• Smart Media Journal
• /
• v.13 no.2
• /
• pp.68-84
• /
• 2024

This paper describes the development of a system and algorithms for high-quality welfare services by recognizing behavior development indicators (activity, sociability, danger) in children aged 0 to 2 years old using action recognition technology. Action recognition targeted 11 behaviors from lying down in 0-year-olds to jumping in 2-year-olds, using data directly obtained from actual videos provided for research purposes by three nurseries in the Gwangju and Jeonnam regions. A dataset of 1,867 actions from 425 clip videos was built for these 11 behaviors, achieving an average recognition accuracy of 97.4%. Additionally, for real-world application, the Edge Video Analyzer (EVA), a behavior analysis device, was developed and implemented with a region-specific random frame selection-based PoseC3D algorithm, capable of recognizing actions in real-time for up to 30 people in four-channel videos. The developed system was installed in three nurseries, tested by ten childcare teachers over a month, and evaluated through surveys, resulting in a perceived accuracy of 91 points and a service satisfaction score of 94 points.

• International conference on construction engineering and project management
• /
• 2024.07a
• /
• pp.1282-1282
• /
• 2024

Given the widespread use of intelligent surveillance cameras at construction sites, recent studies have introduced vision-based deep learning approaches. These studies have focused on enhancing the performance of vision-based excavator activity recognition to automatically monitor productivity metrics such as activity time and work cycle. However, acquiring a large amount of training data, i.e., videos captured from actual construction sites, is necessary for developing a vision-based excavator activity recognition model. Yet, complexities of dynamic working environments and security concerns at construction sites pose limitations on obtaining such videos from various surveillance camera locations. Consequently, this leads to performance degradation in excavator activity recognition models, reducing the accuracy and efficiency of heavy equipment productivity analysis. To address these limitations, this study aimed to conduct sensitivity analysis of excavator activity recognition performance based on surveillance camera location, utilizing synthetic videos generated from a game-engine-based virtual environment (Unreal Engine). Various scenarios for surveillance camera placement were devised, considering horizontal distance (20m, 30m, and 50m), vertical height (3m, 6m, and 10m), and horizontal angle (0° for front view, 90° for side view, and 180° for backside view). Performance analysis employed a 3D ResNet-18 model with transfer learning, yielding approximately 90.6% accuracy. Main findings revealed that horizontal distance significantly impacted model performance. Overall accuracy decreased with increasing distance (76.8% for 20m, 60.6% for 30m, and 35.3% for 50m). Particularly, videos with a 20m horizontal distance (close distance) exhibited accuracy above 80% in most scenarios. Moreover, accuracy trends in scenarios varied with vertical height and horizontal angle. At 0° (front view), accuracy mostly decreased with increasing height, while accuracy increased at 90° (side view) with increasing height. In addition, limited feature extraction for excavator activity recognition was found at 180° (backside view) due to occlusion of the excavator's bucket and arm. Based on these results, future studies should focus on enhancing the performance of vision-based recognition models by determining optimal surveillance camera locations at construction sites, utilizing deep learning algorithms for video super resolution, and establishing large training datasets using synthetic videos generated from game-engine-based virtual environments.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.1143-1148
• /
• 2006

사용자 친화형 유저 인터페이스 구현을 위해 인간의 손 형상을 실시간으로 인식하는 연구의 중요성이 부각되고 있다. 그러나 인간의 손은 자유도가 크기 때문에 손 형상을 정확히 인식하기란 매우 어렵고 또한 피부색과 유사한 색을 가지는 복잡한 배경에서는 더욱 곤란하다. 본 논문에서는 별도의 센서를 부착하지 않고 카메라를 사용하여 피부색 정보에 의한 손 형상을 분할한 후 손가락 끝 점을 찾는다. 찾은 손가락 끝점을 이용하여 방향을 탐지하는 알고리즘에 대해 기술한다. 이 방법은 템플리트 매칭을 이용하여 손가락 끝 점을 탐색한 후 찾은 손 가락 끝 점과 손목의 중심을 이용하여 전, 후, 좌, 우 방향을 탐지한다. 제안하는 방법을 이용하여 3D가상현실 공간에서의 Navigation에 응용하였으며, 실험결과 전진, 후진 및 좌측, 우측의 방향전환도 매우 좋은 결과를 보였다. 또한 본 논문에서 제안하는 방법은 마우스, 키보드, 조이스틱 등의 조작 없이 전, 후, 좌, 우 방향전환을 사용자가 직관적으로 지시함으로써 보다 자연스러운 인간과 컴퓨터의 상호작용을 제공할 수 있을 것이다.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.2
• /
• pp.178-183
• /
• 2008

Human motion analysis is researched as a new method for human-robot interaction (HRI) because it concerns with the key techniques of HRI such as motion tracking and pose recognition. To analysis human motion, extracting features of human body from sequential images plays an important role. After finding the silhouette of human body from the sequential images obtained by CCD color camera, the skeleton model is frequently used in order to represent the human motion. In this paper, using the silhouette of human body, we propose the feature extraction method based on hybrid skeleton for detecting human motion. Finally, we show the effectiveness and feasibility of the proposed method through some experiments.

• Korean Journal of Computational Design and Engineering
• /
• v.14 no.1
• /
• pp.42-49
• /
• 2009

Virtual manufacturing technology has been applied in actual production sites with the development of virtual reality technology. However, the current virtual manufacturing technology requires experts for application of the system. Furthermore, the sense of reality is diminished as the entire simulation is driven by virtual objects. In contrast, mixed reality can visualize virtual objects and an actual work place simultaneously, and thus the sense of reality of the virtual manufacturing simulation can be improved. This paper introduces a method that applies mixed reality in the manufacturing process, and proposes a method to adapt general safety sign post in the factory instead of a black square marker for visual fiducial recognition.

• Journal of Drive and Control
• /
• v.17 no.4
• /
• pp.72-79
• /
• 2020

To ensure the safety of Advanced Driver Assistance Systems (ADAS) or autonomous vehicles, it is important to recognize the vehicle position, and specifically, the increased accuracy of the lateral position of the vehicle is required. In recent years, the quality of GPS signals has improved a lot and the price has decreased significantly, but extreme urban environments such as tunnels still pose a critical challenge. In this study, we proposed stable and precise lane recognition and tracking methods to solve these two issues via fusion of AVM images and vehicle sensor data using an extended Kalman filter. In addition, the vehicle's lateral position recognition and the abnormal state of RTK GPS were determined using this approach. The proposed method was validated via actual vehicle experiments in urban areas reporting multipath and signal disconnections.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.01a
• /
• pp.271-272
• /
• 2023

본 논문에서는 게임 개발 플랫폼인 언리얼 엔진 4를 사용하여 동작 인식 기반 복싱 게임 콘텐츠를 개발하였다. Google 사의 미디어파이프(MediaPipe) 오픈소스를 통해 웹캠으로 플레이어의 동작을 인식하며, 미디어파이프의 Pose Landmarks를 기준으로 게임 내의 캐릭터와 매핑하여 캐릭터 동작을 제어하여 복싱 게임에 대한 동작, 자세, 반응속도 등을 연습할 수 있는 콘텐츠를 즐길 수 있다. 제안한 동작 인식 기반 게임은 MediaPipe 기술을 이용하여 자신의 동작으로 게임 캐릭터를 제어하여 더 강한 몰입감을 느낄 수 있고, 비싼 VR 기기들 없이 웹캠만 있으면 어디서든 즐길 수 있고, 다양한 콘텐츠를 싼 가격에 즐길 수 있다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.01a
• /
• pp.481-483
• /
• 2023

본 논문에서는 어린이통학버스 안전사고 예방을 위한 지능형 탑승객 모니터링 시스템을 개발한다. 지능형 탑승객 모니터링은 통학버스 내 설치된 카메라로 부터 촬영되는 영상을 실시간으로 분석한 후 통학버스 내 발생할 수 있는 다양한 이벤트를 운전자 또는 교사에게 적시에 통보하여 잠재적 안전사고를 지능적으로 회피할 수 있도록 지원하는 시스템을 말한다. 제안한 시스템은 Yolov4, DeepSort, MediaPipe등의 인공지능 관련 SW기술을 활용하여 영상을 분석한 후 싸움과 같은 이상행동, 정차 후 잔류 인원 발생, 하차자와 차량 간의 안전거리 확보 여부를 포함하는 3가지 이벤트를 인식한 후 운전자 또는 교사에게 알림을 제공한다.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.12
• /
• pp.123-129
• /
• 2014

In accordance with the development of various convergence devices, cameras are being used in many types of the systems such as security system, driver assistance device and so on, and a lot of people are exposed to these system. Therefore the system should be able to recognize the human behavior and support some useful functions with the information that is obtained from detected human behavior. In this paper we use a machine learning approach based on 2D image and propose the human behavior pattern recognition methods. The proposed methods can provide valuable information to support some useful function to user based on the recognized human behavior. First proposed one is "phone call behavior" recognition. If a camera of the black box, which is focused on driver in a car, recognize phone call pose, it can give a warning to driver for safe driving. The second one is "looking ahead" recognition for driving safety where we propose the decision rule and method to decide whether the driver is looking ahead or not. This paper also shows usefulness of proposed recognition methods with some experiment results in real time.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.1
• /
• pp.219-223
• /
• 2020

To guarantee AI model's prominent recognition rate and recognition precision, obtaining the large number of data is essential. In this paper, we propose transfer learning-based object detection algorithm for maintaining outstanding performance even when the volume of training data is small. Also, we proposed a tranfer learning network combining Resnet-50 and YOLO(You Only Look Once) network. The transfer learning network uses the Leeds Sports Pose dataset to train the network that detects the person who occupies the largest part of each images. Simulation results yield to detection rate as 84% and detection precision as 97%.