• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• fall
• /
• pp.260-263
• /
• 2021

코로나19로 인해 출입 관련 시스템도 비대면으로 변화하고 있다. 변화에 맞추어 비대면으로 출입자를 관리할 수 있는 프로그램을 개발하여 접촉 위험을 줄이고 출입자 모니터링에 실용성을 제공하고자 한다. 본 연구에서는 Raspberry Pi 카메라에 Alchera Face Authentication API를 적용하여 얼굴인식을 실시하며 정보를 AWS 클라우드에서 저장·관리 하는 시스템을 개발하였다. 챗봇 서비스를 통해 출입자를 확인할 수 있으며 메신저에서 쉽게 클라우드에 접근하여 정보를 확인할 수 있게 하였다. 이를 통해, 특정 장소를 비대면으로 관리하며 간편하게 출입자를 모니터링할 수 있을 것으로 기대한다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.433-434
• /
• 2023

본 논문에서는 미디어파이프와 아이트래킹의 손동작 및 눈 위치 인식을 이용하여 차량 내 조작할 수 있는 다양한 기능을 감압식 버튼이 아닌 카메라를 이용한 동작 기능을 제공해주는 차량 인포테인먼트시스템을 제안한다. 인공지능 모델은 Open-CV 구조를 활용하여 학습을 진행하였고, 라즈베리파이를 이용하여 구현하였다. 제안된 시스템은 운전자를 위해 설계된 다양한 동작들을 시각 정보로 전달해 운전 중 불편함을 대체할 수 있을 뿐만 아니라, 설치 및 사용방법이 간편하여 활용도가 높을 것으로 기대된다.

• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.982-983
• /
• 2023

본 논문은 시각장애인에게 큰 위험이 되는 주거 공간 내 위험 요소들을 감시하고, 경고를 알리는 "시각장애인 주거 공간 내 스마트 안전 시스템"을 제안한다. 주요 특징은 다음과 같다. 첫째, 정해진 시간 또는 음성 명령이 입력되면 이동로봇이 각 방을 자율주행으로 순회하며 위험 요소를 측정한다. 레이저 거리 센서를 활용한 단차 측정값, 열화상 카메라를 활용한 물웅덩이 위험도, 웹캠을 활용한 물체 변위 값을 측정하여 서버에 전송한다. 둘째, 각 방에 설치된 인체 감지 센서를 통해 사용자의 접근을 인식하고, 방에 들어가기 전 서버에 저장된 측정 데이터를 기반으로 스마트 스피커를 통해 위험 요소를 알린다. 셋째, 앱을 통해 관리자가 즉시 검사, 기록조회, 방 정보 수정 등을 이용하여 시스템을 관리한다.

• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.848-849
• /
• 2023

본 논문은 로봇의 실내 환경에서의 자율성을 높이기 위한 SLAM과 내비게이션 방법을 제시한다. 2D LiDAR와 카메라를 이용하여 위치를 인식하고 사람과 장애물을 의미론적으로 검출하며, ICP와 RTAB-map, YOLOv3를 통합하여 Semantic Map을 생성하고 실내 환경에서 자율성을 유지한다. 이 연구를 통해 로봇이 복잡한 환경에서도 높은 수준의 자율성을 유지할 수 있는지 확인하고자 한다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.2A
• /
• pp.95-104
• /
• 2012

Pattern recognition using ensemble classifiers is composed of strong classifier which consists of many weak classifiers. In this paper, we used feature extraction to organize strong classifier using static camera sequence. The strong classifier is made of weak classifiers which considers environmental factors. So the strong classifier overcomes environmental effect. Proposed method uses binary foreground image by frame difference method and the boosting is used to train crowdedness model and recognize crowdedness using features. Combination of weak classifiers makes strong ensemble classifier. The classifier could make use of potential features from the environment such as shadow and reflection. We tested the proposed system with road sequence and subway platform sequence which are included in "AVSS 2007" sequence. The result shows good accuracy and efficiency on complex environment.

• The Journal of the Korea Contents Association
• /
• v.6 no.1
• /
• pp.31-40
• /
• 2006

In this paper, we propose a gait recognition method based on gait silhouette sequences. Features of gait are affected by the variation of gait direction. Therefore, we synthesize silhouettes to canonical form by using planar homography in order to reduce the effect of the variation of gait direction. The planar homography is estimated with only the information which exist within the gait sequences without complicate operations such as camera calibration. Even though gait silhouettes are generated from an individual person, fragments beyond common characteristics exist because of errors caused by inaccuracy of background subtraction algorithm. In this paper, we use the Principal Component Analysis to analyze the deviated characteristics of each individual person. PCA used in this paper, however, is not same as the traditional strategy used in pattern classification. We use PCA as a criterion to analyze the amount of deviation from common characteristic. Experimental results show that the proposed method is robust to the variation of gait direction and improves separability of test-data groups.

• Journal of Korea Game Society
• /
• v.12 no.4
• /
• pp.45-51
• /
• 2012

Current virtual game interfaces cannot comprehend our metaphor, cannot reflect on our natural behavior aspect, cannot make us immerse into a game, and makes a barrier between virtual game space and our real behavior. It is very meaningful issue to use real objects tightly related to human-being's behaviors or reactions for interacting with game applications. Interactive Augmented Reality interfaces may augment users' perception of the real world by adding virtual information to it. We attempted an experiment on camera-based non-marker interface for online TCG application. This experiment uses real TCG cards which are recognized by our two phases Image KeyPoint Extraction/Matching Algorithm. These initiative experiments not only enlarge immersion and reality to the game, but also make real and virtual world seamless.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.3
• /
• pp.1-10
• /
• 2010

As various interfacing devices for computational machines are being developed, a new HCI method using hand motion input is introduced. This interface method is a vision-based approach using a single camera for detecting and tracking hand movements. In the previous researches, only a skin color is used for detecting and tracking hand location. However, in our design, skin color and shape information are collectively considered. Consequently, detection ability of a hand increased. we proposed primary orientation edge descriptor for getting an edge information. This method uses only one hand model. Therefore, we do not need training processing time. This system consists of a detecting part and a tracking part for efficient processing. In tracking part, the system is quite robust on the orientation of the hand. The system is applied to recognize a hand written number in script style using DNAC algorithm. Performance of the proposed algorithm reaches 82% recognition ratio in detecting hand region and 90% in recognizing a written number in script style.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.8B
• /
• pp.1549-1557
• /
• 1999

In this paper, we propose a face recognition algorithm based on the differential image method-DCT This algorithm uses neural networks which is flexible for noise. Using the same condition (same luminous intensity and same distance from the fixed CCD camera to human face), we have captured two images. One doesn't contain human face. The other contains human face. Differential image method is used to separate the second image into face region and background region. After that, we have extracted square area from the face region, which is based on the edge distribution. This square region is used as the characteristics region of human face. It contains the eye bows, the eyes, the nose, and the mouth. After executing DCT for this square region, we have extracted the feature vectors. The feature vectors were normalized and used as the input vectors of the neural network. Simulation results show 100% recognition rate when face images were learned and 92.25% recognition rate when face images weren't learned for 30 persons.

• The KIPS Transactions:PartB
• /
• v.12B no.4 s.100
• /
• pp.387-394
• /
• 2005

This paper describes an idea for determining self-localization using visual landmark. The critical geometric dimensions of a pentagon are used here to locate the relative position of the mobile robot with respect to the pattern. This method has the advantages of simplicity and flexibility. This pentagon is also provided nth a unique identification, using invariant features and colors that enable the system to find the absolute location of the patterns. This algorithm determines both the correspondence between observed landmarks and a stored sequence, computes the absolute location of the observer using those correspondences, and calculates relative position from a pentagon using its (ive vortices. The algorithm has been implemented and tested. In several trials it computes location accurate to within 5 centimeters in less than 0.3 second.