CN-121998993-A - Automatic cutting method for aviation ground service lifting platform integrating image processing

CN121998993ACN 121998993 ACN121998993 ACN 121998993ACN-121998993-A

Abstract

The invention discloses an automatic cutting method of an aviation ground service lifting platform integrating image processing, which particularly relates to the technical field of aviation maintenance auxiliary equipment, and comprises the steps of obtaining a multi-angle infrared image and a depth image of an aircraft target operation area, identifying target contour features, generating a three-dimensional contour feature set, projecting the feature set to an aircraft three-dimensional model, constructing an operation area point cloud image, estimating a height difference value and an edge curvature, generating a lifting platform foundation matching contour curve, combining platform size parameters and an operation task boundary, generating an optimal cutting track sequence, controlling a numerical control cutting device to complete automatic cutting of a structural plate, obtaining contact pressure distribution of the edge of the platform after cutting, combining the aircraft model to perform error comparison and compensation cutting, and automatically outputting a completion signal after the laminating precision reaches a set threshold.

Inventors

YU SHUIQIANG
LOU XIAOQING
ZHENG DONG
MA ZHENJIAN
XIONG HUAYUN

Assignees

杭州尚研机械制造有限公司

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (8)

1. An automatic cutting method of an aviation ground service lifting platform integrating image processing is characterized by comprising the following steps: Acquiring multi-angle image data of a target operation area of an aircraft, wherein the image data comprises at least one group of infrared images and depth images, and performing feature recognition on the image data through a fusion convolutional neural network and an edge extraction algorithm to generate a target contour feature set; Projecting the target profile feature set to a three-dimensional space model of the surface of the aircraft, constructing a three-dimensional point cloud picture of an operation area, estimating the height difference values and the edge curvatures of different areas, and generating a basic matching profile curve of the lifting platform; Based on the lifting platform foundation matching profile curve, combining the platform initial size parameter and the operation task boundary to generate an optimal cutting track sequence for cutting the platform structure; inputting the optimal cutting track sequence to a numerical control cutting device, and controlling the platform structural plate to automatically cut according to an optimal path; Feeding back contact pressure distribution after cutting of the platform structure, comparing the pressure distribution with a three-dimensional model of the surface of the aircraft in real time, and adjusting compensation cutting parameters of the part with the cutting error exceeding the threshold value; when the lamination precision of the cut platform structure and the target operation area reaches a preset tolerance threshold, outputting a cutting completion signal, and automatically entering a locking deployment state.
2. The method for automatically cutting the aviation ground service lifting platform by fusion image processing according to claim 1, wherein the step of constructing the three-dimensional point cloud image of the working area comprises the following steps: Based on the space coordinates of each three-dimensional edge point in the target contour feature set, calling a pre-established three-dimensional space model of the surface of the aircraft, and executing rigid coordinate transformation and scale matching processing to enable the edge points and the surface model of the aircraft to be in the same reference coordinate system; After coordinate alignment is completed, mapping each edge point to a grid unit corresponding to the aircraft surface model by adopting a nearest neighbor search mode to form a surface constraint projection point set; And performing density interpolation and discrete point compensation processing on the surface constraint projection point set to generate a continuously distributed three-dimensional point cloud image of the working area.
3. The method for automatically cutting the lifting platform for aviation ground service by fused image processing according to claim 2, wherein the step of generating the basic matching contour curve of the lifting platform comprises the following steps: based on the three-dimensional point cloud image of the operation area, carrying out space partition on the point cloud according to a preset grid size, extracting the maximum height value and the minimum height value of the edge points in each grid unit, and calculating an area height difference value; Fitting curve segments to edge point distribution of adjacent grid units, estimating edge curvature by adopting a second derivative change rate, and judging continuity and variability of edge change; Extracting a representative edge path point set as a control node of a cutting curve of the lifting platform according to a joint weight scoring rule of the height difference value and the edge curvature; And connecting the control nodes according to a space sequence, and generating a lifting platform basic matching contour curve in a B spline curve fitting mode.
4. The method for automatically cutting an aviation ground service lifting platform with fused image processing of claim 1, wherein the step of generating an optimal cutting track sequence for cutting a platform structure comprises the following steps: based on the lifting platform foundation matching profile curve, extracting a control node sequence, calculating Euclidean distance between adjacent nodes, and judging local variation amplitude of the curve; Constructing a platform boundary constraint model by combining initial size parameters of the lifting platform, geometrically cutting a matched profile curve and a platform boundary, and removing curve segments exceeding the processing range of the platform; according to the boundary information of the operation task, a multi-target path optimization algorithm is adopted, path continuity, minimum cutter moving distance and material utilization rate are set as optimization targets, and a cutting path topology is reconstructed; and executing five times of Bezier curve smoothing processing on the optimized path nodes to generate an optimal cutting track sequence which can be processed by the platform.
5. The method for automatically cutting an aviation ground service lifting platform by fused image processing according to claim 1, wherein the step of comparing the pressure distribution with a three-dimensional model of the surface of the aircraft in real time and adjusting the compensating cutting parameters of the part with the cutting error exceeding the threshold value comprises the following steps: after the cutting of the platform structure is completed, acquiring real-time pressure data of a contact area between the edge of the platform and the surface of the aircraft, and generating a pressure distribution map; Carrying out space point cloud registration on the contact pressure peak coordinates in the pressure distribution map and the three-dimensional model of the surface of the aircraft, and calculating a deviation vector field between the cutting contour and the theoretical laminating boundary; if the local deviation vector amplitude exceeds the cutting error threshold, extracting an initial cutting track point of the corresponding edge section; and correcting the coordinates of the track control points based on the deviation direction and the tangential component, updating the cutting path, and regenerating a processing control instruction for compensating cutting.
6. The method for automatically cutting the aviation ground service lifting platform by fusion image processing according to claim 1, wherein in the step of carrying out feature recognition on image data through a fusion convolutional neural network and an edge extraction algorithm, a multi-scale convolutional neural network model based on a U-Net structure is adopted to carry out edge region segmentation on an infrared image, the neural network model comprises at least five layers of encoders and five layers of decoders, and feature extraction is carried out through a convolutional group with a convolutional kernel size of 3×3.
7. The method for automatically cutting the aviation ground service lifting platform with the integrated image processing function according to claim 6, wherein the edge extraction algorithm is a Canny edge detection algorithm, image binarization processing is performed on the basis of an infrared image edge heat map, the binarization threshold value is automatically determined through an Otsu method, and an 8-neighborhood connection strategy is adopted for edge connection.
8. The method for automatically cutting the aviation ground service lifting platform with the integrated image processing according to claim 2, wherein in the step of constructing the three-dimensional point cloud image of the operation area, an inverse distance weighted interpolation method is adopted to conduct point supplementing processing on the sparse area, and the interpolation weight is the square inverse of Euclidean distance between interpolation points and adjacent edge points.

Description

Automatic cutting method for aviation ground service lifting platform integrating image processing Technical Field The invention relates to the technical field of aviation maintenance auxiliary equipment, in particular to an automatic cutting method of an aviation ground service lifting platform integrating image processing. Background The aviation ground service lifting platform is used as important auxiliary equipment for guaranteeing maintenance, overhaul, assembly and emergency rescue operation of a large-scale aircraft, and is widely applied to military and civil aviation bases. However, in the process of operation preparation and scene adaptation, the problem that the size of the existing lifting platform is not matched with the operation area generally exists, and particularly in the maintenance task of a remote or special-shaped structure body needing quick response, the platform cannot achieve accurate fitting of a complex outline structure, so that the operation efficiency is low, the risk is high, and even the surface structure of an aircraft is possibly damaged. The lifting platform structure of the current mainstream is mostly assembled by fixed modularization, and intelligent recognition and dynamic cutting capability are lacked. When the complex air inlet channel contour of a military fighter, the abnormal hanging frame area of an unmanned aerial vehicle, the propeller hub connection node of a rotor unmanned helicopter and other irregular spaces are met, the platform needs to be manually measured and manually cut, so that the deployment period is long, the operation precision is poor, the platform is easy to collide with a machine body due to misjudgment, and irreversible damage is caused. Especially in the low illumination environment at night, the temporary deployment site and the combat readiness state, the platform needs to have the capabilities of image guidance, self-adaptive recognition and automatic cutting, and can meet urgent demands of modern ground service on rapidity, high safety and module intelligence. Therefore, there is a need for an automatic cutting method for an aviation ground service lifting platform integrating image processing, which can realize self-adaptive cutting of a platform structure in a complex scene and improve the intellectualization and reliability of aviation ground service operation. Disclosure of Invention The invention aims to provide an automatic cutting method for an aviation ground service lifting platform integrating image processing, which aims to solve the defects in the background technology. In order to achieve the purpose, the invention provides the following technical scheme that the automatic cutting method of the aviation ground service lifting platform integrating image processing comprises the following steps: Acquiring multi-angle image data of a target operation area of an aircraft, wherein the image data comprises at least one group of infrared images and depth images, and performing feature recognition on the image data through a fusion convolutional neural network and an edge extraction algorithm to generate a target contour feature set; Projecting the target profile feature set to a three-dimensional space model of the surface of the aircraft, constructing a three-dimensional point cloud picture of an operation area, estimating the height difference values and the edge curvatures of different areas, and generating a basic matching profile curve of the lifting platform; Based on the lifting platform foundation matching profile curve, combining the platform initial size parameter and the operation task boundary to generate an optimal cutting track sequence for cutting the platform structure; inputting the optimal cutting track sequence to a numerical control cutting device, and controlling the platform structural plate to automatically cut according to an optimal path; Feeding back contact pressure distribution after cutting of the platform structure, comparing the pressure distribution with a three-dimensional model of the surface of the aircraft in real time, and adjusting compensation cutting parameters of the part with the cutting error exceeding the threshold value; when the lamination precision of the cut platform structure and the target operation area reaches a preset tolerance threshold, outputting a cutting completion signal, and automatically entering a locking deployment state. Preferably, the step of constructing the three-dimensional point cloud image of the working area includes: Based on the space coordinates of each three-dimensional edge point in the target contour feature set, calling a pre-established three-dimensional space model of the surface of the aircraft, and executing rigid coordinate transformation and scale matching processing to enable the edge points and the surface model of the aircraft to be in the same reference coordinate system; After coordinate alignment is completed, mapping each edge point to a g