Human activity recognition (HAR) has been actively researched in fields such as healthcare to understand and analyze human behavior in human-robot interaction. However, most studies have struggled to recognize activities like turning and motion transitions, which are often associated with dynamic balance. Therefore, we propose a novel HAR approach using a single sensor to collect and early fuse motion and position data. The aim is to enhance the accuracy of motion classification for daily activities and those that cause imbalance, which have traditionally been difficult to recognize. We constructed a quarantine room environment for data collection and to evaluate the impact of the suggested features on behavior. Five deep learning models were trained and evaluated to identify the optimal model. The collected data was classified and analyzed by the selected model, which demonstrated an average accuracy of 98.96%.

This paper proposes an IMU method for location tracking in power plants and indoor environments without GPS. IMU-based sensors use accelerometer, angular accelerometer, earth magnetometer, and altimeter. It is a method for recognizing the movement of pedestrians or moving objects. However, errors can be caused, as noise and bias increase due to long-term measurement. VIO-SLAM type sensor T265, which uses a combination of cameras and IMU, and can accurately track paths in invisible spaces, is used in this study. In addition, this type of sensor can be corrected in real time with a filter function inserted into the sensor and errors can be minimized. As a comparison experiment with the encoder, it is possible to evaluate the location of the scanner within a ±10 mm error from the actual distance in 1,500 × 700 (mm) space. The usefulness of this method is verified by measuring real specimens of boiler pipes and tubes, which are the major components of power plants.