 1992.06a
• /
• pp.495-497
• /
• 1992

• The Korean Journal of Nuclear Medicine Technology
• /
• v.21 no.2
• /
• pp.80-85
• /
• 2017

Purpose The D530c have cadmium zinc telluride(CZT) detectors that are arranged focus on the heart. This structural characteristic allows for quicker imaging without rotation, but this is sensitive to patient movement and can affect the test results. The aim of this study is to optimize the image quality by reducing patient movement during the examination. Materials and Methods We analyzed the patients' movements, and performed various activities such as provided patient education about correct breathing techniques and avoiding patient movements, and created breathing correction tools to minimize patient movement during exam. The 70 patients who underwent myocardial perfusion SPECT with D530c in November 2016 were categorized as the group before the corrective steps. Another 70 patients who underwent the procedure with D530c from February 14, 2017 to February 21, 2017 were categorized as the improvement group. Images acquired during stress and at rest were compared and analyzed by measuring the durations of heart movements over certain distances (4 mm, 8 mm, 12 mm, or more) noted on the x-, y-, and zaxes. Results After the activities, the durations of heart movements decreased in the images acquired both under stress and at rest. In particular, there were no large motions greater than 12 mm recorded in the stress images after the improvement. There was a significant difference (p<0.005) in the 4-mm and 8-mm fluctuations on the X-axis and the 8-mm fluctuations on the Z axis in the stress images, but there was no significant difference (p>0.005) in the other stress and rest intervals. Conclusion The decrease in the time of motion occurrence due to the 4 mm fluctuation distance that can occur through breathing can be understood as a result of the breathing being corrected through training and motion prevention tools. It is expected that the image quality will be improved by reducing the occurrence time according to the variation distance of 8 mm or 12 mm, which is expected as the actual movement of the patient other than the breathing.

• Cartoon and Animation Studies
• /
• s.40
• /
• pp.33-53
• /
• 2015

The movement of characters is one of the crucial elements to deliver their emotion flowing inside. Though it is the same movement, it may appear or be expressed differently according to the character's personality or emotion or the particular situation. The purpose of this study is to analyze not only the movement found superficially in an animation but also a character's internal emotion and attitude with Laban's movement analysis system, particularly effort, one of its analysis categories, and examine how effectively Laban's movement analysis often employed at the circles of dance can analyze movement in an animation. is about a monster that constantly makes efforts to realize its dream to be a scarer. Functional movement forms the most part, but expressive movement to show how a character thinks or feels also appears harmoniously. Characters' externally shown movement can express their internal emotion properly sometimes, but they also often move expressing their feelings in moderation. Therefore, this study analyzes the movement of characters found in the four scenes of with LMA's effort. According to the findings, at the scene where Michael enters the door leading to the human world following the scarer, the emotional state of Michael envious of the scarer is expressed with the Vision Drive giving the strong feel of dreaming. At the scene of the second game to choose the best scare team, it shows us the Spell Drive with its careful and light movement having clear intention to survive at the game. At the scene where there is a party held for the teams that have survived, it shows the Passion Drive of being eagerly expressing happy and delightful feelings without considering what is around. At the scene where Michael and Sullivan are pursued by people, the Action Drive was used to express movement that was heavy and strong and was getting faster gradually by focusing the feelings of the characters in haste into one place.

• Journal of Korea Multimedia Society
• /
• v.10 no.8
• /
• pp.967-980
• /
• 2007

Video objects of interest play an important role in representing the video content and are useful for improving the performance of video retrieval and compression. The objects of interest may be a main object in describing contents of a video shot or a core object that a video producer wants to represent in the video shot. We know that any object attracting one's eye much in the video shot may not be an object of interest and a non-moving object may be an object of interest as well as a moving one. However it is not easy to define an object of interest clearly, because procedural description of human interest is difficult. In this paper, a set of four filtering conditions for extracting moving objects of interest is suggested, which is defined by considering variation of location, size, and moving pattern of moving objects in a video shot. Non-moving objects of interest are also defined as another set of four extracting conditions that are related to saliency of color/texture, location, size, and occurrence frequency of static objects in a video shot. On a test with 50 video shots, the segmentation method based on the two sets of conditions could extract the moving and non-moving objects of interest chosen manually on accuracy of 84%.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.4
• /
• pp.811-819
• /
• 2017

A behavior pattern analysis algorithm based on descriptors consists of information of a moving object and temporal histogram is proposed. Background learning is performed first for detecting, tracking and analyzing moving objects. Each object is identified using an association of the center of gravity of objects and tracked individually. A temporal histogram represents a motion pattern using positions of the center of gravity and time stamp of objects. The characteristic and behavior of objects are figured out by comparing each coordinates of a position history in the histogram. Behavior information which is comprised with numbers of a start and end frame, and coordinates of positions of objects is stored and managed in the linked list. Descriptors are made with the stored information and the video retrieval algorithm is designed. We confirmed the higher retrieval accuracy compare with conventional methods.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.39 no.6
• /
• pp.49-58
• /
• 2002

Based on electrophysiological retina mechanism, a retina model is proposed, which has similar response characteristics compared with the real primate retina. Photoreceptors, horizontal cells, and bipolar cells are modeled based on the previously studied retina models. And amacrine cells known to have relation to movements detection, and bipolar cell terminals are newly modeled using 3 NDP mechanism. The proposed model verified by analyzing the spatial response characteristics to stationary and moving stimuli, and characteristics for different speeds. Through this retina model, human vision system could be applied to computer vision systems for movement detection, and it could be the basic research for the implantable artificial retina.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.6
• /
• pp.24-34
• /
• 2007

Application systems that easy to access a information have been developed by IT growth and a human life variation. In this paper, we propose a application system to register a 3D footwear model using a monocular camera. In General, a human motion analysis research to body movement. However, this system research a new method to use a foot movement. This paper present a system process and show experiment results. For projection to 2D foot plane from 3D shoe model data, we construct processes that a foot tracking, a projection expression and pose estimation process. This system divide from a 2D image analysis and a 3D pose estimation. First, for a foot tracking, we propose a method that find fixing point by a foot characteristic, and propose a geometric expression to relate 2D coordinate and 3D coordinate to use a monocular camera without a camera calibration. We make a application system, and measure distance error. Then, we confirmed a registration very well.

• Proceedings of the Korea Society of Information Technology Applications Conference
• /
• 2007.05a
• /
• pp.38-46
• /
• 2007

본 논문에서는 실시간으로 획득된 영상을 분석하여 움직이는 다수 물체를 검출하고, 카메라를 자동 제어하여 관심 보행자만을 추적하는 시스템을 제안한다. 다수 물체 영역 검출은 차영상과 이전변환 밀도값을 이용한다. 검출된 다수 물체 영역에서 사람의 구조적 정보와 형태 정보를 이용하여 나무들의 흔들림으로 인한 영역이나 차량의 움직임 영역은 제거되고, 관심 보행자 영역만을 검출하였다. 관심 보행자 추적은 무게중심 차를 이용한 움직임 정보와 k-means 알고리즘으로 구한 세 점의 평균 색상 정보를 이용한다. 원거리 관심 보행자는 인식률을 높이기 위해 줌을 실행하여 확대하고, 관심 보행자의 화면상 위치에 따라 카메라 방향을 자동으로 조정하여 관심 보행자반을 연속적으로 추적한다. 실험 결과, 제안한 시스템은 실시간으로 움직이는 다수 물체를 검출하고, 사람의 구조적 특정과 형태 정보로 관심 보행자만을 검출할 수 있었고, 움직임 정보와 색상정보로 관심 보행자를 연속적으로 추적할 수 있었다.