Civil Aircraft ISS Oriented Runway Surveillance Algorithm Research

The safety of civilian aircraft increased significantly with the development of aviation technology. Especially, the application of Integrated Surveillance System (ISS) can provide traffic, weather, terrain and other information to flight crew and give the optimal strategies to avoid hazard. ISS is an efficient solution of safety when the aircraft is in the air. However, the ISS can't supplies adequate safety insurance when the aircraft is approaching with low visibility or on the airport runway. On the one hand, aircraft landing accident rate dominates in the entire flight procedures. When the aircraft reaches decision height, the landing procedure can be carried on if the pilot can see the runway clearly and the flight status is normal. Current ISS can help pilots determine the fitness of flight status, while it can't help the pilot to see the runway clearly with low visibility. On the other hand, the TCAS function of ISS function made the aircraft collision accidents in the air almost disappeared. However, the accidents caused by collision between the aircraft and other aircrafts or ground vehicles on the runway and the accidents that the ground crew's deaths caused by inhalation into aircraft engine occur quite frequently. Also, the TCAS vertical collision avoidance strategy and the hazard of high radiation power to peoples on the ground and other features currently and in the future limit its possibilities to solve collision accidents between aircraft and the objects on the runway. ARINC 768 standardizes the interface functional specifications of ISS, and it also suggests new features for future ISS function which includes runway monitoring and alerting capabilities. This feature provides the flight crew with the information on and around the airport runway, thereby enhancing pilot situational awareness of the runway environment and will help to improve flight safety. This paper introduces runway surveillance methods into ISS to meet two practical demands: improving pilots'awareness of runway environment under low visibility conditions during approaching and preventing collisions between the aircraft and other objects on the runway.(1) To improve pilots'awareness of runway environment under low visibility conditions during approaching, the paper puts up a novel ISS Runway Recognition and Approach Assistance function. The IR images of the front vision are taken and the runways are detected and recognized to help pilots see the runway clearly in low visibility conditions and improve flight safety. The image pre-processing methods are first introduced to provide proper data for detection and recognition. Then the lines filtering, runway lines filtering, runway area filtering rules are designed after lines detection. The undetected error and insane error elimination methods are designed to improve the robustness of runway recognition.(2) To avoid the collisions between aircraft and other objects, the paper puts up a novel ISS Runway Traffic Collision Avoidance function. The imaging equipment are used to track and survey other aircraft, vehicles, people and other objects on the runway to help pilots to avoid collisions. The first solution combine the particle filter tracking algorithm based on adaptive spatio-temporal codebook detection model of are proposed as the first solution. The algorithm adds adaptive process to spatio-temporal codebook which can efficiently improve the objects detection abilities and suppress noise disturbance. The detection results are followed by particle filter tracking algorithm which can efficiently improve the reliability of the tracking algorithm. Compare with current codebook methods the adaptive codebook has the robustness against noise disturbance. Particle filters using detection results can swiftly capture moving objects and provide reliable tracking process. To implement tacking and surveillance, the mean shift tracking algorithm using variational bayesian inference theory frame as the first solution. In this algorithm, the position and shape of the ellipsoidal region from the previous frame are used as the initial values. We can simultaneously gain the unbiased estimation of the position of the tracked object and the ellipse that approximates the shape of the object through simple iterative procedure. This algorithm is able to robustly track the objects in different situations. It can also adapt to changes in shape and scale of the object.These two functions together above form the ISS Runway Surveillance function, which can compensate the defect that the ISS can only ensure the safety of aircraft in the air. While the ISS with Runway Surveillance can supply the whole flight procedures protections which includes glide out from the terminal, take off, cruise, landing, taxiing back to the terminal.