Method, device, equipment and storage medium for identifying corner features of aircraft structural parts

Abstract

This application discloses a method, apparatus, device, and storage medium for identifying corner features of aircraft structural parts. By performing a closed process on a two-dimensional contour image of an aircraft structural part at any viewing angle, eliminating interference from missing pixels, and obtaining a contour image of the structural part, the contour line is truncated at the corner points of the contour image to obtain multiple arc lines. The center of the characteristic area is determined based on the distribution of the arc lines, and the rounded corner feature and bottom corner feature in the image are then determined. Finally, the corner feature is determined based on the distribution of the rounded corner area and bottom corner area. The method of this application can identify corner features at any viewing angle without any restrictions on the projection angle, i.e., no manual identification is required, thus achieving automated corner feature recognition and improving recognition efficiency and accuracy.

Description

Technical Field

[0001] The present application relates to the field of manufacturing and processing of aviation structural parts, and in particular to a method, device, equipment and storage medium for identifying corner features of aircraft structural parts. Background Art

[0002] The machining of aircraft structural components relies heavily on CNC machine tools. Only when materials meet strength and stiffness requirements, high-precision machine tools, and well-designed CNC machining programs are used can these components be produced to meet these requirements. However, due to the complexity of aircraft structural components, existing methods for determining the properties of patch spatial topological relationships are unable to describe all possible scenarios. Programming methods rely heavily on manual experience, and as a result, CNC machining programs in the aviation industry have always been primarily developed using manual programming.

[0003] Before programming a CNC machining program, it is necessary to identify and analyze the structural feature units in the three-dimensional model of the structural part. This is one of the most core tasks of programming, but the manual identification of corner features is inefficient. Summary of the Invention

[0004] The main purpose of this application is to provide a method, device, equipment and storage medium for identifying corner features of aircraft structural parts, aiming to solve the technical problem of low efficiency in manual identification of corner features.

[0005] To achieve the above objectives, the present application provides a method for identifying corner features of aircraft structural parts, comprising:

[0006] Acquire a contour image of the aircraft structural component to be programmed; wherein the contour image is obtained by performing feature region closing processing on a two-dimensional projection image of a three-dimensional model of the aircraft structural component to be programmed;

[0007] Performing cluster center processing on the extreme points of the contour image to obtain corner points of the contour image;

[0008] Based on the corner points, the contour line of the contour image is cut off to obtain the center of the feature area;

[0009] Identifying the rounded corner area and the bottom corner area of the contour image according to the center of the characteristic area;

[0010] A corner feature recognition result of the aircraft structural component to be programmed is obtained according to the fillet area and the bottom angle area.

[0011] Optionally, before the step of obtaining the contour image of the aircraft structural component to be programmed, the method further includes:

[0012] Acquiring a two-dimensional projection image of the three-dimensional model of the aircraft structural component to be programmed;

[0013] Performing expansion processing on the low-grayscale area of the two-dimensional projection image at multiple angles using a 1*2 convolution kernel to obtain expansion results of the low-grayscale area at each angle;

[0014] Fusing the multiple low-grayscale area expansion results to obtain a closed contour image;

[0015] At all angles, a 2*2 convolution kernel is used to expand the high grayscale area of the closed contour image to obtain the contour image.

[0016] Optionally, the step of performing expansion processing on the low-grayscale area of the two-dimensional projection image at multiple angles using a 1*2 convolution kernel to obtain the low-grayscale area expansion result at each angle includes:

[0017] The low-grayscale area of the two-dimensional projection image is expanded using a 1*2 convolution kernel at angles of 0°, 45°, 90°, and 135°, respectively, to obtain low-grayscale area expansion results at four angles.

[0018] Optionally, the step of performing cluster center processing on the extreme points of the contour image to obtain corner points of the contour image includes:

[0019] Performing filtering on the contour image to obtain a plurality of filtered images;

[0020] performing differential processing on the filtered image to obtain a plurality of extreme points;

[0021] Perform cluster center processing on the extreme points to obtain corner points of the contour image.

[0022] Optionally, the step of truncating the contour line of the contour image based on the corner point to obtain the center of the feature area includes:

[0023] Draw a reference circle with a radius of a with the corner point as the center; where a is a preset value;

[0024] Based on the reference circle, the contour line of the contour image is cut off to obtain a line segment image; wherein the line segment image contains a plurality of circular arc segments;

[0025] Performing full circle fitting on the arc segments to obtain multiple initial circles;

[0026] Obtaining multiple first circle centers according to the radius of the initial circle, the coordinates of the center of the initial circle and a preset radius range;

[0027] The initial circles corresponding to the first circle centers are grouped to obtain the circle centers of the feature areas.

[0028] Optionally, the step of performing group determination on the initial circles corresponding to the first circle centers to obtain the circle centers of the feature areas includes:

[0029] Determine whether there are two first circle centers with the same center coordinates within a preset bottom angle arc distance threshold range;

[0030] If so, the first circle center is the group center, and the two arc segments corresponding to the group center are used as the first arc segment and the second arc segment respectively; wherein the distance between the first arc segment and the group center is greater than the distance between the second arc segment and the group center;

[0031] Draw a ray from the center of the grouped circles to the first arc segment to determine a connected area;

[0032] Extracting a potential feature area image according to the center ray range and the connected area;

[0033] determining, based on the potential feature region image, whether the connected region is a rounded corner region;

[0034] If so, the group center is the center of the feature area.

[0035] Optionally, the step of determining whether the connected area is a rounded area based on the potential feature area image includes:

[0036] Performing first-order difference on the potential feature region image to obtain the contour line of the connected region;

[0037] Obtaining the area value and aspect ratio value of the minimum rectangle in the connected area according to the contour line;

[0038] Setting an area threshold and an aspect ratio threshold according to the connected area;

[0039] If the area value of the minimum area rectangle is smaller than the area threshold, and the aspect ratio value of the minimum area rectangle is smaller than the aspect ratio threshold, then the connected region is the rounded corner region.

[0040] Optionally, the step of identifying the rounded corner area and the bottom corner area of the contour image according to the center of the characteristic area includes:

[0041] From the center of the characteristic area to the centroid of the connected area, the characteristic area can be divided into a first area, a second area, a third area, a fourth area and a fifth area;

[0042] If the grayscale values of the first region, the third region, and the fifth region are 255, 255, 255, and the grayscale values of the second region and the fourth region are 0, 0, 0;

[0043] Then the third area is the bottom corner area, and the fifth area is the rounded corner area.

[0044] In addition, to achieve the above-mentioned purpose, the present application also provides an aircraft structural component corner feature recognition device, comprising:

[0045] A contour image acquisition module is used to acquire a contour image of the aircraft structural component to be programmed; wherein the contour image is obtained by performing a feature region closing process on a two-dimensional projection image of a three-dimensional model of the aircraft structural component to be programmed;

[0046] A corner point acquisition module, configured to perform cluster center processing on the extreme points of the contour image to obtain the corner points of the contour image;

[0047] A circle center acquisition module, configured to cut off the contour line of the contour image based on the corner point to obtain the circle center of the feature area;

[0048] a rounded corner and bottom corner region recognition module, configured to recognize the rounded corner region and bottom corner region of the contour image according to the center of the characteristic region;

[0049] The corner feature recognition result acquisition module is used to obtain the corner feature recognition result of the aircraft structural component to be programmed based on the fillet area and the bottom angle area.

[0050] In addition, to achieve the above-mentioned purpose, the present application also provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-mentioned method.

[0051] In addition, to achieve the above-mentioned purpose, the present application also provides a computer-readable storage medium, on which a computer program is stored, and a processor executes the computer program to implement the above-mentioned method.

[0052] The beneficial effects that can be achieved by this application.

[0053] The present application provides a method, apparatus, device, and storage medium for identifying corner features of aircraft structural parts. The method comprises: obtaining a contour image of an aircraft structural part to be programmed; wherein the contour image is a two-dimensional projection image of a three-dimensional model of the aircraft structural part to be programmed, obtained after a feature region closure process; performing cluster center processing on the extreme points of the contour image to obtain corner points of the contour image; truncating the contour line of the contour image based on the corner points to obtain the center point of the feature region; identifying the rounded corner region and bottom corner region of the contour image based on the center point of the feature region; and obtaining a corner feature recognition result of the aircraft structural part to be programmed based on the rounded corner region and bottom corner region. Specifically, the method performs a closure process on the two-dimensional contour image of the aircraft structural part at any viewing angle to eliminate interference from missing pixels, obtains a contour image of the structural part, truncates the contour line based on the corner points of the contour image to obtain multiple arcs, determines the center point of the feature region based on the distribution of the arcs, and then determines the rounded corner feature and bottom corner feature in the image. Finally, the corner feature is determined based on the distribution of the rounded corner region and bottom corner region. The method of the present application can realize corner feature recognition at any viewing angle without any restriction on the projection angle, that is, it does not require manual recognition, realizes the automation of corner feature recognition, and improves the efficiency and accuracy of recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A schematic diagram of the computer device structure of the hardware operating environment involved in the embodiments of the present application;

[0055] Figure 2 A schematic flow chart of a method for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0056] Figure 3 A schematic diagram of the functional modules of a device for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0057] Figure 4 A two-dimensional projection image of an aircraft structural part according to a method for identifying corner features of an aircraft structural part provided in an embodiment of the present application;

[0058] Figure 5 A schematic diagram of a non-enclosed area in a two-dimensional projection image of an aircraft structural component according to a method for identifying corner features of an aircraft structural component provided by an embodiment of the present application;

[0059] Figure 6 A schematic diagram of a non-closed contour in a two-dimensional projection image of an aircraft structural part according to a method for identifying corner features of an aircraft structural part provided by an embodiment of the present application;

[0060] Figure 7A partial schematic diagram of a low-grayscale region expansion result of a two-dimensional projection image of an aircraft structural part at an angle of 0° in a method for identifying corner features of an aircraft structural part provided by an embodiment of the present application;

[0061] Figure 8 A partial schematic diagram of a low-grayscale region expansion result at a 45° angle of a two-dimensional projection image of an aircraft structural part in a method for identifying corner features of an aircraft structural part provided by an embodiment of the present application;

[0062] Figure 9 A partial schematic diagram of a low-grayscale region expansion result at a 90° angle of a two-dimensional projection image of an aircraft structural part in a method for identifying corner features of an aircraft structural part provided by an embodiment of the present application;

[0063] Figure 10 A partial schematic diagram of a low-grayscale region expansion result of a two-dimensional projection image of an aircraft structural part at an angle of 135° in a method for identifying corner features of an aircraft structural part provided by an embodiment of the present application;

[0064] Figure 11 A partial schematic diagram of a closed contour image of a method for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0065] Figure 12 A partial schematic diagram of a contour image of a method for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0066] Figure 13 A schematic diagram of corner points of a contour image of a method for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0067] Figure 14 A line segment image of a method for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0068] Figure 15 A schematic diagram of a local arc segment and an initial circle center of a line segment image in a method for identifying corner features of aircraft structural parts provided by an embodiment of the present application;

[0069] Figure 16 A schematic diagram of grouped circle centers in a method for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0070] Figure 17 A schematic diagram of a potential feature area image of a method for identifying corner features of aircraft structural parts provided in an embodiment of the present application;

[0071] Figure 18 A schematic diagram of a fillet area and a bottom corner area in a method for identifying corner features of an aircraft structural component provided in an embodiment of the present application;

[0072] Figure 19 This is the corner feature recognition result of a corner feature recognition method for aircraft structural parts provided in an embodiment of the present application.

[0073] The realization of the objectives, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0074] It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application.

[0075] The main solution of the embodiments of the present application is: a method, device, equipment and storage medium for identifying corner features of aircraft structural parts are proposed, which obtains a contour image of the aircraft structural part to be programmed; wherein the contour image is a two-dimensional projection image of the three-dimensional model of the aircraft structural part to be programmed obtained after feature area closure processing; cluster center processing is performed on the extreme points of the contour image to obtain the corner points of the contour image; based on the corner points, the contour line of the contour image is truncated to obtain the center of the feature area; according to the center of the feature area, the rounded corner area and the bottom corner area of the contour image are identified; and according to the rounded corner area and the bottom corner area, the corner feature recognition result of the aircraft structural part to be programmed is obtained.

[0076] Conventional technology relies on CNC machine tools to manufacture aircraft structural components. Only when materials meet strength and stiffness requirements, high-precision machine tools, and well-designed CNC machining programs are used can these components be produced. However, due to the complexity of aircraft structural components, existing methods for determining the properties of spatial topological relationships in patches are unable to describe all possible scenarios. Programming relies heavily on manual experience, and as a result, CNC machining programs in the aviation industry have always been primarily developed manually.

[0077] Before programming a CNC machining program, it is necessary to identify and analyze the structural feature units in the three-dimensional model of the structural part. This is one of the most core tasks of programming, but the efficiency and accuracy of manual identification of corner features are low.

[0078] To this end, the present application provides a solution. By closing the two-dimensional contour image of an aircraft structural component at any viewing angle, the interference caused by missing pixels is eliminated, and the contour image of the structural component is obtained. The contour line is truncated at the corner points of the contour image to obtain multiple arc lines. The center of the feature area is determined based on the distribution of the arc lines, and the rounded corner feature and bottom corner feature in the image are further determined. The corner feature is finally determined based on the distribution of the rounded corner area and bottom corner area. The present application method can realize corner feature recognition at any viewing angle without any restrictions on the projection angle, that is, no manual recognition is required, and the corner feature recognition is automated, improving the efficiency and accuracy of recognition. At the same time, the bottom corner area is determined based on the double arc segments to determine the center of the same group of circles, and the rounded corner area is determined by the rectangle with the smallest area, thereby realizing accurate identification of the bottom corner area and the rounded corner area; further, the grayscale value change of the area connecting the center of the circle and the center of mass is used to determine the corner area characteristics, thereby further improving the recognition accuracy; further, the recognition results of the method of the present application can be subsequently mapped back to the three-dimensional model of the structural part to determine the data information of the specific area, and the generation of the structural part corner tool path data can be completed based on this information, thereby realizing the automated programming of the corner features.

[0079] Reference Figure 1 , Figure 1 This is a schematic diagram of the computer device structure of the hardware operating environment involved in the embodiment of the present application.

[0080] like Figure 1 As shown, the computer device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (WIreless-FIdelity, WI-FI) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) memory, or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk memory. The memory 1005 may optionally be a storage device independent of the aforementioned processor 1001.

[0081] Those skilled in the art will understand that Figure 1The structure shown in the figure does not constitute a limitation on the computer device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0082] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a data storage module, a network communication module, a user interface module and an electronic program.

[0083] exist Figure 1 In the computer device shown, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the computer device of the present invention can be set in the computer device, and the computer device calls the aircraft structural component corner feature recognition device stored in the memory 1005 through the processor 1001 and executes the aircraft structural component corner feature recognition method provided in the embodiment of the present application.

[0084] Reference Figure 2 Based on the hardware devices of the aforementioned embodiments, an embodiment of the present application provides a method for identifying corner features of aircraft structural parts, comprising:

[0085] S10: Acquire a contour image of the aircraft structural component to be programmed; wherein the contour image is obtained by performing a feature region closing process on a two-dimensional projection image of a three-dimensional model of the aircraft structural component to be programmed;

[0086] During implementation, the recognition of aircraft structural component corner features was performed on a 64-bit Windows 7 system using the OpenCV 4.5.2 image processing library, Release X64, SDK version 10.0.22000.0, an Intel(R) Xeon(R) W-2223 3.60GHz processor, 32GB of RAM at 3200MHz, and a 256GB hard drive. The software development platform was Visual Studio 2019, programming in C++. The corresponding aircraft structural component 3D model design software was CATIA V5.21. To ensure recognition accuracy, the acquired 2D projection images required specific display parameters: the grayscale of the contour lines was black (0, 0, 0), and the grayscale of other areas was white (255, 255, 255).

[0087] The image capture tool based on CATIA software obtains the two-dimensional projection image of the three-dimensional model of the aircraft structure under any viewing angle, such as Figure 4As shown in , the acquired image only contains contour information, and no other non-structural parts are captured. However, due to the difference in resolution, the contour lines of different feature areas may be discontinuous in the two-dimensional projection image at any viewing angle, such as Figure 5 As shown, these distinct feature regions should actually be closed. To improve subsequent recognition accuracy, these regions should be closed, forming closed feature segments. By closing these regions, we can obtain the contour image of the aircraft structural component to be programmed, improving feature recognition accuracy.

[0088] As an optional embodiment, before the step of obtaining the contour image of the aircraft structural component to be programmed, the method further includes: obtaining a two-dimensional projection image of the three-dimensional model of the aircraft structural component to be programmed; performing expansion processing on the low-grayscale region of the two-dimensional projection image at multiple angles using a 1*2 convolution kernel to obtain low-grayscale region expansion results at each angle; fusing the multiple low-grayscale region expansion results to obtain a closed contour image; and performing expansion processing on the high-grayscale region of the closed contour image at all angles using a 2*2 convolution kernel to obtain the contour image.

[0089] In the specific implementation process, a two-dimensional projection image of the three-dimensional model of the aircraft structure to be programmed is obtained, and convolution expansion processing is performed on it using the following relationship:

[0090]

[0091] Among them, i and j represent the row and column values of the corresponding pixels in the image, m and n represent the length and width of the convolution kernel, F(m,n) represents the convolution kernel function, and G(i,j) represents the image before convolution processing.

[0092] The low grayscale area is expanded based on the convolution kernel of 1*2 (1 pixel high, 2 pixels wide) at multiple angles to fill the contour line of the corresponding angle, and the results of the low grayscale area expansion at each angle are fused to obtain the following: Figure 6 The non-closed contour shown is closed, and the result is Figure 11 The closed contour image shown.

[0093] In the full angle range of 0°-360°, the high grayscale area is expanded based on the 2*2 (2 pixel high, 2 pixel wide) convolution kernel, and the following is obtained: Figure 12 The contour image shown restores the shape of the feature unit to the width before processing within the full angle range, ensuring the processing refinement and the step size of the convolution kernel movement is also 1, which is convenient for subsequent feature recognition.

[0094] As an optional implementation, the step of using a 1*2 convolution kernel to expand the low grayscale area of the two-dimensional projection image at multiple angles to obtain the low grayscale area expansion results at each angle includes: using a 1*2 convolution kernel to expand the low grayscale area of the two-dimensional projection image at angles of 0°, 45°, 90° and 135° to obtain the low grayscale area expansion results at four angles.

[0095] In the specific implementation process, the low grayscale area of the two-dimensional projection image is expanded at multiple angles to close the contour area. This embodiment selects four angles of 0°, 45°, 90° and 135° to process and obtain the following Figures 7 to 10 The expansion results shown can achieve expansion in eight directions: 0°, 360°, 45°, 315°, 90°, 270°, 135°, and 225°.

[0096] At an angle of 0°, a 1*2 convolution kernel is used to expand the low-grayscale area, so that the potential feature areas in the 0° and 360° directions are closed. The specific steps are: use a 1-unit pixel distance to move the convolution kernel, and use the left point as the convolution kernel anchor point. When the grayscale value of the point on the right is 0, 0, 0, the grayscale value of the anchor point is transformed to 0, 0, 0, otherwise it remains unchanged until all pixel positions in the image are processed; at an angle of 45°, a 1*2 convolution kernel is used to expand the low-grayscale area, so that the potential feature areas in the 45°315° directions are closed. The specific steps are: use a 1-unit pixel distance to move the convolution kernel, and use the point corresponding to the upper right corner as the anchor point. When the grayscale value of the pixel point at the corresponding position on the lower left is 0, 0, 0, the grayscale value of the pixel point at the anchor point is transformed to 0, 0, 0, otherwise it remains unchanged; at a 90° angle, a 1*2 convolution kernel is used to expand the low-grayscale area, so that the potential feature areas in the 45°315° directions are closed. At the angle, a 1*2 convolution kernel is used to expand the low grayscale area, so that the potential feature areas in the 90° and 270° directions are closed. The specific steps are: the convolution kernel moving step is 1 unit pixel for processing, and the point corresponding to the upper side position is used as the anchor point. When the grayscale value of the pixel at the corresponding position on the lower side is 0, 0, 0, the grayscale value of the pixel at the anchor point is transformed to 0, 0, 0, otherwise it remains unchanged; at the 135° angle, a 1*2 convolution kernel is used to expand the low grayscale area, so that the potential feature areas in the 135° and 225° directions are closed. The specific steps are: the convolution kernel moving step is 1 unit pixel for processing, and the point corresponding to the upper left position is used as the anchor point. When the grayscale value of the pixel at the corresponding position on the lower right side is 0, 0, 0, the grayscale value of the pixel at the anchor point is transformed to 0, 0, 0, otherwise it remains unchanged.

[0097] S20: performing cluster center processing on the extreme points of the contour image to obtain corner points of the contour image;

[0098] In the specific implementation process, Figure 13 As shown, a corner point refers to a point with particularly prominent attributes in certain aspects, an isolated point with the maximum or minimum intensity in certain attributes, or the end point of a line segment. A corner point can effectively reduce the amount of information while retaining important features of the image graphics, making its information content very high, effectively improving the speed of calculation, facilitating reliable image matching, and improving the accuracy and efficiency of image processing. Regardless of the angle of the structural parts in the two-dimensional projection image, the corner point will have a large information response in filters of different scales, and has high recognition accuracy and reliability. This embodiment recognizes corner features based on the corner points of the contour image, is not affected by the viewing angle, and can also ensure the accuracy of the initial basic data.

[0099] As an optional implementation, the step of performing cluster center processing on the extreme points of the contour image to obtain the corner points of the contour image includes: performing filtering processing on the contour image to obtain multiple filtered images; performing differential processing on the filtered images to obtain multiple extreme points; and performing cluster center processing on the extreme points to obtain the corner points of the contour image.

[0100] In the specific implementation process, the acquisition of image corner points needs to be based on three single passes, so the contour image needs to be grayscaled first; when the following relationship is used, the image is filtered based on different weight differences:

[0101] F(x,y,σ,p,n)=filter(x,y,σ,p,n)*h(x,y)

[0102]

[0103] Where x and y represent the pixel column and row values in the image, respectively, σ represents the standard deviation, p represents the constant coefficient, n represents the constant coefficient raised to the power of n, and the value of n can be 4. h(x,y) is the grayscale image of the contour image. According to the filtering formula, filtered images can be obtained after filtering at different filtering values, and the number of filtered images corresponds to the constant coefficient raised to the power of n. In this embodiment, the original grayscale image h(x,y) is downsampled to 1 / 4, 1 / 8, and 1 / 16 of its original size, respectively. Filtering is performed on these four scales (the original size image and the three downsampled images), resulting in 4 filtered images for each scale, totaling 16 filtered images.

[0104] The grayscales of the corresponding positions of the filtered images of the same scale are subtracted to obtain differential images of the filtered differences of the same scale, and the extreme points with larger responses are located in the differential images. In this embodiment, based on the pixel points that are not on the boundary and have an 8-neighborhood position relationship, the grayscale values of the 9 pixels above and below the corresponding position areas of the 8 neighborhoods of the plane image and the different filtered images of the same scale are compared in turn, and a total of 26 (8+9+9) pixel positions are compared. After the processing is completed, all images are upsampled to the same scale (the size of the original grayscale image). If the grayscale value of a certain pixel point is the largest at the corresponding position in the two images with different filter coefficients, the corresponding point is the extreme point, otherwise it is not an extreme point.

[0105] Because there are multiple high-frequency areas near the intersection of line segments (i.e., corner points), the extreme points determined in the above steps may only be points near the corner points, rather than the corner points themselves. In order to improve the accuracy of corner point recognition, the extreme points need to be clustered. The center point of the cluster is the corner point in the final contour image. Figure 13 As shown. The cluster center is processed through the following relationship:

[0106]

[0107] Among them, center represents the cluster center of the extreme points, S5*5 means that the statistical extreme value is an image area of 5*5 size. Because the width value of the contour in the acquired contour image is between 1 and 2, the width of 5 can surround the contour, so this embodiment selects this range, and num represents the number of extreme points in the 5*5 area.

[0108] S30: Based on the corner point, cut off the contour line of the contour image to obtain the center of the feature area;

[0109] In the specific implementation process, Figure 14 As shown, a circle is drawn with the corner point as the center, the contour line of the contour image is truncated, and the center of the feature area is determined according to the distribution characteristics of the truncated contour line. Subsequently, the feature area can be determined according to the center of the feature area to identify the corner feature.

[0110] As an optional embodiment, the step of truncating the contour line of the contour image based on the corner point to obtain the center of the feature area includes: drawing a reference circle with a radius of a with the corner point as the center; wherein a is a preset value; based on the reference circle, truncating the contour line of the contour image to obtain a line segment image; wherein the line segment image contains multiple arc segments; performing whole circle fitting on the arc segments to obtain multiple initial circles; obtaining multiple first circle centers based on the radius of the initial circle, the center coordinates of the initial circle and a preset radius range; performing group judgment on the initial circles corresponding to the first circle centers to obtain the center of the feature area.

[0111] In the specific implementation process, this embodiment uses each corner point as the center of the circle and a radius of 6 pixels to draw a reference circle, and cuts off the part where the contour image intersects with the reference circle, so that the original contour line is cut into multiple straight line segments and arc segments, and the following is obtained: Figure 14 The line segment image shown in FIG is used to fit a circle with the arc segment in the line segment image to obtain multiple initial circles. The equation of the initial circle can be obtained according to the coordinates of each point on the arc segment, as shown in FIG. Figure 15 The center and radius of the circle are shown. Based on the aircraft structural component design, a radius range for potential feature areas is set. Initial circles with radii within this range are retained, and their centers are designated as the first circle centers. A determination is made as to whether the arcs on the initial circles corresponding to each first circle center are true feature area segments. If so, the first circle center is designated as the center of the feature area.

[0112] As an optional embodiment, the step of performing grouping judgment on the initial circles corresponding to the first circle center and obtaining the circle center of the feature area includes: judging whether there are two first circle centers with the same circle center coordinates within a preset bottom angle arc distance threshold range; if so, the first circle center is the grouped circle center, and the two arc segments corresponding to the grouped circle center are respectively used as the first arc segment and the second arc segment; wherein, the distance between the first arc segment and the grouped circle center is greater than the distance between the second arc segment and the grouped circle center; based on the grouped circle center, a center ray is drawn to the first arc segment to determine a connected area; based on the center ray range and the connected area, a potential feature area image is extracted; based on the potential feature area image, whether the connected area is a rounded corner area is judged; if so, the grouped circle center is the circle center of the feature area.

[0113] In the specific implementation process, according to the arc radius threshold of the aircraft structure to be programmed r Set the bottom corner arc distance threshold range, such as Figure 15 As shown, if two first circle centers are both within the distance threshold range and their coordinates are the same, then the two first circle centers can be determined to be grouped circle centers, that is, the initial circles corresponding to the two first circle centers are concentric circles, and the corresponding two arc segments are potential feature area segments. In theory, the coordinates of the two first circle centers of a grouped circle center are the same (that is, the same point), but due to the thickness of the contour line, image resolution, etc., in actual operation, the coordinates of the two first circle centers that should be grouped circle centers will deviate. The midpoint of the line connecting the two can be found as the grouped circle center. Figure 16As shown, the grouped circle centers correspond to arc segments of two initial circles. The arc segment farther from the grouped circle center is the first arc segment (corresponding to a larger initial circle radius), and the arc segment farther from the grouped circle center is the second arc segment (corresponding to a smaller initial circle radius). Determining the grouped circle centers enables preliminary screening of multiple first circle centers and preliminary judgment of the base angle characteristics (the circles corresponding to the upper and lower arc segments of the base angle area are concentric circles).

[0114] Based on the center of the grouped circle, a ray is drawn from the center of the first arc segment. The area where the first pixel with a grayscale value of 255, 255, 255 is located outside the first arc segment is the research and analysis object area. In this area, this point is used as the seed point to search for connected areas. The search method is: the search is performed in a loop iterative manner with the same grayscale of the four neighborhoods of the initial seed point. That is, if it is determined that there is a point with a similar grayscale value in a certain neighborhood of the initial seed point at the last time, the search is completed in the lower, left, upper, and right directions of the initial seed point. The seed point centered on the neighborhood point is searched until the search is completed. The search returns to the second-to-last connected pixel in the neighborhood for search, and the search is reversed in sequence until the point where the grayscale relationship is satisfied in the first neighborhood of the initial seed point.

[0115] All the grouped circle centers in the image are used to determine the corresponding connected areas according to the above method. The connected areas are potential rounded corner areas, such as Figure 17 As shown, potential feature regions are delineated based on the range of the center rays of a single grouped circle center and the corresponding connected area, and each potential feature region image is extracted. This means that each image contains only one feature region, facilitating subsequent feature recognition. Within this region, a determination is made as to whether the connected region is a true rounded corner region. If so, the potential feature region is a true corner feature region, and the corresponding grouped circle center is the center of the true feature region.

[0116] As an optional embodiment, the step of determining whether the connected area is a rounded corner area based on the potential feature area image includes: performing a first-order difference on the potential feature area image to obtain the contour line of the connected area; obtaining the area value and aspect ratio value of the minimum rectangle in the connected area based on the contour line; setting an area threshold and an aspect ratio threshold based on the connected area; if the area value of the minimum rectangle is less than the area threshold, and the aspect ratio value of the minimum rectangle is less than the aspect ratio threshold, then the connected area is the rounded corner area.

[0117] In the specific implementation process, the potential feature area image is processed through first-order difference to obtain multiple gradient value points. Based on the plane distance minimum criterion, all the gradient value points are connected end to end to obtain the connected area contour line. That is, this method reconnects the contour line previously truncated by the reference circle to form a complete contour line. Since an image has only one feature area, there is only one connected area contour line in the image, and the accuracy of the connected contour line operation is high. The first-order difference is performed using the following relationship:

[0118]

[0119] The directions of the first-order differences in this embodiment are the same as the angles of the above-mentioned contour closing processing for the two-dimensional projection image, namely, the eight directions of 0°, 360°, 45°, 315°, 90°, 270°, 135°, and 225°. The grad in the relationship represents the gradient. represents partial derivative; and Represents the partial derivatives in the X-axis and Y-axis directions respectively. For all pixels in the image, the first-order difference values in 8 directions are compared, and the pixel with the largest difference value is taken as the gradient value point.

[0120] After connecting the contour lines of the connected region, determine whether it is a truly rounded region: determine the minimum rectangle in the connected region and calculate its area and aspect ratio. The minimum rectangle feature is determined as follows: In the Dicker coordinate system of the two-dimensional contour image, the geometric center of the outer contour is the rotation center angle α. The contour line is mapped to countless projection points on the X and Y axes through the mapping function f(*). On each coordinate axis, there is only one extreme point (xmin, xmax, ymin, ymax). Based on these two sets of points, a rectangle with area Sn is obtained. The rotation angle is changed by increasing it by 1°. The angle and area corresponding to the minimum area value can be found in the range [0°, 90°].

[0121] The area threshold and aspect ratio threshold are set according to the characteristics of the standard rounded corner and the minimum area rectangle in its rounded corner area, as well as the actual size of the connected area. If and only if the area value of the minimum area rectangle is less than the area threshold and the aspect ratio value of the minimum area rectangle is less than the aspect ratio threshold, the connected area is considered to be a true rounded corner area, and the corresponding grouped circle center is the center of the feature area.

[0122] S40: Identifying the rounded corner area and the bottom corner area of the contour image according to the center of the characteristic area;

[0123] In the specific implementation process, the corner features include rounded corners and bottom corners. The above steps have determined the rounded corner area and bottom corner area through concentric circles and minimum area rectangles, but only the distribution mode in which the rounded corners are connected with the bottom corners can be determined as corners. Therefore, according to the center of the feature area, the rounded corner area and bottom corner area in the contour image are identified and determined again to improve the recognition accuracy.

[0124] As an optional embodiment, the step of identifying the rounded corner area and the bottom corner area of the contour image based on the center of the feature area includes: dividing the feature area into a first area, a second area, a third area, a fourth area and a fifth area from the center of the feature area to the centroid of the connected area; if the grayscale values of the first area, the third area and the fifth area are 255, 255, 255, and the grayscale values of the second area and the fourth area are 0, 0, 0; then the third area is the bottom corner area, and the fifth area is the rounded corner area.

[0125] In the specific implementation process, a directed line segment is drawn from the center of the feature area to the centroid of the connected area. If the grayscale of the area passed by the directed line segment satisfies the "white-black-white-black-white" change process, such as Figure 18 As shown, the second grayscale white area (the third area) is the bottom corner area, and the third grayscale white area (the fifth area) is the rounded corner area.

[0126] S50: Obtaining a corner feature recognition result of the aircraft structural component to be programmed based on the fillet area and the bottom corner area.

[0127] In the specific implementation process, Figure 19 As shown, the fillet area and the bottom corner area together constitute the corner area. By traversing all first circle centers using the above method, all corner features in the two-dimensional projection image of the aircraft structure to be programmed can be identified.

[0128] It should be understood that the above is only an example and does not constitute any limitation to the technical solution of the present application. Those skilled in the art can make settings based on needs in actual applications and no limitation is made here.

[0129] From the above description, it is not difficult to see that this embodiment obtains the contour image of the structural part by closing the two-dimensional contour image of the aircraft structural part at any viewing angle, eliminating the interference of missing pixels, and then truncating the contour line at the corner points of the contour image to obtain multiple arc lines. The center of the feature area is determined based on the distribution of the arc lines, and then the rounded corner feature and bottom corner feature in the image are determined. Finally, the corner feature is determined based on the distribution of the rounded corner area and bottom corner area. The method of the present application can realize corner feature recognition at any viewing angle without any restrictions on the projection angle, that is, no manual recognition is required, and the corner feature recognition is automated, improving the efficiency and accuracy of recognition. At the same time, the bottom corner area is determined based on the double arc segments to determine the center of the same group of circles, and the rounded corner area is determined by the rectangle with the smallest area, thereby realizing accurate identification of the bottom corner area and the rounded corner area; further, the grayscale value change of the area connecting the center of the circle and the center of mass is used to determine the corner area characteristics, thereby further improving the recognition accuracy; further, the recognition results of the method of the present application can be subsequently mapped back to the three-dimensional model of the structural part to determine the data information of the specific area, and the generation of the structural part corner tool path data can be completed based on this information, thereby realizing the automated programming of the corner features.

[0130] Reference Figure 3 Based on the same inventive concept, an embodiment of the present application further provides a device for identifying corner features of aircraft structural parts, comprising:

[0131] A contour image acquisition module is used to acquire a contour image of the aircraft structural component to be programmed; wherein the contour image is obtained by performing a feature region closing process on a two-dimensional projection image of a three-dimensional model of the aircraft structural component to be programmed;

[0132] A corner point acquisition module, configured to perform cluster center processing on the extreme points of the contour image to obtain the corner points of the contour image;

[0133] A circle center acquisition module, configured to cut off the contour line of the contour image based on the corner point to obtain the circle center of the feature area;

[0134] a rounded corner and bottom corner region recognition module, configured to recognize the rounded corner region and bottom corner region of the contour image according to the center of the characteristic region;

[0135] The corner feature recognition result acquisition module is used to obtain the corner feature recognition result of the aircraft structural component to be programmed based on the fillet area and the bottom angle area.

[0136] It should be noted that the modules in the aircraft structural component corner feature recognition device in this embodiment correspond one-to-one to the steps in the aircraft structural component corner feature recognition method in the aforementioned embodiment. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned aircraft structural component corner feature recognition method, and will not be repeated here.

[0137] In addition, in one embodiment, an embodiment of the present application further provides a computer device, which includes a processor, a memory, and a computer program stored in the memory, and the computer program implements the steps of the method in the aforementioned embodiment when executed by the processor.

[0138] In addition, in one embodiment, an embodiment of the present application further provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method in the aforementioned embodiment are implemented.

[0139] In some embodiments, the computer-readable storage medium may be a memory device such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface mount memory, optical disk, or CD-ROM; or various devices including any one or any combination of the above memories. The computer may be various computing devices including smart terminals and servers.

[0140] In some embodiments, executable instructions may be in the form of a program, software, software module, script, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0141] As an example, executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinating files (e.g., files storing one or more modules, subroutines, or code portions).

[0142] By way of example, executable instructions may be deployed to be executed on one computing device, or on multiple computing devices at one site, or on multiple computing devices distributed across multiple sites and interconnected by a communication network.

[0143] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or system comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or system. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or system comprising the element.

[0144] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0145] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a multimedia terminal device (which can be a mobile phone, a computer, a television receiver, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0146] The above are only preferred embodiments of the present application and do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for identifying corner features of aircraft structural parts, characterized in that: The following steps are involved: Acquire a contour image of the aircraft structural component to be programmed; wherein the contour image is obtained by performing feature region closing processing on a two-dimensional projection image of a three-dimensional model of the aircraft structural component to be programmed; Performing cluster center processing on the extreme points of the contour image to obtain corner points of the contour image; Based on the corner points, the contour line of the contour image is cut off to obtain the center of the feature area; Identifying the rounded corner area and the bottom corner area of the contour image according to the center of the characteristic area; Obtaining a corner feature recognition result of the aircraft structural component to be programmed based on the fillet area and the bottom angle area; The step of truncating the contour line of the contour image based on the corner point to obtain the center of the feature area includes: Draw a reference circle with a radius of a with the corner point as the center; where a is a preset value; Based on the reference circle, the contour line of the contour image is cut off to obtain a line segment image; wherein the line segment image contains a plurality of circular arc segments; Performing full circle fitting on the arc segments to obtain multiple initial circles; Obtaining multiple first circle centers according to the radius of the initial circle, the coordinates of the center of the initial circle and a preset radius range; The initial circles corresponding to the first circle centers are grouped to obtain the circle centers of the feature areas.

2. The method for identifying corner features of aircraft structural parts according to claim 1, wherein: Before the step of obtaining the contour image of the aircraft structural component to be programmed, the method further includes: Acquiring a two-dimensional projection image of the three-dimensional model of the aircraft structural component to be programmed; Performing expansion processing on the low-grayscale area of the two-dimensional projection image at multiple angles using a 1*2 convolution kernel to obtain expansion results of the low-grayscale area at each angle; Fusing the multiple low-grayscale area expansion results to obtain a closed contour image; At all angles, a 2*2 convolution kernel is used to expand the high grayscale area of the closed contour image to obtain the contour image.

3. The method for identifying corner features of aircraft structural parts according to claim 2, wherein: The step of performing expansion processing on the low-grayscale area of the two-dimensional projection image at multiple angles using a 1*2 convolution kernel to obtain the low-grayscale area expansion results at each angle includes: The low-grayscale area of the two-dimensional projection image is expanded using a 1*2 convolution kernel at angles of 0°, 45°, 90°, and 135°, respectively, to obtain low-grayscale area expansion results at four angles.

4. The method for identifying corner features of aircraft structural parts according to claim 1, wherein: The step of performing cluster center processing on the extreme points of the contour image to obtain the corner points of the contour image includes: Performing filtering on the contour image to obtain a plurality of filtered images; performing differential processing on the filtered image to obtain a plurality of extreme points; Perform cluster center processing on the extreme points to obtain corner points of the contour image.

5. The method for identifying corner features of aircraft structural parts according to claim 4, wherein: The step of performing group determination on the initial circles corresponding to the first circle centers to obtain the circle centers of the feature areas includes: Determine whether there are two first circle centers with the same center coordinates within a preset bottom angle arc distance threshold range; If so, the first circle center is the group center, and the two arc segments corresponding to the group center are used as the first arc segment and the second arc segment respectively; wherein the distance between the first arc segment and the group center is greater than the distance between the second arc segment and the group center; Draw a ray from the center of the grouped circles to the first arc segment to determine a connected area; Extracting a potential feature area image according to the center ray range and the connected area; determining, based on the potential feature region image, whether the connected region is a rounded corner region; If so, the group center is the center of the feature area.

6. The method for identifying corner features of aircraft structural parts according to claim 5, wherein: The step of determining whether the connected area is a rounded area based on the potential feature area image includes: Performing first-order difference on the potential feature region image to obtain the contour line of the connected region; Obtaining the area value and aspect ratio value of the minimum rectangle in the connected area according to the contour line; Setting an area threshold and an aspect ratio threshold according to the connected area; If the area value of the minimum area rectangle is smaller than the area threshold, and the aspect ratio value of the minimum area rectangle is smaller than the aspect ratio threshold, then the connected region is the rounded corner region.

7. The method for identifying corner features of aircraft structural parts according to claim 1, wherein: The step of identifying the rounded corner area and the bottom corner area of the contour image according to the center of the characteristic area includes: From the center of the characteristic area to the centroid of the connected area, the characteristic area can be divided into a first area, a second area, a third area, a fourth area and a fifth area; If the grayscale values of the first region, the third region, and the fifth region are 255, 255, 255, and the grayscale values of the second region and the fourth region are 0, 0, 0; Then the third area is the bottom corner area, and the fifth area is the rounded corner area.

8. An aircraft structural component corner feature recognition device, characterized in that: include: A contour image acquisition module is used to acquire a contour image of the aircraft structural component to be programmed; wherein the contour image is obtained by performing a feature region closing process on a two-dimensional projection image of a three-dimensional model of the aircraft structural component to be programmed; A corner point acquisition module, configured to perform cluster center processing on the extreme points of the contour image to obtain the corner points of the contour image; A circle center acquisition module, configured to cut off the contour line of the contour image based on the corner point to obtain the circle center of the feature area; a rounded corner and bottom corner region recognition module, configured to recognize the rounded corner region and bottom corner region of the contour image according to the center of the characteristic region; a corner feature recognition result acquisition module, configured to obtain a corner feature recognition result of the aircraft structural component to be programmed based on the fillet area and the bottom angle area; The step of truncating the contour line of the contour image based on the corner point to obtain the center of the feature area includes: Draw a reference circle with a radius of a with the corner point as the center; where a is a preset value; Based on the reference circle, the contour line of the contour image is cut off to obtain a line segment image; wherein the line segment image contains a plurality of circular arc segments; Performing full circle fitting on the arc segments to obtain multiple initial circles; Obtaining multiple first circle centers according to the radius of the initial circle, the coordinates of the center of the initial circle and a preset radius range; The initial circles corresponding to the first circle centers are grouped to obtain the circle centers of the feature areas.

9. A computer device, characterized in that: The computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and the processor executes the computer program to implement the method according to any one of claims 1 to 7.

Images