Autonomous systems demand high accuracy of positioning. The visual-inertial navigation system (VINS) can provide accurate positioning in a short period but suffer from drift over time. Besides, outlier features caused by moving objects and unstable illuminations degrade the performance of the VINS. Contrarily, the global navigation satellite system (GNSS) can provide reliable and globally referenced positioning in open areas but signal reflections and blockages from surrounding buildings make it challenging in urban canyons. This research presented a sliding window factor graph optimization (FGO) based GNSS/Visual/IMU/Map integration to exploit the complementariness of GNSS and VINS. First, the graduated non-convexity (GNC) is employed to mitigate the outliers involved in the features and GNSS pseudorange measurements. Particularly, the window carrier phase (WCP) and the Doppler frequency are also explored to constrain the relative motion of the consecutive epochs. Second, a novel sliding window (SW) based map matching model is proposed to correct the states using the lightweight OpenStreetMap (OSM). Unlike conventional filtering-based map matching, the states within the sliding window of the FGO are associated with the OSM lane information, effectively exploiting the measurement redundancy from the factor graph model. The effectiveness of the proposed method will be validated using challenging datasets.
Autonomous systems demand high accuracy of positioning. The visual-inertial navigation system (VINS) can provide accurate positioning in a short period but suffer from drift over time. Besides, outlier features caused by moving objects and unstable illuminations degrade the performance of the VINS. Contrarily, the global navigation satellite system (GNSS) can provide reliable and globally referenced positioning in open areas but signal reflections and blockages from surrounding buildings make it challenging in urban canyons. This research presented a sliding window factor graph optimization (FGO) based GNSS/Visual/IMU/Map integration to exploit the complementariness of GNSS and VINS. First, the graduated non-convexity (GNC) is employed to mitigate the outliers involved in the features and GNSS pseudorange measurements. Particularly, the window carrier phase (WCP) and the Doppler frequency are also explored to constrain the relative motion of the consecutive epochs. Second, a novel sliding window (SW) based map matching model is proposed to correct the states using the lightweight OpenStreetMap (OSM). Unlike conventional filtering-based map matching, the states within the sliding window of the FGO are associated with the OSM lane information, effectively exploiting the measurement redundancy from the factor graph model. The effectiveness of the proposed method will be validated using challenging datasets.