Estimation of the Posture of Mobile Depth Sensor for Detecting the Ground and Walls in Terms of Initial Values

Abstract

The author previously applied Extended Kalman Filter for estimating the posture of depth sensor attached to a walking person or a mobile vehicle in conjunction with the surface of the plane, where the posture of the sensor was included in the state vector of the model. However, it could not adapt to situations when the initial value of Extended Kalman Filter was not appropriate. The initial value of the Extended Kalman Filter strongly affects the filtering results, especially for the model when the domain of the state vector has several groups. Different starting point leads to different results. In this paper, setting method of the initial value of the algorithm is proposed. Trial and error approach has proven to work well and has a good property to adapt to various situations. By defining the reset conditions differently, it is shown that the system can detect different surfaces.