• Journal of IKEEE
• /
• v.26 no.2
• /
• pp.249-256
• /
• 2022

The prosthetic leg is a device that performs walking instead of a amputated lower limb, and require a change in locomotion mode by providing the user's intention to respond to a discontinuous locomotion environment. Research has been conducted to detect the users' intentions through biomechanical features inside the socket that directly contacts the cut site in demand for natural locomotion mode changes without external control equipment. However, there is still a need for a sensor system that is suitable for the internal environment of the main body and socket of the cut site. Accordingly, this paper proposed a film-type sensor system that is suitable for the main body characteristics of the cut site, is not affected by the temperature and humidity conditions inside the socket, and is easy to manufacture in various sizes. The proposed sensor is manufactured base on Velostat film and takes into account the pressure measurement characteristics that vary with size. Through the experiment, the change in the internal pressure of the socket due to the intentional posture performance of the wearer was measured, and the possibility of detecting the intention to change the locomotion mode was confirmed.

• Smart Media Journal
• /
• v.11 no.3
• /
• pp.90-101
• /
• 2022

Buses are a popular means of transportation. As such, thorough preparation is needed for passenger safety management. However, the safety system is insufficient because there are accidents such as a death accident occurred when the bus departed without recognizing the elderly approaching to get on in 2018. There is a safety system that prevents pinching accidents through sensors on the back door stairs, but such a system does not prevent accidents that occur in the process of getting on and off like the above accident. If it is possible to predict the intention of bus passengers to get on and off, it will help to develop a safety system to prevent such accidents. However, studies predicting the intention of passengers to get on and off are insufficient. Therefore, in this paper, we propose a 1×1 CNN3D-based getting on and off intention prediction model using skeleton keypoints of passengers extracted from the camera image attached to the bus through UDP-Pose. The proposed model shows approximately 1~2% higher accuracy than the RNN and LSTM models in predicting passenger's getting on and off intentions.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.5
• /
• pp.187-194
• /
• 2021

A person's behavioral recognition is the recognition of what a person does according to joint movements. To this end, we utilize computer vision tasks that are utilized in image processing. Human behavior recognition is a safety accident response service that combines deep learning and CCTV, and can be applied within the safety management site. Existing studies are relatively lacking in behavioral recognition studies through human joint keypoint extraction by utilizing deep learning. There were also problems that were difficult to manage workers continuously and systematically at safety management sites. In this paper, to address these problems, we propose a method to recognize risk behavior using only joint keypoints and joint motion information. AlphaPose, one of the pose estimation methods, was used to extract joint keypoints in the body part. The extracted joint keypoints were sequentially entered into the Long Short-Term Memory (LSTM) model to be learned with continuous data. After checking the behavioral recognition accuracy, it was confirmed that the accuracy of the "Lying Down" behavioral recognition results was high.

• Journal of the Society of Disaster Information
• /
• v.18 no.3
• /
• pp.433-443
• /
• 2022

Purpose: In this paper, we propose a study of deep learning algorithms that estimate and predict sarcopenia by exploiting the high penetration rate of smart devices. Method: To utilize deep learning techniques, experimental data were collected by using the inertial sensor embedded in the smart device. We implemented a smart device application for data collection. The data are collected by labeling normal and abnormal gait and five states of running, falling and squat posture. Result: The accuracy was analyzed by comparative analysis of LSTM, CNN, and RNN models, and binary classification accuracy of 99.87% and multiple classification accuracy of 92.30% were obtained using the CNN-LSTM fusion algorithm. Conclusion: A study was conducted using a smart sensoring device, focusing on the fact that gait abnormalities occur for people with sarcopenia. It is expected that this study can contribute to strengthening the safety issues caused by sarcopenia.

• Journal of the Korean Society of Industry Convergence
• /
• v.26 no.1
• /
• pp.113-119
• /
• 2023

In this paper, we propose an optimization of a pose estimation deep learning model for recognition of abnormal behavior in unmanned stores using radio frequencies. The radio frequency use millimeter wave in the 30 GHz to 300 GHz band. Due to the short wavelength and strong straightness, it is a frequency with less grayness and less interference due to radio absorption on the object. A millimeter wave radar is used to solve the problem of personal information infringement that may occur in conventional CCTV image-based pose estimation. Deep learning-based pose estimation models generally use convolution neural networks. The convolution neural network is a combination of convolution layers and pooling layers of different types, and there are many cases of convolution filter size, number, and convolution operations, and more cases of combining components. Therefore, it is difficult to find the structure and components of the optimal posture estimation model for input data. Compared with conventional millimeter wave-based posture estimation studies, it is possible to explore the structure and components of the optimal posture estimation model for input data using genetic algorithms, and the performance of optimizing the proposed posture estimation model is excellent. Data are collected for actual unmanned stores, and point cloud data and three-dimensional keypoint information of Kinect Azure are collected using millimeter wave radar for collapse and property damage occurring in unmanned stores. As a result of the experiment, it was confirmed that the error was moored compared to the conventional posture estimation model.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.268-268
• /
• 2023

전 지구적 급격한 기후변화로 인해 수문기상인자들의 비선형적 변동성이 발생함과 동시에 가뭄, 홍수와 같은 수재해의 발생빈도 및 강도가 증가하고 있는 추세이다. 이에 따라, 세계의 유수기관 (NASA, ESA 등)에서는 대기모형과 해양 모형의 결합 및 수치해석적 접근법을 활용하여 계절내-계절 (Subseasonal to seasonal; S2S) 예측치를 생산하여 제공하고 있다. 이에 따라, 본 연구에서는 European Centre for Medium-Range Weather Forecast (ECMWF)에서 산정되는 수문기상인자 (강수량, 증발산량 및 유출량)에 대한 정확도를 평가하고자 한다. 연구지역으로는 다양한 기후대 및 토지 피복으로 구성되어 있으며, El-Nino-Southern Oscillation (ENSO), Indian Ocean Diapole (IOD)와 같은 기후 현상이 빈번히 발생하는 호주지역을 대상으로 연구를 수행하였다. ECMWF S2S 자료에 대한 통계적 검증은 1) 지점 기반 관측치와 더불어 2) 물수지 모델 기반 수문 추정치 (The Australian Water Resources Assessment Landscape Model; AWRA-L)와 비교하였다. 연구 결과 S2S 강우 및 증발산량 산정치의 경우 비교적 짧은 예측기간(약 2주)에서 상대적으로 높은 상관관계 (R=0.5~0.6)와 낮은 편차 (강수량 = 0.10 mm/day, 증발산량 = 0.21 mm/day)를 나타내었다. 유출량의 경우, 강우 및 증발산량에 비해 상대적으로 낮은 정확도를 나타내었으며, 예측 기간이 길어짐에 따라 불확실성이 상당히 높아지는 것으로 확인되었다. 이는, S2S 계산과정에서 강우 및 증발산량 뿐만아니라 지표 유출로 도달하기 전까지의 수문기상인자들의 불확실성이 모두 모여 유출량의 불확실성이 높아진 것으로 확인할 수 있었다. 계절적 검증에서는, 강우 및 증발산량 모두 여름철에 높은 상관관계를 나타내었지만 불확실성은 상대적으로 큰 값을 나타내었다. 자세한 분석을 위해, 공간적인 불확실성을 분석해본 결과 ECMWF S2S가 매우 습윤하거나 건조한 지역에서 수문기상인자를 예측하는데 있어 한계성이 나타난 것을 확인하였다. 본 연구를 토대로, 추후 S2S 예측치에 대한 보정과 더불어 미래의 수재해 발생 위험도에 대한 정보를 획득하는데 적용될 수 있을 것으로 판단된다.

• Journal of Advanced Navigation Technology
• /
• v.27 no.4
• /
• pp.481-487
• /
• 2023

Recently, interest in UAM (Urban Air Mobility) has surged as a critical solution to urban traffic congestion and air pollution issues. However, efficient UAM operation requires accurate 3D Point Cloud data processing, particularly in separating the ground and objects. This paper proposes and validates a method for effectively separating ground and objects in a UAM environment, taking into account its dynamic and complex characteristics. Our approach combines attitude information from MEMS sensors with ground plane estimation using RANSAC, allowing for ground/object separation that isless affected by GPS errors. Simulation results demonstrate that this method effectively operates in UAM settings, marking a significant step toward enhancing safety and efficiency in urban air mobility. Future research will focus on improving the accuracy of this algorithm, evaluating its performance in various UAM scenarios, and proceeding with actual drone tests.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.3
• /
• pp.115-121
• /
• 2024

As the radar mounted on a moving platform moves and rotates, the state of the radar's coordinate system also changes. At this time, in order to track target, the target's coordinates should be converted using the platform state measured from the sensor, and tracking performance may deteriorate due to causes such as sensor noise, communication delay, and sensor update cycle. In this paper, to minimize the degradation of tracking performance because of sensor error, we designed a navigation filter to estimate the state of the moving platform and analyzed the effect of improving tracking performance by applying the navigation filter through a simulation test. To design this navigation filter, three filter algorithms were applied and analyzed to confirm the effect of improving platform position and attitude performance for each filter, and the navigation filter designed by applying the highest performance filter algorithm was applied to a tracking simulation test. Finally we confirmed Improvement in tracking performance before and after applying navigation filters.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.1
• /
• pp.9-17
• /
• 2024

It is widely recognized that for 3D human pose estimation (HPE), dataset acquisition is expensive and the effectiveness of augmentation techniques of conventional visual recognition tasks is limited. We address these difficulties by presenting a simple but effective method that augments input images in terms of viewpoints when training a 3D human pose estimation (HPE) model. Our intuition is that meaningful variants of the input images for HPE could be obtained by viewing a human instance in the images from an arbitrary viewpoint different from that in the original images. The core idea is to synthesize new images that have self-occlusion and thus are difficult to predict at different viewpoints even with the same pose of the original example. We incorporate this idea into the training procedure of the 3D HPE model as an augmentation stage of the input samples. We show that a strategy for augmenting the synthesized example should be carefully designed in terms of the frequency of performing the augmentation and the selection of viewpoints for synthesizing the samples. To this end, we propose a new metric to measure the prediction difficulty of input images for 3D HPE in terms of the distance between corresponding keypoints on both sides of a human body. Extensive exploration of the space of augmentation probability choices and example selection according to the proposed distance metric leads to a performance gain of up to 6.2% on Human3.6M, the well-known pose estimation dataset.

• Journal of the Korea Concrete Institute
• /
• v.24 no.6
• /
• pp.727-735
• /
• 2012

Recently, a high-strength strand of 2400 MPa was developed using domestic technologies. In 2011, KS D 7002 was revised to cover the newly developed high-strength strands to support their practical usage. Presently, however, discussions and evaluations are not sufficient on the mechanical properties of the strands and their performance in structural members. Also, there were no detailed reviews on the need to revise the current design code for practical use of the high-strength strands. In this study, flexural behavior of a member with the high-strength strands was estimated through sectional analysis and a review and comparison of the domestic and foreign design codes were conducted considering the analysis results. Also, the need for the revision of the design code was discussed. Such discussion especially focused on the estimation of the stress in strand, which related with various issues such as determination methods for yield point of strands, time-dependent loss of prestressing force, estimation of stress in strand at member failure, and net strain limit for ductile failure of member. The discussion revealed that some parts in the design code need a revision and the further studies are required.