An Improved algorithm for moving object tracking based on EKF
- Authors: Hou, Leichao , Qu, Junsuo , Zhang, Ruijun , Wang, Ting , Ting, KaiMing
- Date: 2018
- Type: Text , Conference proceedings
- Relation: The Euro-China conference on intelligent data analysis and applications; published in Proceedings of the Fifth Euro-China Conference on Intelligent Data Analysis and Applications. ECC 2018. p. 483-490
- Full Text: false
- Reviewed:
- Description: Kalman filter estimates the desired signal from the amount of measurement related to the extracted signal, which is widely used in engineering due to its simple calculation and easy programming on a computer. However, the basic theory originally proposed by Rudolf E. Kalman is for linear systems only, whereas a realistic physical system is often nonlinear. Extended Kalman Filter (EKF) solves nonlinear filtering problems. In this paper, we focus on issues related with targeted object being occluded We combine EKF and Meanshift to track the moving object. Once the object position is predicted by EKF in the center of the object, then the Meanshift algorithm iterates over the initial value of EKF estimation to track the object. Experiments show that the method reduces the object search time and improves the accuracy of the object tracking.
An improved measurement variable estimation model for positioning mobile robot
- Authors: Qu, Junsuo , Hou, Leichao , Zhang, Ruijun , Zhang, Zhiwei , Zhang, Qipeng , Ting, Kaiming
- Date: 2019
- Type: Text , Journal article
- Relation: Interaction Studies Vol. 20, no. 1 (2019), p. 78-101
- Full Text: false
- Reviewed:
- Description: The localization and navigation technology are the key factors in the research of mobile robots. With the demand of smart manufacturing industry and the development of robotics technology, the importance of mobile robot has become increasingly prominent. Mobile robot positioning research is mostly based on odometry, however, it has cumulative errors that would affect the accuracy of positioning results. This paper describes an improved measurement model that suitable from 0 degrees to 180 degrees and used this model in the Extended Kalman Filter (EKF) and Unscented Kalman Filter(UKF) time update step respectively, the method can address the interference of kinematics model predicted position and heading angle, both of them are easily disturbed by noises and other factors. Designing a tracked mobile robot as experimental platform to collect the raw data, conducting experimental research including the performance of hardware platform and autonomous obstacle avoidance, the real-time and stability of remote data interaction, and the accuracy of optimal pose estimation. The simulation results have been verified the accuracy of the improved measurement model applied to UKF.
Research on EKF-based localization method of tracked mobile robot
- Authors: Qu, Junsuo , Zhang, Qipeng , Hou, Leichao , Zhang, Ruijun , Ting, Kaiming
- Date: 2017
- Type: Text , Conference proceedings
- Relation: 2nd International Conference on Computer Engineering, Information Science & Application Technology (ICCIA 2017); Wuhan, China; 8th-9th July 2017; published in ACSR-Advances in Computer Science Research series Vol. 74, p. 175-180
- Full Text:
- Reviewed:
- Description: To estimate the position and heading angle of mobile robot precisely, an measurement variable estimation model was proposed to adapt any angle. Fusing the predictive value of odometry and measurement data of multiple sensors by the Extended Kalman Filtering (EKF) for reducing the accumulative error by using only traditional odometry. The proposed models is verified by Matlab simulation and experimental results.