基于双目视觉-惯性导航的轻型无人机导航算法

Because of the complexity of battlefield environment, the navigation algorithm of light UAV in GNSS signal rejection environment is very important. A nonlinear optimization-based stereo visual-inertial navigation odometer is proposed. The proposed algorithm starts with inertial measurement units (IMU) pre-integration, in which IMU measurements are accumulated between several frames using measurement pre-integration. In the initialization procedure, the pre-integrated IMU measurements and visual observations are tightly fused, and the initial velocity, direction of gravity, and gyroscope bias are estimated by using multiple view geometry (MVG) theory based on the feature-based method. After the initial state estimation converges, a highly precision stereo vision-inertial navigation odometer is obtained by fusing IMU measurements and feature observations. The proposed algorithm is validated on the EuRoC datasets. Experimental results prove that the proposed algorithm has higher accuracy and robustness than those of the most advanced visual-inertial fusion methods in some challenging situations.