• ETRI Journal
• /
• v.33 no.4
• /
• pp.569-579
• /
• 2011

This paper presents a novel approach for human activity recognition (HAR) using the joint angles from a 3D model of a human body. Unlike conventional approaches in which the joint angles are computed from inverse kinematic analysis of the optical marker positions captured with multiple cameras, our approach utilizes the body joint angles estimated directly from time-series activity images acquired with a single stereo camera by co-registering a 3D body model to the stereo information. The estimated joint-angle features are then mapped into codewords to generate discrete symbols for a hidden Markov model (HMM) of each activity. With these symbols, each activity is trained through the HMM, and later, all the trained HMMs are used for activity recognition. The performance of our joint-angle-based HAR has been compared to that of a conventional binary and depth silhouette-based HAR, producing significantly better results in the recognition rate, especially for the activities that are not discernible with the conventional approaches.

• Smart Structures and Systems
• /
• v.25 no.6
• /
• pp.733-749
• /
• 2020

This study presents a computer vision-based approach for representing time evolution of structural damages leveraging a database of inspection images. Spatially incoherent but temporally sorted archival images captured by robotic cameras are exploited to represent the damage evolution over a long period of time. An access to a sequence of time-stamped inspection data recording the damage growth dynamics is premised to this end. Identification of a structural defect in the most recent inspection data set triggers an exhaustive search into the images collected during the previous inspections looking for correspondences based on spatial proximity. This is followed by a view synthesis from multiple candidate images resulting in a single reconstruction for each inspection round. Cracks on concrete surface are used as a case study to demonstrate the feasibility of this approach. Once the chronology is established, the damage severity is quantified at various levels of time scale documenting its progression through time. The proposed scheme enables the prediction of damage severity at a future point in time providing a scope for preemptive measures against imminent structural failure. On the whole, it is believed that the present study will immensely benefit the structural inspectors by introducing the time dimension into the autonomous condition assessment pipeline.

• Journal of Korea Society of Industrial Information Systems
• /
• v.25 no.5
• /
• pp.41-48
• /
• 2020

Recently, as the demand for the mobile platform market in the virtual/augmented/mixed reality field is increasing, experiential content that gives users a real-world felt through a virtual environment is drawing attention. In this paper, as a method of tracking a tracker for user location estimation in a virtual space movement platform for motion capture of trainees, we present a method of estimating 3D coordinates of the 3D cross covariance through the coordinates of the markers projected on the image. In addition, the validity of the proposed algorithm is verified through rigid body tracking experiments.

• Smart Structures and Systems
• /
• v.24 no.5
• /
• pp.587-595
• /
• 2019

Deformation measurement of large scale structures, such as the ground beds of high-rise buildings, tunnels, bridge, and railways, are important for insuring service quality and safety. The pose-relay videometrics method and displacement-relay videometrics method have already presented to measure the pose of non-intervisible objects and vertical subsidence of unstable areas, respectively. Both methods combine the cameras and cooperative markers to form the camera series networks. Based on these two networks, we propose two novel videometrics methods with closed-loop camera series network for deformation measurement of large scale structures. The closed-loop camera series network offers "closed-loop constraints" for the camera series network: the deformation of the reference points observed by different measurement stations is identical. The closed-loop constraints improve the measurement accuracy using camera series network. Furthermore, multiple closed-loops and the flexible combination of camera series network are introduced to facilitate more complex deformation measurement tasks. Simulated results show that the closed-loop constraints can enhance the measurement accuracy of camera series network effectively.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.8
• /
• pp.4026-4040
• /
• 2019

In this paper, we present an approach for detecting slide transitions in lecture videos by introducing the spatio-temporal residual networks. Given a lecture video which records the digital slides, the speaker, and the audience by multiple cameras, our goal is to find keyframes where slide content changes. Since temporal dependency among video frames is important for detecting slide changes, 3D Convolutional Networks has been regarded as an efficient approach to learn the spatio-temporal features in videos. However, 3D ConvNet will cost much training time and need lots of memory. Hence, we utilize ResNet to ease the training of network, which is easy to optimize. Consequently, we present a novel ConvNet architecture based on 3D ConvNet and ResNet for slide transition detection in lecture videos. Experimental results show that the proposed novel ConvNet architecture achieves the better accuracy than other slide progression detection approaches.

• Journal of the Korea Safety Management & Science
• /
• v.25 no.3
• /
• pp.17-22
• /
• 2023

As the importance of artificial intelligence grows rapidly and emerges as a leader in technology, it is becoming an important variable in the next-generation industrial system along with the robot industry. In this study, a safety system was developed using deep learning technology to provide worker safety in a robot workplace environment. The implemented safety system has multiple cameras installed with various viewing directions to avoid blind spots caused by interference. Workers in various scenario situations were detected, and appropriate robot response scenarios were implemented according to the worker's risk level through IO communication. For human detection, the YOLO algorithm, which is widely used in object detection, was used, and a separate robot class was added and learned to compensate for the problem of misrecognizing the robot as a human. The performance of the implemented system was evaluated by operator detection performance by applying various operator scenarios, and it was confirmed that the safety system operated stably.

• Nuclear Engineering and Technology
• /
• v.55 no.2
• /
• pp.493-505
• /
• 2023

Comprehensive condition monitoring of large industry systems such as nuclear power plants (NPPs) is essential for safety and maintenance. In this study, we developed novel system-scale diagnostic technology based on deep-learning and IR thermography that can efficiently and cost-effectively classify system conditions using compact Raspberry Pi and IR sensors. This diagnostic technology can identify the presence of an abnormality or accident in whole system, and when an accident occurs, the type of accident and the location of the abnormality can be identified in real-time. For technology development, the experiment for the thermal image measurement and performance validation of major components at each accident condition of NPPs was conducted using a thermal-hydraulic integral effect test facility with compact infrared sensor modules. These thermal images were used for training of deep-learning model, convolutional neural networks (CNN), which is effective for image processing. As a result, a proposed novel diagnostic was developed that can perform diagnosis of components, whole system and accident classification using thermal images. The optimal model was derived based on the modern CNN model and performed prompt and accurate condition monitoring of component and whole system diagnosis, and accident classification. This diagnostic technology is expected to be applied to comprehensive condition monitoring of nuclear power plants for safety.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.374-375
• /
• 2022

Recently, Metaverse service has been widely used to naturally communicate with a remote location, freeing from time and spatial constraints. In order to produce such contents, it is necessary to restore and synthesize a 3D model based on real space data. In this paper, a 3D-generated reconstruction model is produced based on continuous images using multiple cameras and a technique to correct the reconstructed 3D model is presented. For this. offline multi-camera setup was performed, errors were analyzed on the 3D model created through images obtained from various angles, and correction was performed using a matching technique between image frames. It is expected that 3D reconstructed data can be utilized in various service fields such as culture, tourism, and medical care.

• International conference on construction engineering and project management
• /
• 2013.01a
• /
• pp.21-27
• /
• 2013

Construction safety is a predominant hindrance in in-situ workflow and considered an unresolved issue. Current methods used for safety optimization and prediction, with limited exceptions, are paper-based, thus error prone, as well as time and cost ineffective. In an attempt to exploit the potential of BIM for safety, the objective of the proposed methodology is to automatically predict hazardous on-site conditions related to the route that the dozers follow during the different phases of the project. For that purpose, safety routes used by construction equipment from an origin to multiple destinations are computed using video cameras and their cycle times are calculated. The cycle times and factors; including weather and light conditions, are considered to be independent and identically distributed random variables (iid); and simulated using the Arena software. The simulation clock is set to 100 to observe the minor changes occurring due to external parameters. The validation of this technology explores the capabilities of BIM combined with simulation for enhancing productivity and improving safety conditions a-priori. Preliminary results of 262 measurements indicate that the proposed methodology has the potential to predict with 87% the location of exclusion zones. Also, the cycle time is estimated with an accuracy of 89%.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.191-192
• /
• 2024

고령 인구의 지속적인 증가로 인해 고령자의 안전과 관련된 문제는 주요한 관심사 중 하나로 부상하고 있다. 특히, 고령자들 사이에서 자주 발생하는 낙상 사고는 심각한 건강 문제를 일으킬 수 있으며, 이를 예방하고 대응하는 것은 고령 인구의 삶의 질을 향상하는 데 중요한 역할을 한다. 본 연구는 8대의 카메라로 촬영된 영상과 센서 데이터를 통합한 낙상 감지 기법을 제안한다. 제안한 기법은 MediaPipe를 활용하여 Skeleton Keypoint를 추출하는 이미지 인식 기법과 센서 데이터에서 얻은 특징을 활용하는 센서 기반 기술을 결합하여 낙상 사고의 발생 및 방향을 효과적으로 감지할 수 있다. 이러한 결과를 바탕으로 본 연구는 향후 고령자들의 생활 안전성과 의료 시스템의 효율성을 높이는 데 이바지할 수 있을 것으로 기대한다.