A Research Of Attitude Determination And Positioning For Civil Aircraft Based On Computer Vision

Ramp incidents cause huge economic losses to airlines every year,most of which occur at the gate area.Advanced Visual Docking Guidance System(A-VDGS)provides an automatic way of aircraft docking,which can reduce accidents caused by human factors,and has been proved to greatly reduce the number of accidents at boarding gates.At present,the widely used A-VDGS mostly adopts three-dimensional laser technology.Because the structure of the laser equipment is more complex and its stability is not high,the maintenance is costly.With the maturity of computer vision technology,some systems combine laser technology with computing visual technology to improve the performance of the system.There are few docking systems based on computer vision.Most of the researches in this field in China are in the experimental stage.Therefore,it is innovative and meaningful to study the algorithm of attitude determination and location based on computer vision.On the basis of in-depth study of various image segmentation in computer vision technology,according to the characteristics of aircraft parking process,in this paper,we first use image segmentation to find the exact location of aircraft in video,then transform it to the relative position of aircraft in the airport,and finally calculate the orientation information and stop distance information of aircraft.The main contents are as follows:An image segmentation algorithm which combines the threshold method with the graph cutting method is proposed.First of all,the aircraft engine is detected,and the approximate area of the front wheel of the aircraft is determined according to the engine position.Then,the threshold segmentation method is used to preliminarily segment the front wheel of the aircraft.Finally,the precise location of the tire is determined by using the graph cutting algorithm.The experiment shows that the algorithm combines the automatic segmentation method with the interactive segmentation method effectively.On the one hand,the threshold segmentation algorithm greatly reduces the target area of graph cut algorithm,thereby improving the efficiency of graph cut algorithm.On the other hand,the final segmentation effect is better than that of threshold segmentation algorithm or graph cut algorithm.The algorithm not only has the efficiency of the threshold segmentation algorithm,but also has the accuracy of the graph cutting algorithm.A position orientation algorithm based on homography is proposed.First,a plane coordinate system is set up in the plane of the airport,and then it corresponds to the image plane coordinate system by using the homography.The coordinates of the aircraft on the airport plane are calculated by the aircraft coordinates obtained by the segmentation algorithm,and the distance between the azimuth and the stop line of the aircraft relative to the guide line is determined.The experiment shows that the method is simple and accurate.