MIM gear shaft feature recognition method and device based on Hu invariant moment

Abstract

The invention provides an MIM gear shaft feature recognition method and device based on Hu invariant moment, and the method comprises the steps: obtaining a digital model of a to-be-detected gear, and enabling the digital model to comprise a main tooth profile surface; on the basis of feature information of the main tooth-shaped surface, an identification coordinate system is constructed, and the feature information comprises the normal direction of the main tooth-shaped surface, the midpoint of a tooth-shaped area and the circle center; according to the identification coordinate system and a preset cross section, an inclusion kernel plane set is intercepted, and the inclusion kernel plane set comprises a plurality of inclusion kernel images; on the basis of a scaled Hu invariant moment algorithm, acquiring a scaled Hu invariant moment of the containing nuclear image; and comparing the scaling Hu invariant moments of all the inclusion kernel images in the inclusion kernel plane set, and carrying out inclusion kernel feature identification. According to the invention, the robustness of a gear shaft feature recognition result is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a method and device for MIM gear shaft feature recognition based on Hu invariant moments. Background Art

[0002] With the rapid development of computer software and hardware technology, it has been deeply applied in mechanical manufacturing enterprises, and enterprises have a wide and urgent practical demand for computer-aided process design. In this context, the mechanical design and manufacturing industry can make up for the defects of traditional mechanical design and manufacturing with the help of computer-aided technology, and the market competitiveness of enterprises can be greatly improved.

[0003] The Chinese patent with publication number CN111652925B discloses a method for extracting the global feature Hu invariant moment of a target using single pixel imaging, by constructing a (p+q)-order two-dimensional function G(x,y) such that the two-dimensional function satisfies: G pq (x, y) = x p y q , and then adjust the parameter values ​​of p,q to generate the corresponding two-dimensional function G pq (x, y) modulation information; the generated two-dimensional modulation information is sent to the light modulation system, the illumination light of the target object is modulated to generate structured modulated light, and the modulated light is used to illuminate the target, and finally the intensity value of the target transmission or reflection light is obtained by using the single pixel detection and acquisition system, and the intensity value obtained by detection is used to invert the Hu invariant moment of the global feature of the target. However, the above method is based on the two-dimensional function G pq The Hu invariant moment of (x, y) cannot maintain good rotation, scaling and translation invariance when converting three dimensions to two dimensions, resulting in the robustness of the output results failing to meet the requirements. Therefore, it is very necessary to provide a MIM gear shaft feature recognition method and device based on Hu invariant moment to improve the robustness of the gear shaft feature recognition results. Summary of the invention

[0004] In view of this, the present invention proposes a MIM gear shaft feature recognition method and device based on Hu invariant moment, which obtains the scaled Hu invariant moment of each inclusive core image in the inclusive core plane set by intercepting the inclusive core plane set of the digital model of the gear to be tested in the constructed recognition coordinates, thereby improving the robustness of the gear shaft feature recognition result.

[0005] The present invention provides a method and device for MIM gear shaft feature recognition based on Hu invariant moment, the method comprising:

[0006] Acquire a digital model of the gear to be tested, wherein the digital model includes a main tooth profile surface;

[0007] Based on the characteristic information of the main tooth profile surface, an identification coordinate system is constructed, wherein the characteristic information includes the normal direction of the main tooth profile surface, the midpoint of the tooth profile area, and the center of a circle;

[0008] According to the identified coordinate system and the preset cross section, intercepting an inclusive core plane set, wherein the inclusive core plane set includes a plurality of inclusive core images;

[0009] Based on the scaled Hu invariant moment algorithm, obtaining the scaled Hu invariant moment of the inclusion kernel image;

[0010] The scaled Hu invariant moments of all the inclusion kernel images in the inclusion kernel plane set are compared, and inclusion kernel feature recognition is performed.

[0011] On the basis of the above technical solution, preferably, the method for obtaining the main tooth profile surface includes:

[0012] Acquire the surface data of the gear to be tested, wherein the surface data includes the gear surface and the number of gear surfaces connected to the gear surface;

[0013] Determining whether the number of connected gear surfaces of the gear surface is equal to the number of preset surface shapes;

[0014] If the number of the connected gear surfaces is equal to the preset number of surface profiles, the gear surface is selected as the main tooth profile surface.

[0015] On the basis of the above technical solution, preferably, the construction of the identification coordinate system based on the feature information of the main tooth profile surface specifically includes:

[0016] The normal direction of the main tooth profile is used as the Z axis, and the line between the central axis of the gear to be measured and the rotation central axis of the gear shaft to be measured is used as the X axis to construct the identification coordinate system, wherein the gear to be measured includes the gear shaft to be measured.

[0017] More preferably, the scaled Hu invariant moment algorithm based on the scaled Hu invariant moment sequentially calculates the scaled Hu invariant moment in the inclusive kernel image, specifically comprising:

[0018] Acquire a triangular mesh of the main tooth profile surface, wherein the triangular mesh includes a plurality of triangular node coordinates;

[0019] Performing region segmentation on the inclusion core image so that the inclusion core image forms M×N image sub-blocks;

[0020] Based on the identification coordinate system and the image sub-block, obtaining the plane node coordinates of the image sub-block;

[0021] Acquire a mapping relationship between the triangular mesh and the inclusion kernel image according to the triangular node coordinates and the plane node coordinates, wherein the mapping relationship includes a scaling factor;

[0022] Based on the scaling factors, the scaled Hu invariant moments of the inclusion kernel image are calculated in sequence.

[0023] More preferably, the step of sequentially calculating the scaled Hu invariant moments of the inclusive kernel image based on the scaling factor specifically includes:

[0024] Obtaining the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image;

[0025] The p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image is scaled by a scaling factor to obtain an i order scaled Hu invariant moment.

[0026] More preferably, the obtaining of the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image specifically includes:

[0027]

[0028] Among them, m pq represents the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image, where p and q are both natural numbers and the 0-order origin moment m 00 represents the total volume of the main tooth surface in the triangular mesh, f(x, y) represents the two-dimensional distribution function of the inclusion kernel image, (x, y) represents the two-dimensional coordinates of any point in the inclusion kernel image, u pq represents the geometric center moment of the inclusion kernel image, and Respectively represent the horizontal and vertical coordinates of the centroid of the main tooth profile surface.

[0029] More preferably, scaling the p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image by a scaling factor to obtain an i order scaled Hu invariant moment specifically includes:

[0030]

[0031] Among them, η pq represents the normalized central moment, γ=(p+q) / 2+1, M1 to M7 are scaled Hu invariant moments of order 1 to 7, respectively. represents the scaling factor, μ 00 Represents the zero-order origin moment m 00 The corresponding 0th-order normalized central moment.

[0032] In a second aspect of the present application, a MIM gear shaft feature recognition device based on Hu invariant moment is provided, wherein the feature recognition device comprises a collection module, a processing module and an analysis module, wherein:

[0033] The acquisition module is used to obtain a digital model of the gear to be tested, wherein the digital model includes a main tooth profile surface;

[0034] The processing module is used to construct an identification coordinate system based on the feature information of the main tooth profile surface, wherein the feature information includes the normal direction of the main tooth profile surface, the midpoint of the tooth profile area and the center of the circle, and to intercept an inclusive core plane set according to the identification coordinate system and a preset cross section, wherein the inclusive core plane set includes a plurality of inclusive core images, and to obtain a scaled Hu invariant moment of the inclusive core image based on a scaled Hu invariant moment algorithm;

[0035] The analysis module is used to compare the scaled Hu invariant moments of all the inclusion kernel images in the inclusion kernel plane set and perform inclusion kernel feature recognition.

[0036] In a third aspect of the present application, an electronic device is provided, comprising a processor, a memory, a user interface and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory.

[0037] In a fourth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, and the computer program is executed by a processor to implement the steps of a MIM gear shaft feature recognition method based on Hu invariant moments.

[0038] The MIM gear shaft feature recognition method and device based on Hu invariant moment provided by the present invention has the following beneficial effects compared with the prior art:

[0039] (1) The inclusive core plane set of the digital model of the gear to be tested is intercepted in the constructed recognition coordinates to obtain each inclusive core image in the inclusive core plane set. The inclusive core image is divided into blocks by sliding windows and the scaled Hu invariant moment of the divided blocks is calculated. Finally, the scaled Hu invariant moment of the inclusive core image is used for inclusive core feature recognition. By selecting image features with translation, rotation and scale invariance, the effectiveness and stability of feature extraction of image sub-blocks in the inclusive core plane set are guaranteed, thereby improving the robustness of the gear shaft feature recognition results.

[0040] (2) A mapping relationship is formed between triangular meshes and image sub-blocks in a two-dimensional plane, which qualitatively improves the connectivity of each image and significantly improves the accuracy of feature detection. At the same time, in order to avoid incomplete image information caused by the inconsistent rotation angle between the collected inclusion kernel plane set and the image sub-block, the inclusion kernel image in the inclusion kernel plane set is divided into image sub-blocks one by one to improve the stability of feature extraction and ensure the matching accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 A schematic flow chart of a method for identifying MIM gear shaft features based on Hu invariant moments provided by the present invention;

[0043] Figure 2 A schematic diagram of the distance between the main tooth profile surface and the inclusion core in the current coordinate system provided by the present invention;

[0044] Figure 3 A schematic diagram of a homogenized normal vector provided by the present invention;

[0045] Figure 4 A schematic diagram of a grid after the inclusion kernel image provided by the present invention is segmented into regions;

[0046] Figure 5 A schematic diagram of the structure of the identification device provided by the present invention;

[0047] Figure 6 This is a schematic structural diagram of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] The present application embodiment discloses a method for identifying a characteristic model of an injection molded gear, such as Figure 1 As shown, the method comprises steps S1-S5.

[0050] Step S1, obtaining a digital model of the gear to be tested, wherein the digital model includes a main tooth profile surface.

[0051] In this step, the digital model of the gear to be tested is traversed to obtain the tooth profile of the gear to be tested, and the identification coordinate system is determined according to the selected tooth profile.

[0052] Step S1 includes steps S11-S13, wherein:

[0053] Step S11, obtaining the surface data of the gear to be tested, wherein the surface data includes the gear surface and the number of gear surfaces connected to the gear surface.

[0054] In this step, the surface data is obtained by traversing the digital model of the gear to be tested.

[0055] Step S12, determining whether the number of adjacent gear surfaces of the gear surface is equal to the number of preset surface types.

[0056] In this step, the preset number of surface shapes is the maximum number of adjacent surfaces in the gear to be tested.

[0057] Step S13, if the number of connected gear surfaces is equal to the preset number of surface profiles, then the gear surface is selected as the main tooth profile surface.

[0058] In this step, the number of adjacent surfaces of each gear surface of the gear to be tested is calculated, and the two gear surfaces with the largest number of adjacent surfaces are determined as tooth profile surfaces. One gear surface is selected as the main tooth profile surface, and the other tooth profile surface is selected as the auxiliary tooth profile surface, and the main tooth profile surface is used as the identification coordinate system (the Z-axis coordinates of the auxiliary tooth profile surface are all negative numbers at this time). After the tooth profile surface is determined, the contour line of the tooth profile surface is extracted.

[0059] In one example, if Figure 2 As shown, in the current coordinate system, the distance h1 from the main tooth profile surface to the top of the containing core and the distance h2 from the auxiliary tooth profile surface to the bottom of the containing core are calculated. If h2>h1, a new coordinate system newCSYS is created. This coordinate system is opposite to the current coordinate system. The original coordinate system is hidden and the marks of the main tooth profile surface and the auxiliary tooth profile surface are swapped. The distance between the two surfaces is calculated, which is the parameter tooth thickness b.

[0060] Step S2, constructing an identification coordinate system based on the feature information of the main tooth profile surface, wherein the feature information includes the normal direction of the main tooth profile surface, the midpoint of the tooth profile area, and the center of a circle.

[0061] In this step, the normal direction of the main tooth profile surface is used as the Z axis, and the line between the center axis of the gear to be tested and the rotation center axis of the gear shaft to be tested is used as the X axis to construct an identification coordinate system, wherein the gear to be tested includes the gear shaft to be tested, and a straight line perpendicular to the line connecting the midpoint of the tooth profile area and the center of the main tooth profile surface and passing through the intersection of the Z axis and the X axis is used as the Y axis of the identification coordinate system.

[0062] Furthermore, by calculating Figure 2 The number of adjacent surfaces of each gear is calculated, and the two surfaces with the largest number of adjacent surfaces are taken as tooth profile surfaces, which are recorded as surface A (upper surface) and surface B (lower surface), respectively. The contour lines of the two tooth profile surfaces are extracted, and the distance h1 from the main tooth profile surface to the top of the containing core and the distance h2 from the auxiliary tooth profile surface to the bottom of the containing core are calculated under the tooth profile surface normal. If h2>h1, then surface A is the main tooth profile surface, otherwise surface B is the main tooth profile surface. The center of the main tooth profile surface is the coordinate origin, the normal direction of the main tooth profile surface is the coordinate Z direction, and the line connecting the gear center axis to the rotation center axis is the X axis, and the identification coordinate system is constructed.

[0063] Step S3, intercepting the inclusive core plane set according to the identified coordinate system and the preset cross section, wherein the inclusive core plane set includes a plurality of inclusive core images.

[0064] In this step, the main tooth profile surface uses the XOZ plane as the reference plane to construct a preset section in the identification coordinate system, and extracts the faces with a normal angle of less than 90° to the main plane as the main plane set, and extracts the faces with a normal angle of less than 90° to the secondary plane as the secondary plane set. The continuity of the surface is considered according to the extension relationship, and the discontinuous surface is eliminated. If there is a surface to be eliminated, that is, the mold has a slider structure.

[0065] Step S4, based on the scaled Hu invariant moment algorithm, obtaining the scaled Hu invariant moment of the inclusion kernel image.

[0066] Step S4 includes steps S41-S45, wherein:

[0067] Step S41, obtaining a triangular mesh of the main tooth profile surface, wherein the triangular mesh includes a number of triangular node coordinates. The main tooth profile surface needs to meet the following requirements: the main tooth profile surface is a smooth continuous curved surface; the curved surface has no overlap in the Z direction; the drawn surface of the cover has translation, rotation, and symmetry invariance, and the drawn surface of the cover has no scaling invariance.

[0068] In this step, the digital model of the gear to be tested can be divided into several subdomains by using an adaptive octree space segmentation method. It should be noted that the specific process of dividing the subdomains in this step is as follows: calculate the minimum and maximum values ​​of the point cloud data set in the entire digital model in three directions (x, y, z); and based on these minimum and maximum value data, construct an initial rectangular bounding box containing the digital model of the entire gear to be tested, and use it as the root node of the octree; the root node is evenly divided into eight equal-sized cuboids according to the division conditions, and the obtained child nodes are further divided into eight sub-cubic blocks until the number of point clouds in the subdomain is less than the adaptive parameter K, where K satisfies f(x)=N. 2 / K+N×K, N represents the number of point clouds, and The value of the adaptive parameter K is determined when the above formula reaches an extreme value.

[0069] Using the graph cutting method to construct a triangular mesh in each subdomain, it can be understood that the Delaunay triangulation can be constructed by a point-by-point insertion algorithm. When a new Delaunay node is added, its distance to the closest node in the Delaunay triangle is calculated. If this distance is less than a certain threshold, the node position is updated. When this distance is greater than the threshold, the original structure is retained and the new node is inserted. Finally, the vertex of the Delaunay triangulation is obtained, which not only stores the three-dimensional coordinates but also stores the visual information of this point. Each tetrahedron in the Delaunay triangulation is marked. If all the Delaunay tetrahedrons that the ray from the tetrahedron vertex to the camera center of the corresponding view passes through should be marked as external. And the tetrahedrons that the ray passes in the opposite direction are marked as internal. Due to the influence of noise in the point cloud, the marked tetrahedron will have problems such as island phenomenon. Therefore, starting from the vertex of the tetrahedron, the tetrahedrons in the neighborhood of each vertex are re-marked so that the vertex meets the manifold characteristics, thereby ensuring the manifold characteristics of the edge, which is conducive to the accuracy and speed of subsequent mesh optimization processing. The tetrahedrons in the neighborhood of the vertex are determined and clustered according to the marking value and spatial adjacency. When the number of groups is 1, the vertex is not a reconstructed surface vertex; when the number of groups is 2, the vertex is a reconstructed surface vertex and satisfies the manifold characteristics; when the number of groups is greater than 2, re-mark so that the number of groups is 2. The common triangular faces of the two tetrahedrons obtained by the final marking are approximated as the surface of the object. The denser the Delaunay triangulation, the higher the surface fit obtained.

[0070] The specific process of optimizing the obtained triangular mesh is as follows: Delete duplicate triangles. Since the subdomains obtained by dividing the point cloud space generate overlapping point sets, there will be certain duplicate triangles (triangles ABC, ACB, and BAC, etc.) when forming a triangular mesh. Find one of the triangles and all triangles with the same three vertices, keep one of them, and delete all the remaining triangles. Normalize the triangular mesh. For non-standard meshes, if the spatial quadrilateral has two diagonals, keep the diagonal that divides the maximum angle. If there is only one diagonal and the diagonal does not divide the maximum angle of the quadrilateral, discard this diagonal and connect the other diagonal. Then use the homogenized normal vector to homogenize the plane where the optimized triangular mesh is located.

[0071] See also Figure 3 , the specific process of homogenizing the normal vector is as follows: select triangle T1, the vertices are marked as A, B and C, and the order is ABC; locate triangle T2 with the same sides as triangle T1, and the vertices should be marked as A, B and D (or A, C, E or B, C, F); regardless of the original order of A, B and D, its vertex order is changed to BAD (or ACE or CBF). In this way, expand with the seed triangle until all triangles are traversed and the normal vector is normalized.

[0072] Step S42, performing region segmentation on the inclusion core image so that the inclusion core image forms M×N image sub-blocks.

[0073] In this step, appropriate segmentation parameters M and N are determined so that the size of the segmented sub-blocks is moderate, which is convenient for subsequent processing. The inclusion kernel image is preprocessed, including denoising, contrast enhancement and other operations to improve the segmentation quality. The image is evenly divided into M×N sub-blocks using a regular grid segmentation method to ensure that the segmentation boundary is consistent with the image feature line and that there is a certain overlap between adjacent sub-blocks to avoid information loss. Figure 4 As shown, the step length a of each image sub-block can be expressed as the side length of the inclusive core image in the X direction / M, and the step length b can be expressed as the side length of the inclusive core image in the Y direction / N, wherein the horizontal and vertical coordinates of each node can be expressed as (X-X0) / a, (Y-Y0) / b, respectively.

[0074] Step S43, based on the identification coordinate system and the image sub-block, obtaining the plane node coordinates of the image sub-block.

[0075] In this step, the lower left corner of the image is selected as the default origin (point O), the image centroid position is calculated as the alternative origin, the optimal origin position is determined according to the image features, the pixel coordinate value of the origin is recorded, the Canny operator is used for edge detection, and appropriate high and low thresholds are set for edge refinement to remove noise edges, the Harris corner detector is used to calculate the corner response value, and the corner detection threshold is set for non-maximum suppression. The contour tracking algorithm is used to extract contour key points, the curvature features are calculated to screen significant feature points, the node density threshold is set to merge over-dense nodes, the sparse area nodes are supplemented, and the node distribution is optimized. The image coordinates are converted to plane coordinates, the (x, y) coordinate values ​​of each node are calculated, and the coordinates are corrected based on the image resolution.

[0076] Step S44, obtaining a mapping relationship between the triangular mesh and the inclusion kernel image according to the triangular node coordinates and the plane node coordinates, wherein the mapping relationship includes a scaling factor.

[0077] In this step, we first need to establish a node correspondence table, which contains the following key information: triangular mesh node identification, plane node identification, node distance, and corresponding confidence. In terms of spatial distance calculation, the distance formula between two points in three-dimensional space is used to calculate the triangular mesh node coordinates and the plane node projection coordinates. For nearest neighbor node matching, the KD tree algorithm is mainly used to improve the search efficiency, while setting the maximum matching distance threshold, and focusing on handling many-to-one and one-to-many matching conflicts.

[0078] In the calculation of the local scaling factor, each pair of adjacent nodes needs to be processed: first, the distance between the node pairs on the triangular mesh is calculated, and then the distance between the corresponding plane node pairs is calculated. The ratio of the two is the local scaling factor. At the same time, a distance-based weight factor is introduced, and the weight calculation is performed using an exponential decay function. For the optimization of the global scaling factor, an iterative method is adopted, that is, first, all weighted local scaling factors are accumulated, and then the global scaling factor is obtained by weight normalization. After that, each local scaling factor is updated by linear combination to achieve global optimization.

[0079] The calculation of point error mainly includes recording the original position and the position after mapping, and calculating the position deviation and angle deviation. In terms of shape preservation evaluation, three aspects are mainly examined: the change in side length ratio, angle change and area change. The shape preservation degree is evaluated by comparing the changes in these geometric features before and after mapping. Multiple indicators are used to comprehensively evaluate the mapping quality. The mean square error (MSE) is used to calculate the mean of the sum of squares of all node mapping errors. The topology preservation indicator is used to check whether the adjacency relationship between nodes is maintained to evaluate the degree of preservation of the topological structure. The scaling consistency evaluation is used to calculate the variance of the local scaling factor relative to the global scaling factor to evaluate the uniformity of scaling.

[0080] Step S45, based on the scaling factors, sequentially calculating the scaled Hu invariant moments of the inclusive kernel image.

[0081] In this step, the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image is obtained.

[0082]

[0083] Among them, m pq It represents the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image, where p and q are both natural numbers and the 0-order origin moment m 00 represents the total volume of the main tooth surface in the triangular mesh, f(x, y) represents the two-dimensional distribution function of the inclusive kernel image, (x, y) represents the two-dimensional coordinates of any point in the inclusive kernel image, and u pq represents the geometric center moment of the enclosing kernel image, and Respectively represent the horizontal and vertical coordinates of the centroid of the main tooth profile surface,

[0084] The p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image is scaled by a scaling factor to obtain the i-order scaled Hu invariant moment, specifically including:

[0085] The p+q-order origin moment is scaled by the factor The scaling origin moment obtained after scaling is:

[0086]

[0087] Perform the scaling height H corresponding to the p+q order pq Specifically, it can be expressed as:

[0088]

[0089] According to the invariant moment polynomial space basis theorem (derived from the invariant moment algebraic invariant construction method), as well as the invariance theorem of TS (shift and constrained scaling invariant moment polynomial space basis) and TScR (translation, constrained scaling and rotation invariant moment polynomial space basis), 7 scaled Hu invariant moments can be obtained as shown below:

[0090]

[0091]

[0092] Among them, η pq represents the normalized central moment, γ=(p+q) / 2+1, M1 to M7 are scaled Hu invariant moments of order 1 to 7, respectively. represents the scaling factor, μ 00 Represents the zero-order origin moment m 00 The corresponding 0th-order normalized central moment.

[0093] The mapping relationship between triangular meshes and image sub-blocks in the two-dimensional plane is used to form a qualitative improvement in the connectivity of each image, which can significantly improve the accuracy of feature detection. At the same time, in order to avoid incomplete image information caused by the inconsistent rotation angle between the collected inclusion kernel plane set and the image sub-block, the inclusion kernel image in the inclusion kernel plane set is divided into image sub-blocks one by one to improve the stability of feature extraction and ensure the matching accuracy.

[0094] Step S5, comparing the scaled Hu invariant moments of all the inclusion kernel images in the inclusion kernel plane set, and performing inclusion kernel feature recognition.

[0095] Based on the above method, the present application embodiment discloses a MIM gear shaft feature recognition device based on Hu invariant moment, referring to Figure 5 The feature recognition device 1 includes a collection module 11, a processing module 12 and an analysis module 13, wherein:

[0096] The acquisition module 11 is used to obtain a digital model of the gear to be tested, wherein the digital model includes a main tooth profile surface;

[0097] The processing module 12 is used to construct an identification coordinate system based on the feature information of the main tooth profile surface, wherein the feature information includes the normal direction of the main tooth profile surface, the midpoint of the tooth profile area and the center of the circle, and to intercept an inclusive core plane set according to the identification coordinate system and a preset cross section, wherein the inclusive core plane set includes a plurality of inclusive core images, and to obtain a scaled Hu invariant moment of the inclusive core image based on a scaled Hu invariant moment algorithm;

[0098] The analysis module 13 is used to compare the scaled Hu invariant moments of all the inclusion kernel images in the inclusion kernel plane set and perform inclusion kernel feature recognition.

[0099] In one example, the acquisition module 11 is used to obtain the surface data of the gear to be tested, wherein the surface data includes the gear surface and the number of gear surfaces connected to the gear surface; determine whether the number of connected gear surfaces of the gear surface is equal to the preset surface shape number; if the number of connected gear surfaces is equal to the preset surface shape number, the gear surface is selected as the main tooth profile surface.

[0100] In one example, the processing module 12 is used to construct an identification coordinate system using the normal direction of the main tooth profile as the Z axis and the line connecting the midpoint of the tooth profile area of ​​the main tooth profile and the center of the main tooth profile as the X axis.

[0101] In one example, the processing module 12 is used to obtain a triangular mesh of the main tooth profile, wherein the triangular mesh includes a number of triangular node coordinates; perform sliding window segmentation on the inclusion core image so that the inclusion core image forms M×N image sub-blocks; based on the identification coordinate system and the image sub-block, obtain the plane node coordinates of the image sub-block; based on the triangular node coordinates and the plane node coordinates, obtain the mapping relationship between the triangular mesh and the inclusion core image, wherein the mapping relationship includes a scaling factor; based on the scaling factor, calculate the scaled Hu invariant moment of the inclusion core image in sequence.

[0102] In one example, the processing module 12 is used to obtain the p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image; scale the p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image by a scaling factor to obtain the i-th order scaled Hu invariant moment.

[0103] In one example, the processing module 12 is used to obtain the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image, specifically including:

[0104]

[0105] Among them, m pq It represents the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image, where p and q are both natural numbers and the 0-order origin moment m 00 represents the total volume of the main tooth surface in the triangular mesh, f(x, y) represents the two-dimensional distribution function of the inclusive kernel image, (x, y) represents the two-dimensional coordinates of any point in the inclusive kernel image, and upq represents the geometric center moment of the enclosing kernel image, and They represent the horizontal and vertical coordinates of the centroid of the main tooth profile surface respectively.

[0106] In one example, the processing module 12 is used to scale the p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image by a scaling factor to obtain an i order scaled Hu invariant moment, specifically including:

[0107]

[0108] Among them, η pq represents the normalized central moment, γ=(p+q) / 2+1, M1 to M7 are scaled Hu invariant moments of order 1 to 7, respectively. represents the scaling factor, μ 00 Represents the zero-order origin moment m 00 The corresponding 0th-order normalized central moment.

[0109] See also Figure 6 , is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 5 As shown, the electronic device 2 may include: at least one processor 21 , at least one network interface 24 , a user interface 23 , a memory 25 , and at least one communication bus 22 .

[0110] The communication bus 22 is used to realize the connection and communication between these components.

[0111] The user interface 23 may include a display screen (Display) and a camera (Camera), and the optional user interface 23 may also include a standard wired interface and a wireless interface.

[0112] The network interface 24 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0113] Among them, the processor 21 may include one or more processing cores. The processor 21 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 25, and calling data stored in the memory 25. Optionally, the processor 21 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 21 can integrate one or more combinations of a central processing unit (Central Processing Unit, CPU), a graphics processor (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 21, and it can be implemented separately through a chip.

[0114] Among them, the memory 25 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 25 includes a non-transitory computer-readable storage medium. The memory 25 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 25 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned method embodiments, etc.; the data storage area may store data involved in the above-mentioned method embodiments, etc. The memory 25 may also be optionally at least one storage device located away from the aforementioned processor 21. As Figure 6 As shown, the memory 25 as a computer storage medium may include an operating system, a network communication module, a user interface module, and an application program for the method for identifying the characteristic model of an injection molded gear.

[0115] exist Figure 6In the electronic device 2 shown, the user interface 23 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 21 can be used to call the application program of the recognition method of the injection molding gear feature model stored in the memory 25. When executed by one or more processors, the electronic device executes one or more methods in the above-mentioned embodiments.

[0116] A computer-readable storage medium stores instructions, which, when executed by one or more processors, enable the computer to execute one or more methods in the above-mentioned embodiments.

[0117] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0118] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0119] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0120] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0121] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0122] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0123] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A MIM gear shaft feature recognition method based on Hu invariant moment, characterized in that: The method comprises: Acquire a digital model of the gear to be tested, wherein the digital model includes a main tooth profile surface; Based on the characteristic information of the main tooth profile surface, an identification coordinate system is constructed, wherein the characteristic information includes the normal direction of the main tooth profile surface, the midpoint of the tooth profile area, and the center of a circle; According to the identified coordinate system and the preset cross section, intercepting an inclusive core plane set, wherein the inclusive core plane set includes a plurality of inclusive core images; Based on the scaled Hu invariant moment algorithm, obtaining the scaled Hu invariant moment of the inclusion kernel image; The scaled Hu invariant moments of all the inclusion kernel images in the inclusion kernel plane set are compared, and inclusion kernel feature recognition is performed.

2. The method according to claim 1, characterized in that The method for obtaining the main tooth profile surface comprises: Acquire the surface data of the gear to be tested, wherein the surface data includes the gear surface and the number of gear surfaces connected to the gear surface; Determining whether the number of connected gear surfaces of the gear surface is equal to the number of preset surface shapes; If the number of the connected gear surfaces is equal to the preset number of surface profiles, the gear surface is selected as the main tooth profile surface.

3. The method according to claim 1, characterized in that The constructing of the identification coordinate system based on the feature information of the main tooth profile surface specifically includes: The normal direction of the main tooth profile is used as the Z axis, and the line between the central axis of the gear to be measured and the rotation central axis of the gear shaft to be measured is used as the X axis to construct the identification coordinate system, wherein the gear to be measured includes the gear shaft to be measured.

4. The method according to claim 1, characterized in that The scaled Hu invariant moment algorithm based on the scaled Hu invariant moment sequentially calculates the scaled Hu invariant moment in the inclusive kernel image, specifically including: Acquire a triangular mesh of the main tooth profile surface, wherein the triangular mesh includes a plurality of triangular node coordinates; Performing region segmentation on the inclusion core image so that the inclusion core image forms M×N image sub-blocks; Based on the identification coordinate system and the image sub-block, obtaining the plane node coordinates of the image sub-block; Acquire a mapping relationship between the triangular mesh and the inclusion kernel image according to the triangular node coordinates and the plane node coordinates, wherein the mapping relationship includes a scaling factor; Based on the scaling factors, the scaled Hu invariant moments of the inclusion kernel image are calculated in sequence.

5. The method according to claim 4, characterized in that The step of sequentially calculating the scaled Hu invariant moments of the inclusion kernel image based on the scaling factor specifically includes: Obtaining the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image; The p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image is scaled by a scaling factor to obtain an i order scaled Hu invariant moment.

6. The method according to claim 5, characterized in that The obtaining of the p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image specifically includes: Among them, m pq represents the p+q-order origin moment of the two-dimensional distribution function of the inclusive kernel image, where p and q are both natural numbers and the 0-order origin moment m 00 represents the total volume of the main tooth surface in the triangular mesh, f(x, y) represents the two-dimensional distribution function of the inclusion kernel image, (x, y) represents the two-dimensional coordinates of any point in the inclusion kernel image, u pq represents the geometric center moment of the inclusion kernel image, and Respectively represent the horizontal and vertical coordinates of the centroid of the main tooth profile surface.

7. The method according to claim 6, characterized in that Scaling the p+q order origin moment of the two-dimensional distribution function of the inclusive kernel image by a scaling factor to obtain an i order scaled Hu invariant moment specifically includes: Among them, η pq represents the normalized central moment, μ 00 Represents the zero-order origin moment m 00 The corresponding 0th-order normalized central moment, γ = (p + q) / 2 + 1, M1 to M7 are scaled Hu invariant moments of order 1 to 7, respectively. Represents the scaling factor.

8. A MIM gear shaft feature recognition device based on Hu invariant moment, characterized in that: The feature recognition device (1) comprises a collection module (11), a processing module (12) and an analysis module (13), wherein: The acquisition module (11) is used to obtain a digital model of the gear to be tested, wherein the digital model includes a main tooth profile surface; The processing module (12) is used to construct an identification coordinate system based on feature information of the main tooth profile surface, wherein the feature information includes the normal direction of the main tooth profile surface, the midpoint of the tooth profile area and the center of the circle, and to intercept an inclusive core plane set according to the identification coordinate system and a preset cross section, wherein the inclusive core plane set includes a plurality of inclusive core images, and to obtain a scaled Hu invariant moment of the inclusive core image based on a scaled Hu invariant moment algorithm; The analysis module (13) is used to compare the scaled Hu invariant moments of all the inclusion kernel images in the inclusion kernel plane set and perform inclusion kernel feature recognition.

9. An electronic device, characterized in that: The electronic device (2) comprises a processor (21), a memory (25), a user interface (23) and a network interface (24), wherein the memory (25) is used to store instructions, the user interface (23) and the network interface (24) are used to communicate with other devices, and the processor (21) is used to execute the instructions stored in the memory (25) so that the electronic device (2) executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the MIM gear shaft feature recognition method based on Hu invariant moment described in any one of claims 1-7 are implemented.

Images