×

Machine vision-based method and system for aircraft docking guidance and aircraft type identification

  • US 10,290,219 B2
  • Filed: 07/02/2015
  • Issued: 05/14/2019
  • Est. Priority Date: 08/01/2014
  • Status: Active Grant
First Claim
Patent Images

1. A method for guiding an aircraft docking and recognizing an aircraft model based on machine vision, characterized in that the method comprises:

  • a step for configuring an aircraft berthing scene, in which a monitored scene is divided into regions for separate data processing;

    an image pre-processing step, in which a captured image is pre-processed;

    an aircraft capturing step, in which it is determined that an aircraft exists in the image by recognizing an engine and front wheels of the aircraft in the image;

    an aircraft tracking step, in which the image of the engine and the front wheels of the aircraft recognized in the aircraft capturing step is continuously tracked and updated in real time;

    an aircraft positioning step, in which the aircraft is positioned in real time, and a deviation degree of the aircraft from a guide line and a distance between the aircraft and a stop line are accurately determined; and

    an information displaying step, in which the deviation degree of the aircraft from the guide line and the distance between the aircraft and the stop line determined in the aircraft positioning step are outputted and displayed, wherein the aircraft capturing step further comprises;

    a background elimination step, in which a dynamic distribution of a background in the scene is simulated with a single Gaussian background model, a model of the background is established and the present image and the background model are differentiated to eliminate the background to obtain a foreground region;

    a shade elimination step, gray values of respective pixels in the foreground region are computed to find the maximum gray value g max and the minimum gray value g min thereof, and shading elimination is performed on the regions with gray values lower than T=g min+(g max−

    g min)*0.5;

    a region classification step, in which a template of a standard front view of an aircraft region is created, a target region is extracted through change detection, and a vertical projection curve of the target region is calculated, a correlation coefficient between the vertical projection curve of the target region and a vertical projection curve of the template of the standard front view of the aircraft region is calculated, if the correlation coefficient is larger than or equals to a classification threshold, the target region is determined as an aircraft; and

    a feature verification step, in which it is further determined whether the target region is an aircraft by detecting the recognized engine and front wheel of the aircraft.

View all claims
  • 2 Assignments
Timeline View
Assignment View
    ×
    ×