Inertial Sensor Nodes Research Articles

Data Glove with Bending Sensor and Inertial Sensor Based on Weighted DTW Fusion for Sign Language Recognition

There are numerous communication barriers between people with and without hearing impairments. Writing and sign language are the most common modes of communication. However, written communication takes a long time. Furthermore, because sign language is difficult to learn, few people understand it. It is difficult to communicate between hearing-impaired people and hearing people because of these issues. In this research, we built the Sign-Glove system to recognize sign language, a device that combines a bend sensor and WonderSense (an inertial sensor node). The bending sensor was used to recognize the hand shape, and WonderSense was used to recognize the hand motion. The system collects a more comprehensive sign language feature. Following that, we built a weighted DTW fusion multi-sensor. This algorithm helps us to combine the shape and movement of the hand to recognize sign language. The weight assignment takes into account the feature contributions of the sensors to further improve the recognition rate. In addition, a set of interfaces was created to display the meaning of sign language words. The experiment chose twenty sign language words that are essential for hearing-impaired people in critical situations. The accuracy and recognition rate of the system were also assessed.

Robotic and sensor technology for frozen shoulder Rehabilitation

A common clinical shoulder problem is frozen shoulder. The degree of shoulder joint movement must be measured as part of the rehabilitation procedure. These assessments can be used to assess the severity of a patient's disease, set rehabilitation goals, and determine the appropriate level of activity difficulty, as well as comprehend the rehabilitation's impacts. In the healthcare industry, robots are rapidly being used; as they are in many other industries, Specifically for hospital logistics support, surgery, and rehabilitation. Because rehabilitation affects millions of individuals all around the world, there has been steady growth in the rehabilitation robotics sector over the last decade, with the advent of new technology aimed at overcoming the different challenges that this industry faces. A model is proposed in this study for identifying movements produced during frozen shoulder rehabilitation activities. The model is made up of wearable wireless sensor network (WSN) inertial sensor nodes that were built particularly for this investigation and enable for the identification of inertial sensor nodes of physiological movements anywhere. We describe current robotic systems for shoulder rehabilitation in this work, with an emphasis on indicators and further emerging technologies that, when used with robots, can enhance the advantages of rehabilitation for regaining shoulder function.

Wearable Sensor Clothing for Body Movement Measurement during Physical Activities in Healthcare.

This paper presents a wearable wireless system for measuring human body activities, consisting of small inertial sensor nodes and the main hub for data transmission via Bluetooth for further analysis. Unlike optical and ultrasonic technologies, the proposed solution has no movement restrictions, such as the requirement to stay in the line of sight, and it provides information on the dynamics of the human body’s poses regardless of its location. The problem of the correct placement of sensors on the body is considered, a simplified architecture of the wearable clothing is described, an experimental set-up is developed and tests are performed. The system has been tested by performing several physical exercises and comparing the performance with the commercially available BTS Bioengineering SMART DX motion capture system. The results show that our solution is more suitable for complex exercises as the system based on digital cameras tends to lose some markers. The proposed wearable sensor clothing can be used as a multi-purpose data acquisition device for application-specific data analysis, thus providing an automated tool for scientists and doctors to measure patient’s body movements.

Paddle Stroke Analysis for Kayakers Using Wearable Technologies.

Proper stroke posture and rhythm are crucial for kayakers to achieve perfect performance and avoid the occurrence of sport injuries. The traditional video-based analysis method has numerous limitations (e.g., site and occlusion). In this study, we propose a systematic approach for evaluating the training performance of kayakers based on the multiple sensors fusion technology. Kayakers’ motion information is collected by miniature inertial sensor nodes attached on the body. The extend Kalman filter (EKF) method is used for data fusion and updating human posture. After sensor calibration, the kayakers’ actions are reconstructed by rigid-body model. The quantitative kinematic analysis is carried out based on joint angles. Machine learning algorithms are used for differentiating the stroke cycle into different phases, including entry, pull, exit and recovery. The experiment shows that our method can provide comprehensive motion evaluation information under real on-water scenario, and the phase identification of kayaker’s motions is up to 98% validated by videography method. The proposed approach can provide quantitative information for coaches and athletes, which can be used to improve the training effects.

Retracted] Image Recognition of Badminton Swing Motion Based on Single Inertial Sensor

This article analyzes the method of reading data from inertial sensors. We introduce how to create a 3D scene and a 3D human body model and use inertial sensors to drive the 3D human body model. We capture the movement of the lower limbs of the human body when a small number of inertial sensor nodes are used. This paper introduces the idea of residual error into the deep LSTM network to solve the problem of gradient disappearance and gradient explosion. The main problem to be solved by wearable inertial sensor continuous human motion recognition is the modeling of time series. This paper chooses the LSTM network which can handle time series as well as the main frame. In order to reduce the gradient disappearance and gradient explosion problems in the deep LSTM network, the structure of the deep LSTM network is adjusted based on the residual learning idea. In this paper, a data acquisition method using a single inertial sensor fixed on the bottom of a badminton racket is proposed, and a window segmentation method based on the combination of sliding window and action window in real‐time motion data stream is proposed. We performed feature extraction on the intercepted motion data and performed dimensionality reduction. An improved Deep Residual LSTM model is designed to identify six common swing movements. The first‐level recognition algorithm uses the C4.5 decision tree algorithm to recognize the athlete’s gripping style, and the second‐level recognition algorithm uses the random forest algorithm to recognize the swing movement. Simulation experiments confirmed that the proposed improved Deep Residual LSTM algorithm has an accuracy of over 90.0% for the recognition of six common swing movements.

Canoeing Motion Tracking and Analysis via Multi-Sensors Fusion.

Coaches and athletes are constantly seeking novel training methodologies in an attempt to improve athletic performance. This paper proposes a method of rowing sport capture and analysis based on Inertial Measurement Units (IMUs). A canoeist’s motion was collected by multiple miniature inertial sensor nodes. The gradient descent method was used to fuse data and obtain the canoeist’s attitude information after sensor calibration, and then the motions of canoeist’s actions were reconstructed. Stroke quality was performed based on the estimated joint angles. Machine learning algorithm was used as the classification method to divide the stroke cycle into different phases, including propulsion-phase and recovery-phase, a quantitative kinematic analysis was carried out. Experiments conducted in this paper demonstrated that our method possesses the capacity to reveal the similarities and differences between novice and coach, the whole process of canoeist’s motions can be analyzed with satisfactory accuracy validated by videography method. It can provide quantitative data for coaches or athletes, which can be used to improve the skills of rowers.

Towards Wearable-Inertial-Sensor-Based Gait Posture Evaluation for Subjects with Unbalanced Gaits.

Human gait reflects health condition and is widely adopted as a diagnostic basis in clinical practice. This research adopts compact inertial sensor nodes to monitor the function of human lower limbs, which implies the most fundamental locomotion ability. The proposed wearable gait analysis system captures limb motion and reconstructs 3D models with high accuracy. It can output the kinematic parameters of joint flexion and extension, as well as the displacement data of human limbs. The experimental results provide strong support for quick access to accurate human gait data. This paper aims to provide a clue for how to learn more about gait posture and how wearable gait analysis can enhance clinical outcomes. With an ever-expanding gait database, it is possible to help physiotherapists to quickly discover the causes of abnormal gaits, sports injury risks, and chronic pain, and provides guidance for arranging personalized rehabilitation programs for patients. The proposed framework may eventually become a useful tool for continually monitoring spatio-temporal gait parameters and decision-making in an ambulatory environment.

Inertial Sensor-Based Analysis of Equestrian Sports Between Beginner and Professional Riders Under Different Horse Gaits

A correct riding posture is essential for a good performance in equestrian. This paper proposes a method of equestrian analysis based on body sensor networks. First of all, raw sensor data corresponding to professional (PRO) riders and beginners (BEGs) under different horse gaits were collected by inertial sensor nodes. Then, after an initial sensor alignment with the reduced error, a kind of motion reconstruction method based on the gradient descent method was used to fuse the sensor data to constantly update the rider’s attitude information. In addition, to evaluate the tracking accuracy of joint angle, an Northern Digital optical system was deployed to verify the accuracy of our method. Using the attitude information, the motions of riders in equestrian sports were reconstructed by combining the method of human kinematics. Finally, in view of the two aspects of exercise intensity and joint angle, a quantitative analysis of key features between PRO and BEG in walking and rising trot was carried out by combining with the kinematic analysis method. Experiments conducted in this paper demonstrate that the estimation errors are well controlled, and our method can better reflect the similarities and differences between PRO and BEG in equestrian sports and provide quantitative data for equestrian coaches or riders, which can be used to improve the riding skills of riders.

Gamified platform for rehabilitation after total knee replacement surgery employing low cost and portable inertial measurement sensor node

This paper introduces an innovative gamified rehabilitation platform comprising of a mobile game and a custom sensor placed on the knee, intended for patients that have undergone Total Knee Replacement surgery, in collaboration with the General Hospital in Chania. Initial testing of the system is conducted in the Hospital Orthopaedic Clinic, in collaboration with Orthopeadic Surgeons and Physiotherapists. The application uses a single custom-made, light, portable and low-cost sensor node consisting of an Inertial Measurement Unit (IMU) attached on a lower limb in order to capture its orientation in space in real-time, while the patient is completing a physiotherapy protocol. The aim is to increase patient engagement during physiotherapy by motivating the user to participate in a game. The proposed sensor node attached on the lower limb provides input to the gamified experience displayed on an Android mobile device, offering feedback to the patient in relation to whether the performed exercises were accurately conducted. A classification algorithm is proposed that automatically classifies an exercise in real-time as correct or incorrect, according to physiotherapists’ set criteria. The game projects a graphical image of the patient’s limb motion as part of a 3D computer graphics scene. It then classifies the exercise performed during physiotherapy as accurately performed or not and increases patient compliance via a reward system. Our goal is to reduce the need for the physical presence of a physiotherapist by aiding the efficient performance of exercise sessions at any location, e.g. at home, indoors and outdoors by just utilizing a light sensor and an Android device. Initial testing of the application in the Chania’s General Hospital Orthopaedic Clinic, Greece, indicates that patient engagement is enhanced in most cases, even when elderly patients are concerned.

Quantitative assessment of motor function in patients with Parkinson's disease using wearable sensors

Motor dysfunction is the main clinical symptom and diagnosis basis of patients with Parkinson's disease (PD). A total of 30 subjects were recruited in this study, including 15 PD patients (PD group) and 15 healthy subjects (control group). Then 5 wearable inertial sensor nodes were worn on the bilateral upper limbs, lower limbs and waist of subjects. When completing the 6 paradigm tasks, the acceleration and angular velocity signals from different parts of the body were acquired and analyzed to obtain 20 quantitative parameters which contain information about the amplitude, frequency, and fatigue degree of movements to assess the motor function. The clinical data of the two groups were statistically analyzed and compared, and then Back Propagation (BP) Neural Network was used to classify the two groups and predict the clinical score. The final results showed that most of the parameters had significant difference between the two groups, ten times of 5-fold cross validation showed that the classification accuracy of the BP Neural Network for the two groups was 90%, and the predictive accuracy of Hoehn-Yahr (H-Y) staging and unified PD rating scale (UPDRS) Ⅲ score of the patients were 72.80% and 68.64%, respectively. This study shows the feasibility of quantitative assessment of motor function in PD patients using wearable sensors, and the quantitative parameters obtained in this paper may have reference value for future related research.

Aerobic Exercise Recognition Through Sparse Representation Over Learned Dictionary by Using Wearable Inertial Sensors

Aerobic exercise is conducive to reducing the risks of cardiovascular disease and central arterial stiffness. However, it can also cause some health hazards (such as tissue oxidative damage), especially for the elderly. It is essential to recognize and monitor different aerobic exercises for the health of exercisers. In this paper, a multi-sensor monitoring system is established for aerobic exercise recognition, and a novel recognition algorithm based on dictionary learning algorithm and sparse representation is proposed. Eight volunteers are invited to carry out ten activities, and five wireless inertial sensor nodes are used to collect the sensor data. Several experiments are implemented to verify the effectiveness of the recognition algorithm proposed in the paper. According to the experimental results, our method achieves the best performance than four other recognition algorithms including decision tree C4.5, naive Bayes, support vector machine and sparse representation. Besides, the other two aspects are also studied in the paper, one is the effect of different binding positions of sensors on classification results, and the other is the effect of selecting different features. The results of the experiments show that two sensor nodes attached to the right wrist and the left thigh achieve better result, and the feature “correlation coefficient” is not important to recognize different aerobic exercises that are investigated in our paper.

An activity recognition model using inertial sensor nodes in a wireless sensor network for frozen shoulder rehabilitation exercises.

This paper proposes a model for recognizing motions performed during rehabilitation exercises for frozen shoulder conditions. The model consists of wearable wireless sensor network (WSN) inertial sensor nodes, which were developed for this study, and enables the ubiquitous measurement of bodily motions. The model employs the back propagation neural network (BPNN) algorithm to compute motion data that are formed in the WSN packets; herein, six types of rehabilitation exercises were recognized. The packets sent by each node are converted into six components of acceleration and angular velocity according to three axes. Motor features such as basic acceleration, angular velocity, and derivative tilt angle were input into the training procedure of the BPNN algorithm. In measurements of thirteen volunteers, the accelerations and included angles of nodes were adopted from possible features to demonstrate the procedure. Five exercises involving simple swinging and stretching movements were recognized with an accuracy of 85%–95%; however, the accuracy with which exercises entailing spiral rotations were recognized approximately 60%. Thus, a characteristic space and enveloped spectrum improving derivative features were suggested to enable identifying customized parameters. Finally, a real-time monitoring interface was developed for practical implementation. The proposed model can be applied in ubiquitous healthcare self-management to recognize rehabilitation exercises.

A system of human vital signs monitoring and activity recognition based on body sensor network

Purpose – The purpose of this paper is to develop a health monitoring system that can measure human vital signs and recognize human activity based on body sensor network (BSN). Design/methodology/approach – The system is mainly composed of electrocardiogram (ECG) signal collection node, blood oxygen signal collection node, inertial sensor node, receiving node and upper computer software. The three collection nodes collect ECG signals, blood oxygen signals and motion signals. And then collected signals are transmitted wirelessly to receiving node and analyzed by software in upper computer in real-time. Findings – Experiment results show that the system can simultaneously monitor human ECG, heart rate, pulse rate, SpO2 and recognize human activity. A classifier based on coupled hidden Markov model (CHMM) is adopted to recognize human activity. The average recognition accuracy of CHMM classifier is 94.8 percent, which is higher than some existent methods, such as supported vector machine (SVM), C4.5 decision tree and naive Bayes classifier (NBC). Practical implications – The monitoring system may be used for falling detection, elderly care, postoperative care, rehabilitation training, sports training and other fields in the future. Originality/value – First, the system can measure human vital signs (ECG, blood pressure, pulse rate, SpO2, temperature, heart rate) and recognizes some specific simple or complex activities (sitting, lying, go boating, bicycle riding). Second, the researches of using CHMM for activity recognition based on BSN are extremely few. Consequently, the classifier based on CHMM is adopted to recognize activity with ideal recognition accuracies in this paper.

Impact of Sensor Misplacement on Dynamic Time Warping Based Human Activity Recognition using Wearable Computers.

Daily living activity monitoring is important for early detection of the onset of many diseases and for improving quality of life especially in elderly. A wireless wearable network of inertial sensor nodes can be used to observe daily motions. Continuous stream of data generated by these sensor networks can be used to recognize the movements of interest. Dynamic Time Warping (DTW) is a widely used signal processing method for time-series pattern matching because of its robustness to variations in time and speed as opposed to other template matching methods. Despite this flexibility, for the application of activity recognition, DTW can only find the similarity between the template of a movement and the incoming samples, when the location and orientation of the sensor remains unchanged. Due to this restriction, small sensor misplacements can lead to a decrease in the classification accuracy. In this work, we adopt DTW distance as a feature for real-time detection of human daily activities like sit to stand in the presence of sensor misplacement. To measure this performance of DTW, we need to create a large number of sensor configurations while the sensors are rotated or misplaced. Creating a large number of closely spaced sensors is impractical. To address this problem, we use the marker based optical motion capture system and generate simulated inertial sensor data for different locations and orientations on the body. We study the performance of the DTW under these conditions to determine the worst-case sensor location variations that the algorithm can accommodate.

Feasibility Study and Performance Analysis of a Gyroless Orientation Tracker

Inertial orientation tracking systems commonly use three types of sensors: accelerometers, magnetometers, and gyroscopes. The angular rate signal is used to obtain a dead reckoning estimate, whereas the gravitational and local magnetic field measures allow us to apply a correction and to obtain a drift-free result. Considering the present market of inertial MEMS sensors, the current consumption of gyroscopes represents a major part of the power budget of wireless inertial sensor nodes, which should be minimized given the mobility of the application. This paper introduces an orientation tracking algorithm, based on an unscented Kalman filter, that does not require angular rate data for tracking human movements up to 450 °/s , which is a reasonable value for many applications. Since accelerometers measure other accelerations beside gravity and magnetometers are prone to magnetic disturbances, adaptive techniques are applied in order to reduce the influence on the estimations. The performance of the system is quantitatively analyzed and compared to an estimator that includes angular rate information.

A Wearable Inertial Sensor Node for Body Motion Analysis

A wearable wireless inertial sensor node for body motion analysis is designed and implemented. A triaxial accelerometer, a biaxial gyroscope, and a yaw rate gyroscope are employed to sense the accelerations and angular rates of the object attached. These three ICs, a wireless mote, the power circuit, and a modified printed inverted-F antenna are integrated on a four-layer printed circuit board. To achieve the compact size for comfortable wearing, the printed antenna is deliberately designed as small as possible, while maintaining a reasonable antenna performance. The proposed inertial sensor node is also easily adaptable to applications with different power requirements because of the consideration of the periodic and moving-event wakeup in the software design. The static characteristics and Allan deviations of the implemented node are measured and analyzed. The false alarm rate of a moving detection based on Bayes rule is presented and a threshold is suggested. Finally, the raw data of several body motions are measured and those behaviors are observed apparently. The proposed inertial sensor node is compact and easily wearable; hence, it is feasible to be applied to the body motion analysis.

Structural Action Recognition in Body Sensor Networks: Distributed Classification Based on String Matching

Mobile sensor-based systems are emerging as promising platforms for healthcare monitoring. An important goal of these systems is to extract physiological information about the subject wearing the network. Such information can be used for life logging, quality of life measures, fall detection, extraction of contextual information, and many other applications. Data collected by these sensor nodes are overwhelming, and hence, an efficient data processing technique is essential. In this paper, we present a system using inexpensive, off-the-shelf inertial sensor nodes that constructs motion transcripts from biomedical signals and identifies movements by taking collaboration between the nodes into consideration. Transcripts are built of motion primitives and aim to reduce the complexity of the original data. We then label each primitive with a unique symbol and generate a sequence of symbols, known as motion template, representing a particular action. This model leads to a distributed algorithm for action recognition using edit distance with respect to motion templates. The algorithm reduces the number of active nodes during every classification decision. We present our results using data collected from five normal subjects performing transitional movements. The results clearly illustrate the effectiveness of our framework. In particular, we obtain a classification accuracy of 84.13% with only one sensor node involved in the classification process.