CN-121999036-A - Hinge angle measuring method, device and system of rodless aircraft traction system

CN121999036ACN 121999036 ACN121999036 ACN 121999036ACN-121999036-A

Abstract

The invention discloses a non-contact measurement method, a device and a system for the hinge angle of a rodless aircraft traction system, wherein an industrial photogrammetry camera is arranged on the central axis of a tractor, the camera and the ground level are ensured, the rotation characteristics of the aircraft in the traction operation process are shot by adopting the camera, mask segmentation is carried out on the aircraft in an image through a pre-training YOLOv n-seg neural network, the image is output through graying after type conversion, the specific coordinate information of the aircraft contour is obtained by applying an edge detection operator, the optimal contour coordinate is obtained through Gaussian filtering and abnormal edge processing, and the measured hinge angle information is obtained by utilizing ellipse fitting. The invention realizes the non-contact high-precision measurement of the hinge angle of the aircraft traction system by combining an advanced image processing technology and a machine learning algorithm, has the advantages of visual and convenient reading of the measurement result and high degree of automation, and is suitable for an automatic and intelligent aircraft traction operation environment.

Inventors

Zhu hengjia
XU ZISHUO
XU YITONG
ZHANG WEI

Assignees

中国民航大学

Dates

Publication Date: 20260508
Application Date: 20241107

Claims (10)

1. A method for non-contact measurement of articulation angle of a rodless aircraft traction system, the method comprising: acquiring an aircraft image; performing mask segmentation on the aircraft image by using a pre-training model to obtain an image comprising an aircraft contour; carrying out gray processing on the image comprising the plane outline to obtain a binary image; according to the binary image, utilizing an edge detection operator to obtain coordinate information of the aircraft contour; Performing ellipse fitting on the coordinate information of the aircraft contour to obtain ellipse axis information; And obtaining the hinge angle information of the rodless aircraft traction system according to the elliptical axis information.
2. The method for non-contact measurement of the articulation angle of a rodless aircraft traction system according to claim 1, characterized in that the pre-training model is obtained by training through a neural network, comprising: Constructing an image dataset with instance segmentation tags, the image dataset comprising an aircraft image, the instance segmentation tags comprising information of a target frame and a sampled image; Training and learning the image data set by adopting a neural network model, wherein a loss function adopted in the training process is as follows: Loss=1-CIOU Wherein IOU is the ratio of intersection and union of the real target frame and the neural network model prediction target frame in the image dataset, (x pre ,y pre ) is the center point position information of the neural network model prediction target frame, (x rea ,y rea ) is the center point position information of the real target frame in the image dataset, c is the diagonal distance of the minimum closure region containing the real target frame and the prediction target frame at the same time, d is used for measuring the similarity of the aspect ratio, w rea and h rea are the width and the height of the real target frame respectively, and w pre and h pre are the width and the height of the prediction target frame respectively.
3. The method for non-contact measurement of the articulation angle of a rodless aircraft traction system of claim 1, the method is characterized in that the acquiring of the aircraft image comprises the following steps: Detecting the shot image and judging whether an effective target is detected or not; if not, the image is shot again until an effective target is detected; Wherein the effective target comprises an aircraft nose belly.
4. The method for non-contact measurement of the articulation angle of a rodless aircraft traction system according to claim 1, characterized in that said obtaining the coordinate information of the aircraft contour using an edge detection operator comprises: and detecting edges and obtaining the coordinate information of the optimized plane contour by using a Canny operator and an abnormal edge processing method.
5. The method of non-contact measurement of articulation angle for a rodless aircraft traction system of claim 4, wherein the detecting edges by Canny operator comprises: carrying out Gaussian filtering treatment on the binary image; the Sobel operator is used to calculate the gradient intensity and direction of each pixel point in the image, According to the gradient strength and the gradient direction, a strong edge and a weak edge are determined by adopting a double-threshold method; tracking the boundary by adopting a hysteresis technology, determining a weak edge connected with a strong edge, and determining the final position of the edge to obtain the coordinate information of the aircraft contour: P=[(x 1 ,y 1 ),...(x n ,y n )] Where P is a set of contour points, and (x i ,y i ) is coordinate information of the ith contour point.
6. The method for non-contact measurement of the articulation angle of a rodless aircraft traction system of claim 5, wherein the abnormal edge handling method comprises: carrying out area calculation on each contour detected by the Canny operator, and carrying out abnormal edge detection on the contour with the largest area; the abnormal edge detection of the contour with the largest area comprises the following steps: And calculating whether any one of the adjacent vertical line segments is parallel to the coordinate axis of the pixel coordinate system, detecting whether the length of the line segment exceeds a minimum threshold value, and deleting edge point information corresponding to the line segment if the length of the line segment exceeds the minimum threshold value to obtain the coordinate information of the optimized plane contour.
7. The method of claim 6, wherein said performing an ellipse fit on the coordinate information of the aircraft profile comprises: defining an ellipse fitting function as: F(α,X)=ax 2 +bxy+cy 2 +dx+ey+f=0 Wherein, alpha= [ a, b, c, d, e, f ] T ,X＝[x 2 ,xy,y 2 ,x,y,1] T , Establishing an optimization function And solving the elliptic parameters by adopting a least square method.
8. A non-contact measurement device for the articulation angle of a rodless aircraft traction system, the device comprising: the aircraft image acquisition module is used for acquiring aircraft images; the mask segmentation module is used for carrying out mask segmentation on the aircraft image by utilizing the pre-training model to obtain an image comprising the aircraft outline; The binary image acquisition module is used for carrying out gray processing on the image comprising the airplane outline to obtain a binary image; the coordinate information acquisition module is used for acquiring coordinate information of the aircraft contour by utilizing an edge detection operator according to the binary image; the ellipse fitting module is used for performing ellipse fitting on the coordinate information of the aircraft contour to obtain ellipse axis information; and the hinge angle information acquisition module is used for acquiring the hinge angle information of the rodless aircraft traction system according to the elliptical axis information.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any of claims 1-7 when the program is executed.
10. A non-contact measurement system for the articulation angle of a rodless aircraft traction system, characterized in that it comprises a camera module, a rodless aircraft tractor system, a computer device, which is the computer device of claim 9.

Description

Hinge angle measuring method, device and system of rodless aircraft traction system Technical Field The invention belongs to the technical field of artificial intelligence and measurement, and particularly relates to a non-contact measurement method, device and system for a hinge angle of a rodless aircraft traction system and computer equipment. Background In the field of modern aviation, ground traction operations of an aircraft are important components for ensuring safe and efficient operation of the aircraft. The rodless aircraft tractor is used as important airport ground guarantee equipment, is widely used by virtue of the advantages of small occupation of working space, low energy consumption and the like, and has important significance for researching the safety and the intellectualization of the aircraft traction operation by researching the hinge angle measuring method of the rodless aircraft traction system. The conventional method for measuring the hinge angle of the traction system of the rodless aircraft is not mature, the conventional method for measuring the hinge angle needs to install inertial navigation devices on the tractor and the towed system respectively, and the hinge angle is obtained by measuring the angle difference value of the tractor and the towed system, but the method is inconvenient to install and high in manufacturing cost, and the tractor cannot directly obtain the pose information of the aircraft due to the particularity of the aircraft traction system, so that the non-contact hinge angle measurement of the traction system of the rodless aircraft cannot be completed. Disclosure of Invention Aiming at the defects or improvement demands of the prior art, the invention provides a non-contact measurement method, a non-contact measurement device and computer equipment for the hinge angle of a rodless aircraft traction system, which are used for performing non-contact high-precision measurement on the hinge angle of the aircraft traction system, are intuitive and convenient in measurement result reading, have high automation degree, and are suitable for an automatic and intelligent aircraft traction operation environment. In order to achieve the above purpose, the invention adopts the following technical scheme: In some embodiments, a method of non-contact measurement of articulation angle of a rodless aircraft traction system is provided, the method comprising: acquiring an aircraft image; performing mask segmentation on the aircraft image by using a pre-training model to obtain an image comprising an aircraft contour; carrying out gray processing on the image comprising the plane outline to obtain a binary image; according to the binary image, utilizing an edge detection operator to obtain coordinate information of the aircraft contour; Performing ellipse fitting on the coordinate information of the aircraft contour to obtain ellipse axis information; And obtaining the hinge angle information of the rodless aircraft traction system according to the elliptical axis information. In some embodiments, the pre-training model is trained by a neural network, comprising: Constructing an image dataset with instance segmentation tags, the image dataset comprising an aircraft image, the instance segmentation tags comprising information of a target frame and a sampled image; Training and learning the image data set by adopting a neural network model, wherein a loss function adopted in the training process is as follows: Loss=1-CIOU Wherein IOU is the ratio of intersection and union of the real target frame and the neural network model prediction target frame in the image dataset, (x pre,ypre) is the center point position information of the neural network model prediction target frame, (x rea,yrea) is the center point position information of the real target frame in the image dataset, c is the diagonal distance of the minimum closure region containing the real target frame and the prediction target frame at the same time, d is used for measuring the similarity of the aspect ratio, w rea and h rea are the width and the height of the real target frame respectively, and w pre and h pre are the width and the height of the prediction target frame respectively. In some embodiments, the acquiring an aircraft image includes: Detecting the shot image and judging whether an effective target is detected or not; if not, the image is shot again until an effective target is detected; Wherein the effective target comprises an aircraft nose belly. In some embodiments, the obtaining coordinate information of the aircraft contour using the edge detection operator includes: and detecting edges and obtaining the coordinate information of the optimized plane contour by using a Canny operator and an abnormal edge processing method. In some embodiments, the detecting edges by a Canny operator includes: carrying out Gaussian filtering treatment on the binary image; the Sobel operator is used to calculate the gr