双目视觉辅助GNSS在恶劣环境下导航定位

视觉里程计能够在复杂环境下提供短时间的高精度导航定位,全球卫星导航系统(GNSS)具有全天候、全球性和误差不随时间积累的特性,但是在恶劣环境多路径效应下,GNSS定位精度会变差甚至不可用.为了研究在复杂环境下视觉里程计辅助GNSS导航定位技术,首先介绍了视觉里程计的导航定位原理;然后在卡尔曼滤波器中将GNSS定位结果和视觉里程计定位结果进行了松组合处理;并利用视觉里程计定位结果和预测的视觉里程计误差实现了GNSS在恶劣环境下的导航定位.基于KITTI数据集的模拟验证结果表明,设计的组合方案能够在恶劣环境下持续提供可靠的导航定位.

Stereo vision-aided GNSS navigation and positioning in harsh environments

The visual odometer can provide short-term and precise navigation and positioning in complex environments.GNSS has the characteristics of all-weather.global and no accumulate error over time.But under the multipath effect of harsh environments,GNSS positioning accuracy will be worse or even unusable.In order to study the visual odometer-aided GNSS navigation and positioning technology in complex environments,the navigation positioning principle of the visual odometer is introduced firstly;then the GNSS positioning results and the visual odometer positioning results are loosely combined in a Kalman filter;utilizing the visual odometer positioning results and the predicted visual odometer positioning errors to achieve GNSS navigation positioning in harsh environments.The simulation verification results based on the KITTI data set show that the designed combined scheme can continuously provide reliable navigation and positioning in harsh environments.