No abstract

No abstract

Efficient multi‐sensor fusion is crucial to provide accurate pose estimates for navigating various next‐generation autonomous vehicles such as self‐driving cars, personal air vehicles, urban air mobilities, and electronic vertical take‐off and landing aircraft with respect to a unified global reference frame. The integration of the Global Navigation Satellite System (GNSS) and Inertial Navigation System (INS) provides globally uniform coordinate information, and the Visual‐Inertial Odometry (VIO) gives stable and accurate local pose estimates. However, the error of the integrated INS/GNSS increases rapidly in GNSS‐challenged environments, and the VIO system suffers from the long‐term error accumulation. To cope with these difficulties, this paper proposes a new tightly coupled Visual‐Inertial‐GNSS (VIG) system to achieve robust and drift‐free pose estimation utilising GNSS raw measurements, image features, and inertial measurements. A novel compressed state constraint Kalman Filter (CSCKF) is formulated to combine time‐propagated feature measurements and differential GNSS observations to aid INS. Compared to the conventional integrated navigation systems, the proposed CSCKF‐based VIG system is advantageous in maintaining the minimum number of states without unnecessary state augmentation and no loss of performance. The proposed system is evaluated by field experiments in different GNSS availability situations. The results show that accuracy is improved significantly in GNSS‐degraded environments compared to that of the conventional systems.

This paper proposes a multi-baseline method to estimate absolute coordinates of point clouds and the camera attitude parameter utilizing feature points in successive images. Conventionally, 3D map generation methodologies have been based on images acquired by aerial or land vehicles. Pixels corresponding to known landmarks are manually identified at first. Next, the coordinates are directly geo-referenced and automatically allocated to pixels with high-quality Global Positioning System (GPS) and Inertial Navigation System (INS). However, it is difficult to obtain accurate coordinates by the conventional methodology with low-cost GPS and INS. With camera positions and attitude parameters, image-based point clouds can be compensated accurately. A simulation was carried out to evaluate the performance of the proposed method.