Human Activity Recognition Using Thigh Angle Derived from Single Thigh Mounted IMU Data
MetadataShow full item record
Accurate human activity recognition is a challenging topic of research in many areas. A common approach to activity recognition is to use accelerometers and/or gyroscopes to detect trunk or leg movement. This paper present a novel approach to detect human activities based on thigh angle computed using data from a single thigh mounted Inertial Measurement Unit (IMU). As this work forms a component of a system underdevelopment to assist the vision impaired in indoor navigation, activities common in indoor pedestrian tracking such as sitting, standing and walking were considered in the development of the algorithm. This algorithm uses simple signal processing techniques including peak detection, zero crossing detection and timers to identify the activity based on the thigh angle computed by fusing accelerometer and gyroscope. This allows implantation of the algorithm in a general purpose low end microcontroller. To reduce the number of input parameters to the algorithm, it was assumed that accelerometer y–axis is aligned with the thigh such that gyroscopic x–data represents angular velocity of the forward and backward movement of the thigh. The algorithm has shown above 78% accuracy in detecting standing, above 92%accuracy for walking and no measured errors for sitting, in a test conducted with a limited number of samples with ideal testing conditions. These results indicate that this less computationally intense algorithm gives promising results in activity detection in indoor pedestrian navigation applications.
Showing items related by title, author, creator and subject.
Abhayasinghe, Nimsiri; Murray, Iain (2012)Way finding is one of the main difficulties that vision impaired people face, especially in indoor environments. Although Global Positioning System (GPS) based navigation is possible outdoors, the accuracy of GPS is not ...
Kramer, Annika (2009)Visual perception is our most important sense which enables us to detect and recognise objects even in low detail video scenes. While humans are able to perform such object detection and recognition tasks reliably, most ...
Human gait modelling with step estimation and phase classification utilising a single thigh mounted IMU for vision impaired indoor navigationAbhayasinghe, Kahala Nimsiri (2016)This research is focused on human gait modelling for infrastructure free inertial navigation for vision impaired. A pedometer based on a single thigh mounted gyroscope, an efficient algorithm to estimate thigh flexion and ...