Abstract
Motion analysis using wearable sensors is an essential research topic with broad mathematical foundations and applications in various areas, including engineering, robotics, and neurology. This paper presents the use of the global navigation satellite system (GNSS) for detecting and recording the position of a moving body, along with signals from additional sensors, for monitoring of physical activity and analyzing heart rate dynamics during running on route segments of different slopes and speeds.
This method provides an alternative to the heart monitoring on the treadmill ergometer in the cardiology laboratory. The proposed computational methodology involves digital data preprocessing, time synchronization, and data resampling to enable their correlation, feature extraction both in time and frequency domains, and classification.
The datasets include signals acquired during ten experimental runs in the selected area. The motion patterns detection involves segmenting the signals by analysing the GNSS data, evaluating the patterns, and classifying the motion signals under different terrain conditions.
This classification method compares neural networks, support vector machine, Bayesian, and $k$ -nearest neighbour methods. The highest accuracy of 93.3 % was achieved by using combined features and a two-layer neural network for classification into three classes with different slopes.
The proposed method and graphical user interface demonstrate the efficiency of multi-channel and multi-dimensional signal processing with applications in rehabilitation, fitness movement monitoring, neurology, cardiology, engineering, and robotic systems.