Lidar-Millimeter Wave Radar Information Fusion Multi-Target Detection Based on Unscented Kalman Filter and Covariance Intersection Algorithm

EasyChair Preprint no. 6977

Abstract

Aiming at the problems of missed detection of occluded vehicles and long-distance targets when using Lidar for target detection in intelligent driving vehicles,  this paper proposes a lidar-millimeter wave radar information fusion multi-target detection method based on the unscented Kalman filter and the covariance intersection algorithm. According to the data collected by the sensor, the UKF is used to generate the state estimation, and the CI algorithm is used to form the state estimation into a fusion state estimation. The effectiveness of the method is verified by Matlab simulation experiments, and compared with the algorithm based on Joint Probabilistic Data Association (JPDA) and Gaussian mixture probability hypothesis density (GMPHD) algorithm. The result is UKF-CI algorithm has higher accuracy for multi-target detection, and the effect is more obvious.

Keyphrases: Covariance Intersection Algorithm, information fusion, multi-target detection, Unscented Kalman filter algorithm

@Booklet{EasyChair:6977,
  author = {Fan Le and Hong Mo and Yinghui Meng},
  title = {Lidar-Millimeter Wave Radar Information Fusion Multi-Target Detection Based on Unscented Kalman Filter and Covariance Intersection Algorithm},
  howpublished = {EasyChair Preprint no. 6977},

  year = {EasyChair, 2021}}