An automatic identification method for airport operator management

Abstract

The invention belongs to the technical field of video analysis, in particular to an automatic identification method for airport operator management. The method uses artificial intelligence algorithm to identify the aircraft parking state, and can extract the data of several frames before and after the aircraft parking to process, so as to save computational power and reduce the false alarm rate. The main advantages are as follows: 1, Using artificial intelligence algorithm to identify the aircraft parking status, the all-day monitoring can be realized through the intelligent recognition algorithm, and the cost of manual monitoring is saved; 2, By extracting the data of several frames just before and after the aircraft stops for processing, the computational power can be saved and the falsealarm rate can be reduced. 3, The automatic identification of aircraft parking and the detection of welcoming personnel by using template method can be changed flexibly with the alteration of apron and civil aviation specifications in practice, and the scheme has strong portability.

Description

technical field [0001] The invention belongs to the technical field of video analysis, and in particular relates to an automatic identification method used for airport operator management. Background technique [0002] The management of the personnel on the tarmac is an important aspect of airport operation management. It is required that two staff members guide the aircraft to the specified position during the process of the aircraft landing to a stable stop. One person is located at the nose wheel position when the aircraft is in a stable stop, and one person is located in a position within the pilot's line of sight in front of the nose of the aircraft when the aircraft is in a stable stop. The management of the on-board personnel focuses on two aspects: first, the aircraft must be in place at a certain time before landing; second, the aircraft is required to be properly guided during the docking process. [0003] At present, the management of the onboard personnel is mai...

AI Technical Summary

Problems solved by technology

[0004] With the development of computer vision technology, some people propose to use the method of target recognition to identify the person on board, so as to analyze the time and behavior of him in place. This method requires real-time processing of each frame of data, which consumes a lot of computing power. And there are many false alarms. After actual testing, if the method of processing each frame is used, the false alarm rate is about 25%.

[0005] According to the current operation and management mode, manual inspection is adopted. The scope and time of inspection are very limited, and 24-hour inspection cannot be achieved. If you want to inspect multiple aircraft positions at the same time, you need to send additional manpower, which is time-consuming and labor-intensive. It is very uneconomical; if video or image browsing is used for patrol inspection, there will be a huge amount of browsing workload and the problem of easy missed inspections, and video browsing is a way of tracing after the event, and timely warning cannot be achieved

[0006] If computer vision technology is used to process pictures in real time, there will be a huge amount of data, too many processing targets, and no pertinence, especially the interference and occlusion caused by irrelevant scenes will cause high false detection rate and missed detection rate. no less

Images