• 한국인터넷방송통신학회논문지
• /
• 제13권2호
• /
• pp.171-179
• /
• 2013

영상처리에서 인식에 관련된 기술들은 환경에 아주 많은 영향을 받게 되는데 이러한 인식률을 결정짓는 요소 중인 파라미터는 환경에 적절한 값을 얼마나 잘 선택하느냐에 따라서 인식률의 큰 차이를 보인다. 본 논문은 눈동자 추적 알고리즘이 사람이나 실험 환경의 변화에 따라 인식률이 저하되는 현상을 보완하기 위한 성능 향상 및 환경에 적응하는 시스템의 구현에 대한 방법이다. 최적의 파라미터를 얻기 위해 전 처리에 사용되는 이진화 알고리즘의 문턱값을 학습이 필요한 시기를 적절히 판단해 강화학습을 이용하여 다시 학습시켜 인식률을 향상시키는 방법을 사용했다. 실험데이터를 수집하기 위해 입력 장치는 가격이 저렴하고 일반적인 웹 카메라를 사용 하였으며 얼굴 영역에 해당하는 많은 양의 이미지를 수집하여 강화학습의 적응력을 실험하였다. 이미지의 그룹을 다양하게 변화시켜 실험한 결과 강화학습을 사용한 경우 그렇지 않은 경우에 비해 작게는 3% 많게는 14%가량의 성능이 향상됨을 확인하였다. 이렇게 성능이 향상된 눈동자 추적 시스템은 휴먼 컴퓨터 인터랙션 분야에 효과적으로 활용될 수 있을 것이다.

• 디지털콘텐츠학회 논문지
• /
• 제7권1호
• /
• pp.67-74
• /
• 2006

기존의 화상 회의 시스템에서는 카메라가 고정되어 있어서 사용자의 움직임에 제약을 주어 사용자를 부자연스럽게 한다. 이러한 부자연스러움을 해결하기 위해서는 얼굴의 움직임을 추적해야 하는데, 이때 얼굴 전체를 정보로 추적하는 것은 얼굴 전체를 하나의 특징으로 규정짓기도 힘들고 연산 시간이 많이 걸린다는 문제점을 가지고 있다. 따라서, 얼굴의 움직임을 효율적으로 추적하기 위해서는 얼굴상의 몇 개의 특징점을 이용하는 것이 바람직하다. 본 논문은 화상 회의에서 자연스러운 사용자 인터페이스를 위한 자동 얼굴 추적 시스템의 필수적인 요소인 눈 위치 검출의 효과적인 방법에 대하여 논한다. 눈은 얼굴 내에서 가장 뚜렷하며 단순한 특징을 가지고 있으므로 얼굴을 추적하기 위한 가장 중요한 정보가 된다. 본 논문에서 제안한 알고리즘은 얼굴 후보 영역 추출 단계를 거친 얼굴 후보 영역들에 대해 적용되며, 기존 방법들에 비해 조명에 특별한 제약을 받지 않으며 얼굴 크기와 안경에 대한 제약도 가지고 있지 않다. 또한, 화상 회의 환경에 대한 on-line 실험에서 좋은 결과를 나타냈다.

• 한국실내디자인학회논문집
• /
• 제23권1호
• /
• pp.122-131
• /
• 2014

This study aimed to conduct an eye-tracking experiment carried with facade images as a medium and to do research on 'the interest of the eyes' resulted from people's gazing phenomena. This study secured gazing data which appeared according to visual response and analyzed gazing phenomena to find the basic theory of 'the interest of the eyes' as a methodological role, which consumer interest and attention could be grafted when a plan and a design for space design was made. Data terms used in eye-tracking backgrounds and the movement of the eyes were investigated in literature review. Twenty (20) facade images were selected through a case study to get experimental stimuli for the related experiment. Thirty (30) subjects (men and women) suitable for the experiment were recruited to conduct an eye-tracking experiment. After the experiment, five (5) areas were set up in the facade image to identify the focus level of interest and attention. The level of interest and focus was connected to the interest of the eyes. The analysis to study the interest of the eyes was based on nine (9) items such as sequence, entry time, dwell time, hit ratio, revisits, revisitors, average fixation, first fixation and fixation count. Through gaze analysis, the following conclusion was drawn about the 'interest level of sight' for gaze frequency. The interest level can be interpreted to be higher for faster sequence, shorter entry time, longer all fixation(ms) for dwell time, faster all saccade(%), higher hit ratio, more revisits, more revisitors, longer average fixation, faster and longer first fixation, and more fixation count, and the person can be said to have felt interest faster and/or more.

• 디지털콘텐츠학회 논문지
• /
• 제13권2호
• /
• pp.159-167
• /
• 2012

눈동자 추적 기술은 사용자의 의도를 이해하기 위해 눈 움직임을 탐지한다. 이 기술은 최근 점차적으로 발전하였으며 현재에는 여러 경우에 사용될 수 있다. 예를 들어, 장애인이 눈으로 컴퓨터를 사용하게 할 수 있다. 이 논문은 장애인을 위한 눈 추적 시스템 설계 원리, 구체적 적용 사항, 일반적으로 어떻게 적용되고 있는지에 대해 설명한다. 그리고 이를 구현하기 위한 눈동자 자동 추적 알고리즘을 제시한다. 자동 적용 알고리즘은 자동 초점으로 명확한 눈의 이미지를 잡기 위해 렌즈의 움직임을 조정하는 신호 피드백에 기반을 둔다. 이 CCD 카메라 자동 초점 기법은 외부 환경 빛의 강도 변화에 자동 적응하는 기능을 가진다. 이것은 수동 조절의 문제점을 피하고 조절의 정확성에 향상을 가져온다.

• 한국멀티미디어학회논문지
• /
• 제25권2호
• /
• pp.374-381
• /
• 2022

This study analyzes the visual perceptual difference of observers in the artworks created by human artists and artificial intelligence(AI) through eye-tracking. More specifically, the study analyzes the degree of visual attention through a fixation experiment on non-linguistic sources such as the formation and expression of artworks. As a result of this study, the subjects had guessed that one out of four artworks were created by AI (in actuality, 61.1% of the artworks were created by The Next Rembrandt). This demonstrates that most of the subjects hardly recognized the difference between the artwork of human artists and AI. From the comparative analysis of visual perceptual differences found through eye-tracking, more visual attention was found to be demanded for catching details of more stimulating visuals compared to less stimulating visuals. In the gender difference analysis, both of the female and male subjects were likely to stare more intently at the flowers of still-life paintings (Deep Dream & Vincent Van Gogh) while the eyes of a portrait painting (Rembrandt & The Next Rembrandt); this demonstrates no significant differences in gender. Various opinions on AI and art creation from different perspectives arose, therefore, this research is meaningful in a way that it suggests an objective examination through experiments with an artistic perspective.

• Transactions on Control, Automation and Systems Engineering
• /
• 제3권4호
• /
• pp.229-235
• /
• 2001

Recently, service area has been emerging field f robotic applications. Even though assistant robots play an important role for the disabled and the elderly, they still suffer from operating the robots using conventional interface devices such as joysticks or keyboards. In this paper we propose an efficient computer interface using real-time eye-gaze tracking system. The inputs to the proposed system are images taken by a camera and data from a magnetic sensor. The measured data is sufficient to describe the eye and head movement because the camera and the receiver of a magnetic sensor are stationary with respect to the head. So the proposed system can obtain the eye-gaze direction in spite of head movement as long as the distance between the system and the transmitter of a magnetic position sensor is within 2m. Experimental results show the validity of the proposed system in practical aspect and also verify the feasibility of the system as a new computer interface for the disabled.

• Journal of Multimedia Information System
• /
• 제9권2호
• /
• pp.137-144
• /
• 2022

This research aims to identify the relationship between visual perception and emotion by the types of fear responses elicited from warning images on cigarette packages as well as the effectiveness of the size of such images through questionnaires and eye-tracking experiments with twenty university students from the colleges based in Busan. The research distinguished and analyzed the warning images as rational appeals and emotional appeals by the degree of fear and disgust and the result concurred with the research conclusions of Maynard that people would naturally avoid eye contact when presented with a warning image on cigarette packages. Also, eye avoidance was highly identified with larger (75%) warning images. While the previous research mostly adopted the self-rated validation method, this research tried to make the methodology more objective by adopting both questionnaires and eye-tracking experiments. Through this research, authors contribute to finding effective warning images on cigarette packages in a way to increase public awareness of the dangers of smoking and discourage smoking. Further research is recommended to explore the effectiveness of using explicit images on cigarette packages by the types of smokers such as heavy smokers, normal smokers, and non-smokers.

• 한국실내디자인학회논문집
• /
• 제25권1호
• /
• pp.73-80
• /
• 2016

Street is the most fundamental component of city and place to promote diverse actions of people. Pedestrians gaze at various street environments. A visual gaze means that there are interesting elements and these elements need to be preferentially improved in the street environment improvement project. Therefore, this study aims to set up the priorities of street environment components by analyzing eye movements from a pedestrian perspective. In this study, street environment components were classified into road, street facility, building(facade) and sky and as street environment images, three "Streets of Youth" situated in Gwangbok-ro, Seomyeon and Busan University of Busan were selected. The experiment targeted 30 males and females in their twenties to forties. After setting the angle of sight through a calibration test, an eye-tracking experiment regarding the three images was conducted. Lastly, the subjects were asked to fill in questionnaires. The following three conclusions were obtained from the results of the eye-tracking experiment and the survey. First, building was the top priority of street environment components and it was followed by street facility, road and sky. Second, as components to be regarded as important, fast 'Sequence', many 'Fixation Counts' and 'Visit Counts', short 'Time to First Fixation' and long 'Fixation Duration' and 'Visit Duration' were preferred. Third, after voluntary eye movements, the subjects recognized the objects with the highest gaze frequency and the lowest gaze frequency.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2006년도 춘계종합학술대회
• /
• pp.841-846
• /
• 2006

A real time system realization for binocular eyeball tracking cursor on the computer monitor screen is presented in the paper. The processing for searching iris and tracking the cursor are that a facial is acquired by the small CCD camera, convert it into binary image, search for the eye two using the five region mask method in the eye surroundings and the side four points diagonal positioning method is searched the each iris. The tracking cursor is matched by measuring the iris central moving position, The cursor controlling is achieved by comparing two related distances between the iris maximum moving and the cursor moving to calculate the moving stance from gazing position and screen. The experimental result are obtained by examining some adults person on the system.

• Journal of information and communication convergence engineering
• /
• 제7권2호
• /
• pp.131-134
• /
• 2009

The red-eye effect in photography occurs when using a photographic flash very close to the camera lens, in ambient low light due to in experience. Once occurred, the photographer needs to remove it with image tool that requires time consuming, skillful process. In this paper, we propose a new method to extract and remove such red-eye area automatically. Our method starts with transforming ROB space to YCbCr and HSI space and it extracts the face area by using skin color information. The target red-eye area is then extracted by applying 8-direction contour tracking algorithm and removed. The experiment shows our method's effectiveness.