• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.11
• /
• pp.1649-1654
• /
• 2021

In this paper, we introduce a regularization of long short-term memory (LSTM) based fall detection system using TensorFlow that can detect falls that can occur in the elderly. Fall detection uses data from a 3-axis acceleration sensor attached to the body of an elderly person and learns about a total of 7 behavior patterns, each of which is a pattern that occurs in daily life, and the remaining 3 are patterns for falls. During training, a normalization process is performed to effectively reduce the loss function, and the normalization performs a maximum-minimum normalization for data and a L2 regularization for the loss function. The optimal regularization conditions of LSTM using several falling parameters obtained from the 3-axis accelerometer is explained. When normalization and regularization rate λ for sum vector magnitude (SVM) are 127 and 0.00015, respectively, the best sensitivity, specificity, and accuracy are 98.4, 94.8, and 96.9%, respectively.

• Journal of Internet Computing and Services
• /
• v.10 no.3
• /
• pp.17-25
• /
• 2009

This paper presents a methodology for a fall detection using the feature extraction method based on the neural network with weighted fuzzy membership functions (NEWFM). The distributed non-overlap area measurement method selects the minimized number of input features by removing the worst input features one by one. Nineteen number of wavelet transformed coefficients captured by a triaxial accelerometer are selected as minimized features using the non-overlap area distribution measurement method. The proposed methodology shows that sensitivity, specificity, and accuracy are 95%, 97.25%, and 96.125%, respectively.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.516-518
• /
• 2022

In this paper, we introduce a dependence of number of nodes of hidden-layer in fall detection system using Long Short-Term Memory that can detect falls. Its training is carried out using the parameter theta(θ), which indicates the angle formed by the x, y, and z-axis data for the direction of gravity using a 3-axis acceleration sensor. In its learning, validation is performed and divided into training data and test data in a ratio of 8:2, and training is performed by changing the number of nodes in the hidden layer to increase efficiency. When the number of nodes is 128, the best accuracy is shown with Accuracy = 99.82%, Specificity = 99.58%, and Sensitivity = 100%.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.6
• /
• pp.1226-1274
• /
• 2022

In this paper, we dealt with a safety helmet for detecting dangerous situations that focuses on falling accidents and gas leaks, which are the main causes of industrial accidents. the fall situation range was set through gravity acceleration measurement using an acceleration sensor, and as a result, a fall detection rate of 80% could be confirmed. .In addition, the dangerous gas concentration was measured through a gas sensor, and when a digital value of 188 or more was output through a serial monitor, it was determined as a gas dangerous situation, and a fall warning message and a gas warning message could be checked through a smart-phone application produced based on the app inventor program.

• Journal of Sensor Science and Technology
• /
• v.31 no.2
• /
• pp.96-101
• /
• 2022

As the elderly population gradually increases, the risk of fatal fall accidents among the elderly is increasing. One way to cope with a fall accident is to determine the fall direction before impact using a wearable inertial measurement unit (IMU). In this context, a previous study proposed a method of classifying fall directions using a support vector machine with sensor velocity, acceleration, and tilt angle as input parameters. However, in this method, the IMU signals are processed through several processes, including a Kalman filter and the integration of acceleration, which involves a large amount of computation and error factors. Therefore, this paper proposes a machine learning-based method that classifies the fall direction before impact using IMU raw signals rather than processed data. In this study, we investigated the effects of the following two factors on the classification performance: (1) the usage of processed/raw signals and (2) the selection of machine learning techniques. First, as a result of comparing the processed/raw signals, the difference in sensitivities between the two methods was within 5%, indicating an equivalent level of classification performance. Second, as a result of comparing six machine learning techniques, K-nearest neighbor and naive Bayes exhibited excellent performance with a sensitivity of 86.0% and 84.1%, respectively.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.10a
• /
• pp.265-269
• /
• 2008

Home health care with compact wearable units sounds to be a convenient solution for the elderly people living independently. This paper presents a method to detect fall from the other activities of daily living and also to classify those activities. This kind of ambulatory monitoring enables them to get an emergency help in the case of the fatal fall event and can provide their general health status by observing the activities being performed in daily life. A tri-axial accelerometer sensor is used to get the acceleration anomalies associated with the user's movements. The three axis acceleration data are transferred to the base station sensor node via an IEEE 802.15.4 compliant zigbee module. The base station sensor node sends the data to base station PC for an offline processing. This work shows the feature set preparation using the principal component analysis (PCA) for the designing of neural network. The work includes the most common activities of daily living (ADL) like Rest, Walk and Run along with the detection of fall events from ADL. The angle from the vertical is found to be the most significant feature parameter for classification of fall while mean, standard deviation and FFT coefficients were used as the feature parameter for classifying the other activities under consideration. The accuracy for detection of fall events is 86%. The overall accuracy for ADL and fall is 94%.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.552-554
• /
• 2021

We propose a fall recognition system using a deep learning algorithm using motion data acquired by a Doppler radar sensor. Among the deep learning algorithms, an RNN that has an advantage in time series data is used to recognize falls. The fall data of the Doppler radar sensor has a temporal characteristic as time series data, and the structure of the RNN is sequenced because the result only determines whether a fall or not It is designed in a structure that outputs a fixed size to the input.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.2
• /
• pp.336-342
• /
• 2015

Fall-related injuries are the most common cause of accidental death for the elderly and the most frequent work-related injuries in construction sites. Due to the growing popularity of smartphones, there has been a number of research work related to the use of sensors embedded in the smartphone for fall detection. Falls can be detected easily by measuring the magnitude and direction of acceleration vectors. In general, the direction of the acceleration vector does not show the object movement, but the velocity vector directly indicates the tangential direction in which the object is moving. In this paper, we proposed a new method for computing the fall direction based on the characteristics of the velocity vector extracted from the accelerometer.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.491-500
• /
• 2021

In this study, we propose a method for detecting fall behavior using MS Kinect v2 RGBD Camera-based Human-Skeleton Keypoints and a 2-Stacked Bi-LSTM model. In previous studies, skeletal information was extracted from RGB images using a deep learning model such as OpenPose, and then recognition was performed using a recurrent neural network model such as LSTM and GRU. The proposed method receives skeletal information directly from the camera, extracts 2 time-series features of acceleration and distance, and then recognizes the fall behavior using the 2-Stacked Bi-LSTM model. The central joint was obtained for the major skeletons such as the shoulder, spine, and pelvis, and the movement acceleration and distance from the floor were proposed as features of the central joint. The extracted features were compared with models such as Stacked LSTM and Bi-LSTM, and improved detection performance compared to existing studies such as GRU and LSTM was demonstrated through experiments.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.2
• /
• pp.127-133
• /
• 2021

A recent study of people physically falling focused on analyzing the motions of the falls using a recurrent neural network (RNN) and a deep learning approach to get good results from detecting 2D human poses from a single color image. In this paper, we investigate a detection method for estimating the position of the head and shoulder keypoints and the acceleration of positional change using the skeletal keypoints information extracted using PoseNet from an image obtained with a low-cost 2D RGB camera, increasing the accuracy of judgments about the falls. In particular, we propose a fall detection method based on the characteristics of post-fall posture in the fall motion-analysis method. A public data set was used to extract human skeletal features, and as a result of an experiment to find a feature extraction method that can achieve high classification accuracy, the proposed method showed a 99.8% success rate in detecting falls more effectively than a conventional, primitive skeletal data-use method.