Time filter

Source Type

Hide C.,University of Nottingham | Botterill T.,Geospatial Research Center Ltd.
Institute of Navigation - International Technical Meeting 2010, ITM 2010 | Year: 2010

This paper describes a scheme for pedestrian navigation integrating measurements from a foot-mounted IMU with position and orientation updates from computer vision techniques. By mounting an IMU on a user's foot, the position drift can be substantially reduced since zero velocity updates can be applied every step. However, such a system will still suffer from position drift unless occasional measurements are available from other sensors. This paper describes a novel method for restricting such position drift using an image recognition algorithm. Firstly, a database of images and their locations is constructed over an area of interest. A user then navigates the area using foot-mounted inertial sensors and a video camera. As images are acquired, they are used to search the database of images using the Image Bag-of-Words algorithm. When new images are successfully matched with images in the database, the position from the database is used to update the inertial position using a Kalman filter. Furthermore, when images are successfully matched, orientation updates can be applied by estimating the relative orientation of the two cameras. These measurements can help overcome the limitations of the IMU, in particular the problem with heading drift. The integrated inertial and vision system is demonstrated to provide better than 10m accuracy (typically 1-5m) over a period of 21 minutes, and the paper demonstrates how orientation updates could be applied in the future. Source

Fourie J.,University of Canterbury | Fourie J.,Geospatial Research Center Ltd. | Mills S.,Geospatial Research Center Ltd. | Green R.,University of Canterbury
Image and Vision Computing | Year: 2010

In this article a novel approach to visual tracking called the harmony filter is presented. It is based on the Harmony Search algorithm, a derivative free meta-heuristic optimisation algorithm inspired by the way musicians improvise new harmonies. The harmony filter models the target as a colour histogram and searches for the best estimated target location using the Bhattacharyya coefficient as a fitness metric. Experimental results show that the harmony filter can robustly track an arbitrary target in challenging conditions. We compare the speed and accuracy of the harmony filter with other popular tracking algorithms including the particle filter and the unscented Kalman filter. Experimental results show the harmony filter to be faster and more accurate than both the particle filter and the unscented Kalman filter. © 2010 Elsevier B.V. All rights reserved. Source

Discover hidden collaborations