Machine vision-based method and system for aircraft docking guidance and aircraft type identification
First Claim
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.
2 Assignments
0 Petitions
Accused Products
Abstract
A machine vision-based method and system for aircraft docking guidance and aircraft type identification, comprising: S1: a monitoring scenario is divided into different information processing function areas; S2: a captured image is pre-processed; S3: the engine and the front wheel of an aircraft are identified in the image, so as to confirm that the aircraft has appeared in the image; S4: continuous tracking and real-time updating are performed on the image of the engine and the front wheel of the aircraft captured in step S3; S5: real-time positioning of the aircraft is implemented and the degree of deviation of the aircraft with respect to a guide line and the distance with respect to a stop line are accurately determined; S6: the degree of deviation of the aircraft with respect to the guide line and the distance with respect to the stop line of step S5 are outputted and displayed.
-
Citations
22 Claims
-
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 Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
-
21. A system for guiding an aircraft docking and recognizing an aircraft model based on machine vision, characterized in that the system comprises:
- a photographing device, a central processing device and a display device,
wherein the central processing device is configured to perform; 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, gay 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 Dependent Claims (22)
- a photographing device, a central processing device and a display device,
Specification