A novel vision-aided GNSS/INS Integration method using line features for aviation approach and landing

The approach and landing phase is the busiest and most demanding phase of aviation flight. To meet the high performance requirements in the approach and landing phase. vision system is recently discusscd as an aided navigation mean to be combined with the existing integration of global navigation satellite systems (GNSS) and inenial navigation system (INS). However, vision-aided navigation is quite challenging as the vision system is an angling system, whose navigation errors are approximately proportional to the length of line-of-sights. In this paper, to inhibit the errors of the vision system, line features are applied sincc they arc noise resistant and commonly in sight during the approach and landing phase, then a novel vision-aided GNSS/INS integration method for aviation approach and landing is proposed. First, a line detection method is designed to extract the horizon line and the airport runway lines for vision navigation. Second, a vision navigation model is proposed to obtain the diffidence equation of the aircraft states and the observations. Finally, an extended Kalman Filtering (EKF) method is applied to obtain the position and attitude estimation of the aircraft. Experimental results demonstrate the effectiveness of our proposed algorithm over the existing methods in the position and attitude estimation. Besides, the relative position of the aircraft rcspcct to the airport runways can be obtained, which can improve the flight safety during approach and landing operations.