This short paper is a preliminary report of a gait analysis system using body-installed RGB-D camera and inertial sensors. The camera is installed on the waist and the camera is looking downward so that it can observe foot and floor. During walking, a stance foot and floor is detected, where a stance foot is used as a landmark. A Kalman filter is proposed based on visual inertial odometry. Also, a smoothing algorithm is proposed. Using the proposed algorithm, various gait parameters such as stride length and walking speed can be estimated, which can be used to monitor health conditions of older adults and evaluate functional recovery of injured people. The proposed algorithms are tested with 10 meter straight line walking.

This paper proposed a reliable method for estimating the neck tilting angle while walking using a 3-axis accelerometer. Conventionally tilting angle is estimated by observing the gravitational force that applies on the sensor in static pose. However, while walking the external acceleration of the movement might heavily affect the estimation result. Therefore, we proposed a simple method based on human gait characteristics and step detection to eliminate the external forces in the accelerometer output. The experiment was done with five persons walking in a treadmill with different speeds under the observation of a camera system. The result showed that the proposed method provided an accurate estimation compared with conventional method of direct estimation.

An algorithm for estimating the walking stick movement information is proposed using an inertial sensor attached on the stick. A standard inertial navigation algorithm using an indirect Kalman filter is applied to update velocity and position of the walking stick during movement. The proposed algorithm is verified with three-meter walking experiments.

Welcome to HSI2017, the 10th International Conference on Human System Interactions in 2017 was held at the University of Ulsan in Ulsan, Republic of Korea. The University of Ulsan have organized the conference and the conference is technically co-sponsored by IEEE Industrial Electronics Society. HSI conference series has been one of the most important academic meetings in the field of interactions between human and systems. Until now the HSI conference series have been held in Krakow (Poland) 2008, Catania (Italy) 2009, Rzeszow (Poland) 2010, Yokohama (Japan) 2011, Perth (Australia) 2012, Gdansk (Poland) 2013, Lisbon (Portugal) 2014, Warsaw (Poland) 2015, and Portsmouth (United Kingdom) 2016.

A gait velocity estimation algorithm using the inertial sensors of a smartwatch is proposed. The peaks of accelerometer and gyroscope norms are detected at first. Then a Kalman Filter is employed to recover the peaks that are missed because of the arm swing. The Kalman filter combines the accelerometer and gyroscope norm peaks and robustly detect walking step events even in cases where there is a large arm swing. Walking velocity is then estimated using the step duration. It will be shown in this work that the gait velocity has a good correlation with the inverse of the square of the step duration. The model parameters are calculated by collecting the training data from 25 subjects: each subject walked 50 m six times with different walking speed and different arm swing speed. The standard deviation of walking velocity estimation error is 0.1009 m/s (without person dependent calibration) and 0.0630 m/s (with person dependent calibration). The average precision of 91.7% was achieved for the gait speed testing on the smartwatch platform over all the speed scenarios.

This paper is concerned with compression of multiple inertial sensor data. The compressed data consists of ADPCM compressed data and an adaptive refinement data. Each sensor data is first compressed using ADPCM. To reduce the quantization errors, refinement quantizers are then used, where the total number of bits for the refinement quantizers is fixed. Which inertial sensor data should be further refined is adaptively determined based on quantization error bounds. In the proposed method, total number of bits for the compressed data is fixed; however the number of bits for individual inertial sensor is adaptively determined. Through experiments, it is shown that the proposed method provides better compression quality than the standard ADPCM with the same total number of bits.

A new remote control method of a moving robot is proposed, where a moving robot is moved according to the pointing position and orientation of the remote controller. The remote controller consists of the camera and gyroscopes. The landmark in a moving robot is recognized by the camera in the remote controller and a robot is moved in the camera's pointing position and orientation. The 'virtual link' term is used since the robot is moved as if there is a link between the robot and the remote controller. Gyroscopes are also used in the remote controller so that fast estimation of the camera's pointing position and orientation is possible. The proposed method is verified through experiments.