Lingxiang Zheng

A 3D indoor positioning system based on low-cost MEMS sensors

Abstract A positioning system in the absence of GPS is important in establishing indoor directional guidance and localization. Inertial Measuring Units (IMUs) can be used to detect the movement of a pedestrian. In this paper, we present a three-dimensional (3D) indoor positioning system using foot mounted low cost Micro-Electro-Mechanical System (MEMS) sensors to locate the position and attitude of a person in 3D view, and to plot the path travelled by the person. The sensors include accelerometers, gyroscopes, and a barometer. The pedestrians motion information is collected by accelerometers and gyroscopes to achieve Pedestrian Dead-Reckoning (PDR) which is used to estimate the pedestrian’s rough position. A zero velocity update (ZUPT) algorithm is developed to detect the standing still moment. A Kalman filter is combined with the ZUPT to eliminate non-linear errors in order to obtain accurate positioning information of a pedestrian. The information collected by the barometer is integrated with the accelerometer data to detect the altitude changes and to obtain accurate height information. The main contribution of this research is that the approach proposed fuses barometer and accelerometer in Kalman filter to obtain accurate height information, which has improved the accuracy at x axis and y axis. The proposed system has been tested in several simulated scenarios and real environments. The distance errors are around 1%, and the positioning errors are less than 1% of the total travelled distance. Results indicate that the proposed system performs better than other similar systems using the same low-cost IMUs.

Indoor Pedestrian Navigation with Shoe-Mounted Inertial Sensors

High accuracy in indoor navigation with shoe-mounted inertial sensors attracts a lot of researches in the last decades. In this paper, we build an indoor navigation system using a kind of estimation architecture with shoe-mounted inertial sensors. The architecture consists of zero velocity update (ZUPT) and extended Kalman filter (EKF). The ZUPT during the rest phase of a pedestrian’s foot can be used together with an EKF. The real time EKF runs to estimate the drift error and non-linear error growth of accelerometers and gyroscopes. The algorithm is inspected and verified on an experiment board. It presents a good performance. The position error of our algorithm is below 1% of the actual total traveled distance. It is feasible to obtain a long-term stability and high accuracy in an indoor navigation scenario.

A Foot-Mounted Sensor Based 3D Indoor Positioning Approach

We present a 3D indoor positioning system with foot mounted low cost MEMS sensors. The sensors includes the accelerometers, gyroscopes, and barometer. The output of accelerometers and gyroscopes are used to calculate the zero velocity update (ZUPT) and the movement of one step. The barometer is used to detect the altitude changes. A Kalman filter based framework is used to fusion the outputs of the sensors and estimate the non-linear errors of the position and heading, which increased over time. A particle filter is used to further reduce the errors. The test result shows that the designed system perform well.

A low power and high accuracy MEMS sensor based activity recognition algorithm

Wearable sensors and smart phones have been used in human activity recognitions and can achieve relative high accuracy however the power consumption is also high. In this paper, we propose an activity recognition approach that can achieve high accuracy with low power consumption. Two strategies have been applied to reduce the power consumption. The first strategy is using the hierarchical support vector machine classification algorithm to reduce the computational complexity. The second strategy is to reduce the sensor data sampling rates. Data collected from sensors in low sampling rate were processed using a wider time window for the feature extraction. The experiment results show that the average recognition accuracy of human activities (sitting, standing, walking, and running) in 1 Hz sampling rate can reach 98.50%. It indicates that the proposed approach can effectively extend the battery lifetime while maintaining high prediction accuracy in activity recognition.

A Smartphone Based Hand-Held Indoor Positioning System

In this paper, we present a smartphone-based hand-held indoor positioning system. The system collects data using the accelerometer, gyroscope and gravity virtual sensor sensors embedded in the smartphone. The accelerometer and gravity data are used to detect zero vertical speed and calculate the vertical displacement of each walking step, and then the Pythagorean Theorem is applied to calculate the step length of every step. Gyroscope data is used to estimate the direction angle. The step length and the direction angle of each step is combined to determine the coordinates of each step. A Kalman filter is used to reduce the vertical speed offset caused by accelerometer drift errors. The testing results show good performance of the proposed system.

A Robust Hand Tracking Approach Based on Modified Tracking-Learning-Detection Algorithm

Hand tracking is an essential step for dynamic gesture recognition which catches a lot of attention in the field of gesture interaction. In this paper, we present a robust hand tracking approach for unconstrained videos based on modified Tracking-Learning-Detection (TLD) algorithm, named BP-TLD. By introducing a skin color feature to the model, we make the algorithm more suitable for hand tracking. The experimental results show that BP-TLD has a better performance compared with other tracking algorithms such as TLD, MSEPF and Handvu. It indicates that our approach can meet the requirements of robustness and real-time better for the frontal-view vision-based human computer interactions.

A Novel Energy-Efficient Approach for Human Activity Recognition

In this paper, we propose a novel energy-efficient approach for mobile activity recognition system (ARS) to detect human activities. The proposed energy-efficient ARS, using low sampling rates, can achieve high recognition accuracy and low energy consumption. A novel classifier that integrates hierarchical support vector machine and context-based classification (HSVMCC) is presented to achieve a high accuracy of activity recognition when the sampling rate is less than the activity frequency, i.e., the Nyquist sampling theorem is not satisfied. We tested the proposed energy-efficient approach with the data collected from 20 volunteers (14 males and six females) and the average recognition accuracy of around 96.0% was achieved. Results show that using a low sampling rate of 1Hz can save 17.3% and 59.6% of energy compared with the sampling rates of 5 Hz and 50 Hz. The proposed low sampling rate approach can greatly reduce the power consumption while maintaining high activity recognition accuracy. The composition of power consumption in online ARS is also investigated in this paper.

A Novel Activity Recognition Approach Based on Mobile Phone

This paper presents a novel method for high-accuracy human activity recognition based on mobile phone acceleration sensors. Our approach includes two phases: the feature extraction phase and the classification phase. In feature extraction phase, we process tri-axial acceleration sensor data by combining the Independent Components Analysis (ICA) with the wavelet transform algorithm to get the features. In the classification phase, we apply the Support Vector Machine (SVM) algorithm to distinguish four types of activities (sitting, standing, walking and running). Experimental results show that the approach achieves an average accuracy of 98.78% over four types of activities, which outperforms the traditional method. The high accuracy indicates that this approach may facilitate the mobile phone based human activity recognition application.

An INS/WiFi Indoor Localization System Based on the Weighted Least Squares

For smartphone indoor localization, an INS/WiFi hybrid localization system is proposed in this paper. Acceleration and angular velocity are used to estimate step lengths and headings. The problem with INS is that positioning errors grow with time. Using radio signal strength as a fingerprint is a widely used technology. The main problem with fingerprint matching is mismatching due to noise. Taking into account the different shortcomings and advantages, inertial sensors and WiFi from smartphones are integrated into indoor positioning. For a hybrid localization system, pre-processing techniques are used to enhance the WiFi signal quality. An inertial navigation system limits the range of WiFi matching. A Multi-dimensional Dynamic Time Warping (MDTW) is proposed to calculate the distance between the measured signals and the fingerprint in the database. A MDTW-based weighted least squares (WLS) is proposed for fusing multiple fingerprint localization results to improve positioning accuracy and robustness. Using four modes (calling, dangling, handheld and pocket), we carried out walking experiments in a corridor, a study room and a library stack room. Experimental results show that average localization accuracy for the hybrid system is about 2.03 m.

Evolutionary Particle Filter for Indoor Navigation and Location

Indoor positioning technology can provide accurate location services for pedestrians. MEMS inertial sensors are inexpensive and small in size. Therefore, inertial navigation and positioning become popular research direction. The inertial sensor, which contains 3-axis accelerometer and 3-axis gyroscope, collects the acceleration and angular velocity information. The relative position of the pedestrian is calculated by integrating the acceleration and the angular velocity. The extended Kalman filter estimates attitude, angular velocity, position, velocity and acceleration system state errors. The system state error is updated when the foot touches the ground. Directional drift is the main problem of inertial navigation. Correcting heading by adding auxiliary basic information is one of the more common methods, such as GPS, geomagnetism, and Wi-Fi, but the additional basic information adds to the extra cost. We propose a novel algorithm based on the fact that pedestrians cannot cross the wall during walking. After the extended Kalman filter, the step size and the azimuth change are used as the observed state to establish the walking motion model. Considering the map information, the particle filter estimates the pedestrian position. For the particle impoverishment problem, the mutation operation of the genetic algorithm is used. A healthy male participates in the experiment. The results show an absolute error of 1.6 m.