Workpiece surface quality detection method and device based on point cloud analysis, electronic equipment and storage medium

Abstract

The invention discloses a workpiece surface quality detection method and device based on point cloud analysis, electronic equipment and a storage medium, and belongs to the field of workpiece quality detection.The method comprises the steps that three-dimensional point cloud data of a to-be-detected workpiece is acquired, and gridding processing is conducted on the three-dimensional point cloud data to generate a regular point cloud height image; and based on the regular point cloud height image, generating a basic morphology parameter group representing the morphology of the workpiece, a composite parameter group representing the microscopic geometric characteristics of the workpiece, a material volume parameter group reflecting the bearing characteristics of the surface of the workpiece, and a space parameter group describing the texture direction information of the surface of the workpiece. And inputting the four groups of parameters into a multi-dimensional weighted fusion model for comprehensive calculation, generating a surface quality comprehensive score capable of comprehensively reflecting the surface condition, and finally determining the quality grade of the workpiece according to the score. By implementing the method, the problem that the real condition of the workpiece surface cannot be accurately reflected in the prior art can be solved.

Description

Technical Field

[0001] The present invention relates to the technical field of workpiece quality detection, and in particular to a workpiece surface quality detection method, device, electronic equipment and storage medium based on point cloud analysis. Background Art

[0002] The surface quality of a workpiece is directly related to its performance, lifespan, and reliability, influencing factors such as friction and wear, fatigue resistance, fit accuracy, and appearance. Therefore, accurate inspection of workpiece surface quality is a critical component of modern precision manufacturing and quality control. Analysis methods based on 3D point cloud data, capable of non-contact, high-precision acquisition of 3D surface topography, provide a crucial technical means for comprehensive surface quality assessment and are gaining increasing attention and widespread application in industrial inspection.

[0003] However, some current point cloud data-based workpiece surface quality inspection methods still have certain limitations in practical applications. These methods often focus on calculating a few statistically or highly correlated basic topographical parameters, such as average roughness, to characterize the overall surface profile. While this approach can reflect the surface's undulations to a certain extent, it lacks the ability to fully describe complex three-dimensional topography, making it difficult to fully reveal the surface's microscopic geometric features, texture information in specific directions, and three-dimensional properties related to load or volume. Therefore, evaluation results that rely solely on a limited number of parameters may not be comprehensive and objective, and may not accurately reflect the true condition of the workpiece surface. Summary of the Invention

[0004] The embodiments of the present invention provide a workpiece surface quality detection method, device, electronic device and storage medium based on point cloud analysis, which can solve the problem in the prior art that the real condition of the workpiece surface cannot be accurately reflected.

[0005] An embodiment of the present invention provides a method for detecting workpiece surface quality based on point cloud analysis, comprising:

[0006] Obtain three-dimensional point cloud data of the workpiece to be inspected;

[0007] Performing grid processing on the three-dimensional point cloud data to generate a regular point cloud height image;

[0008] According to the regular point cloud height image, a basic topography parameter group is calculated and generated through integral operation and normalized moment operation;

[0009] According to the regular point cloud height image, a composite parameter group is generated through gradient calculation and curvature extreme value calculation;

[0010] Counting the pixel distribution of each height interval in the regular point cloud height image to generate a height distribution histogram, and generating a load curve with the vertical axis being the height value and the horizontal axis being the cumulative probability percentage based on the height distribution histogram;

[0011] According to the load curve, a material volume parameter group is calculated and generated by using a preset load area ratio threshold;

[0012] According to the regular point cloud height image, a spatial parameter group is generated through spatial domain autocorrelation operation and frequency domain angular spectrum analysis;

[0013] Generate a comprehensive surface quality score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model;

[0014] The quality grade of the workpiece to be inspected is determined based on the comprehensive surface quality score.

[0015] Furthermore, the basic morphological parameter group is calculated and generated by integral operation and normalized moment operation according to the regular point cloud height image, including:

[0016] Performing a global integration operation on the regular point cloud height image to generate an arithmetic mean height;

[0017] Performing a root mean square operation on the regular point cloud height image to generate a root mean square height;

[0018] Calculating the third-order moment of the regular point cloud height image, and performing normalization processing on the calculation result based on the root mean square height to generate skewness;

[0019] Calculating the fourth-order moment of the regular point cloud height image, and performing normalization processing on the calculation result based on the root mean square height to generate kurtosis;

[0020] The arithmetic mean height, the root mean square height, the skewness and the kurtosis are used as a basic morphological parameter group;

[0021] The arithmetic mean height is generated by the following formula:

[0022]

[0023] Where S a is the arithmetic mean height; A is the effective area of the regular point cloud height image; (z(x, y) is the height deviation value at the coordinate (x, y) in the regular point cloud height image;

[0024] The root mean square height is generated by the following formula:

[0025]

[0026] Where S q is the root mean square height;

[0027] The skewness is generated by the following formula:

[0028]

[0029] Where S sk is the skewness;

[0030] The kurtosis is generated by the following formula:

[0031]

[0032] Where S su is the kurtosis.

[0033] Furthermore, the composite parameter group is generated by gradient calculation and curvature extreme value calculation based on the regular point cloud height image, including:

[0034] Performing a gradient operation on the regular point cloud height image to generate a root mean square inclination;

[0035] performing a surface integration operation on the regular point cloud height image to generate a surface development area ratio;

[0036] Performing curvature extreme value detection on the regular point cloud height image to generate an arithmetic mean curvature of the peak vertex;

[0037] The root mean square slope, the surface area ratio and the arithmetic mean curvature of the peak apex are used as a composite parameter group;

[0038] where the root mean square tilt is generated by the following formula:

[0039]

[0040] Where S dq is the root mean square tilt;

[0041] The surface area ratio is generated by the following formula:

[0042]

[0043] Where S dr is the surface area ratio;

[0044] The arithmetic mean curvature of the peak apex is generated by the following formula:

[0045]

[0046] Where S pcis the arithmetic mean curvature of the peak apex; n is the number of maximum curvature points.

[0047] Furthermore, the load area ratio threshold includes a first load area ratio threshold and a second load area ratio threshold; the first load area ratio threshold is greater than the second load area ratio threshold;

[0048] The calculation and generation of the material volume parameter group according to the load curve and a preset load area ratio threshold includes:

[0049] performing an integration operation on the load curve in an area above a height corresponding to a first load area ratio threshold to generate a peak material volume;

[0050] performing an integration operation on a region of the load curve between a first load area ratio threshold and a second load area ratio threshold to generate a core material volume;

[0051] performing a reverse integration operation on a region of the load curve between a first load area ratio threshold and a second load area ratio threshold to generate a core layer dead volume;

[0052] performing an integration operation on the load curve in an area below a height corresponding to a second load area ratio threshold to generate a valley void volume;

[0053] The peak material volume, the core material volume, the core invalid volume and the valley invalid volume are used as a material volume parameter group;

[0054] The peak material volume is generated by the following formula:

[0055]

[0056] Where V mp is the volume of peak material; z p1 is the height value corresponding to the cumulative probability percentage on the load curve being equal to the first load area ratio threshold p1; max is the maximum height of the regular point cloud height image; C(z) is the cumulative probability percentage of the load curve;

[0057] The core material volume is generated by the following formula:

[0058]

[0059] Where V mc is the volume of the core material; z p2 is the height value corresponding to when the cumulative probability percentage on the load curve is equal to the second load area ratio threshold value p2;

[0060] The core void volume is generated by the following formula:

[0061]

[0062] Where V vc is the core layer invalid volume;

[0063] The valley void volume is generated by the following formula:

[0064]

[0065] Where V vv is the ineffective volume of the valley; z min is the minimum height of the regular point cloud height image.

[0066] Furthermore, the spatial parameter group is generated based on the regular point cloud height image through spatial domain autocorrelation operation and frequency domain angular spectrum analysis, including:

[0067] Performing a spatial autocorrelation function operation on the regular point cloud height image to generate an anisotropy index and a minimum autocorrelation length;

[0068] Performing frequency domain angular spectrum energy integration on the regular point cloud height image to generate a texture main direction deviation angle;

[0069] Taking the anisotropy index, the minimum autocorrelation length and the texture main direction deviation angle as a spatial parameter group;

[0070] The anisotropy index is generated by the following formula:

[0071]

[0072] Where S tr is the anisotropy index; S al_long S is the longitudinal offset distance when the autocorrelation function decays to the preset autocorrelation attenuation threshold along the main direction of the texture; al_trans The lateral offset distance when the autocorrelation function decays perpendicular to the main direction of the texture to a preset autocorrelation attenuation threshold;

[0073] The minimum autocorrelation length is generated by the following formula:

[0074] S al =min(S al_long , S d_trans )

[0075] Where S al is the minimum autocorrelation length;

[0076] The texture main direction deviation angle is generated by the following formula:

[0077]

[0078] Where θ is the main direction deviation angle of the texture; F(r,φ) is the frequency domain amplitude of the regular point cloud height image in the polar coordinate system; r is the radial frequency; φ is the frequency domain direction angle; r min is the radial minimum frequency of the spectrum integration; r max is the radial maximum frequency of the spectrum integration; φ design The theoretical design direction angle of the workpiece surface texture.

[0079] Furthermore, the surface quality comprehensive score is generated based on the basic morphology parameter group, the composite parameter group, the volume parameter group and the space parameter group through a preset multi-dimensional weighted fusion model, including:

[0080] The basic morphological parameter group is normalized to generate a morphological feature vector:

[0081] Perform a logarithmic transformation on the composite parameter set to generate a functional feature vector:

[0082] A proportional coupling operation is performed on the material volume parameter group to generate a material distribution vector:

[0083] Nonlinear gain processing is performed on the spatial parameter group to generate a texture feature vector:

[0084] Input the morphological feature vector, functional feature vector, material distribution vector and texture feature vector into the preset multi-dimensional weighted fusion model to generate a comprehensive surface quality score;

[0085] The generated morphological feature vector is specifically:

[0086]

[0087] Where F1 is the morphological feature vector; σ a is the standard deviation of the arithmetic mean height; σ q is the root mean square height standard deviation; μ sk is the mean skewness; σ sk is the standard deviation of skewness; μ ku is the mean kurtosis; σ ku is the standard deviation of kurtosis;

[0088] The functional feature vector is specifically:

[0089] F2=[ln(S dq +1),ln(S dr +1),ln(|S pc |+1)]

[0090] Where F2 is the functional feature vector;

[0091] The material distribution vector is specifically:

[0092]

[0093] Where F3 is the material distribution vector; ∈ is the stabilization constant;

[0094] The texture feature vector is specifically:

[0095]

[0096] Where F4 is the texture feature vector; K θ is the direction deviation gain factor, K θ =1+0.5sin 2 θ; α is the texture normalization coefficient; β is the directional sensitivity coefficient.

[0097] The multi-dimensional weighted fusion model is specifically:

[0098] Q=||w1·F1+w2·F2+w3·F3+w4·F4||

[0099] Where Q is the comprehensive score of surface quality; w1 is the weight coefficient of the morphological feature vector; ω2 is the weight coefficient of the functional feature vector; w3 is the weight coefficient of the material distribution vector; and w4 is the weight coefficient of the texture feature vector.

[0100] Furthermore, after the three-dimensional point cloud data is gridded to generate a regular point cloud height image, the method further includes:

[0101] Calculating the arithmetic mean of the heights of all pixels in the regular point cloud height image to generate a height average;

[0102] For each pixel in the regular point cloud height image, the difference between the height value of the current pixel and the height average value is calculated, and the difference result is updated as the height value of the current pixel.

[0103] Based on the above method embodiments, the present invention provides corresponding device embodiments.

[0104] An embodiment of the present invention provides a workpiece surface quality detection device based on point cloud analysis, comprising: a three-dimensional point cloud data acquisition module, a point cloud gridding processing module, a basic morphology parameter group generation module, a composite parameter group generation module, a material volume parameter group generation module, a spatial parameter group generation module, a surface quality comprehensive score generation module, and a workpiece quality grade determination module;

[0105] The three-dimensional point cloud data acquisition module is used to acquire three-dimensional point cloud data of the workpiece to be inspected;

[0106] The point cloud gridding processing module is used to perform gridding processing on the three-dimensional point cloud data to generate a regular point cloud height image;

[0107] The basic morphology parameter group generating module is used to calculate and generate the basic morphology parameter group according to the regular point cloud height image through integral operation and normalized moment operation;

[0108] The composite parameter group generating module is used to generate a composite parameter group through gradient calculation and curvature extreme value calculation according to the regular point cloud height image;

[0109] The material volume parameter group generation module is used to count the pixel distribution of each height interval in the regular point cloud height image to generate a height distribution histogram, and generate a load curve with the vertical axis as the height value and the horizontal axis as the cumulative probability percentage based on the height distribution histogram; based on the load curve, calculate and generate the material volume parameter group through a preset load area ratio threshold;

[0110] The spatial parameter group generating module is used to generate a spatial parameter group according to the regular point cloud height image through spatial domain autocorrelation operation and frequency domain angular spectrum analysis;

[0111] The surface quality comprehensive score generating module is configured to generate a surface quality comprehensive score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model;

[0112] The workpiece quality grade determination module is used to determine the quality grade of the workpiece to be inspected based on the comprehensive surface quality score.

[0113] Based on the above method embodiment, the present invention provides a corresponding electronic device embodiment.

[0114] An embodiment of the present invention provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the method for detecting workpiece surface quality based on point cloud analysis as described in any one of the above-mentioned method embodiments is implemented.

[0115] Based on the above method embodiment, the present invention provides a corresponding storage medium embodiment.

[0116] An embodiment of the present invention provides a storage medium having a computer program stored thereon, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the workpiece surface quality detection method based on point cloud analysis as described in any one of the above-mentioned method embodiments.

[0117] Compared with the prior art, the present invention has the following beneficial effects:

[0118] Embodiments of the present invention provide a method, apparatus, electronic device, and storage medium for inspecting workpiece surface quality based on point cloud analysis. The method first acquires three-dimensional point cloud data of the workpiece to be inspected and processes it to generate a regular point cloud height image. Next, four sets of key parameters are calculated from this height image, encompassing basic topography, composite features, material volume, and spatial characteristics. Finally, these parameter sets are input into a preset multi-dimensional weighted fusion model to obtain a comprehensive surface quality score, which is then used to determine the workpiece's quality grade.

[0119] This method systematically calculates and generates a multidimensional parameter set from regular point cloud height images, encompassing basic topography, composite microgeometric features, load-bearing material volumetric properties, and spatial characteristics of surface texture. This overcomes the drawback of existing technologies, which rely on a single parameter and struggle to comprehensively and accurately characterize complex three-dimensional surface topography. Furthermore, this solution uses a pre-defined multidimensional weighted fusion model to comprehensively evaluate these diverse parameter sets, generating a unified overall surface quality score. This addresses the one-sided evaluation results and lack of objective, unified assessment criteria of traditional methods, achieving a more accurate and comprehensive assessment of workpiece surface quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0120] Figure 1 It is a flow chart of a method for detecting workpiece surface quality based on point cloud analysis provided by one embodiment of the present invention.

[0121] Figure 2 It is a structural schematic diagram of a workpiece surface quality detection device based on point cloud analysis provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0122] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0123] like Figure 1 As shown, in order to solve the problem that the existing technology cannot accurately reflect the actual condition of the workpiece surface, an embodiment of the present invention provides a workpiece surface quality detection method based on point cloud analysis, which includes at least the following steps:

[0124] Step S1, obtaining three-dimensional point cloud data of the workpiece to be inspected;

[0125] Specifically, high-precision 3D scanning equipment (such as laser scanners or structured light cameras) is used to perform non-contact data collection on the surface of the workpiece to be inspected. Multi-view point cloud stitching technology is then used to construct a complete 3D surface coordinate dataset. Noise points and outliers are removed using a point cloud filtering algorithm. The scanning resolution and field of view are automatically adapted to the workpiece's geometric characteristics and inspection requirements, ensuring complete coverage of key surface microtopography details (such as grooves, scratches, and chatter marks). The 3D point cloud data output from this step provides the raw input for subsequent analysis, and its data quality directly impacts the reliability of surface parameter calculations.

[0126] Step S2: gridding the three-dimensional point cloud data to generate a regular point cloud height image;

[0127] Specifically, the grid parameters are determined based on the spatial distribution of the point cloud data, generating a uniformly distributed 2D grid node array. An interpolation algorithm is used to map the height values of the irregular point cloud to the grid nodes, with the grid spacing smaller than the minimum sampling step of the original point cloud to ensure detail preservation. Ultimately, a height image is formed that is consistent with the digital image matrix structure, with each pixel value corresponding to the 3D height coordinate of a grid node. This step converts the discrete point cloud into a regular image structure, providing standardized input for subsequent surface parameter calculations.

[0128] In a preferred embodiment, after gridding the three-dimensional point cloud data to generate a regular point cloud height image, the method further includes:

[0129] Calculating the arithmetic mean of the heights of all pixels in the regular point cloud height image to generate a height average;

[0130] For each pixel in the regular point cloud height image, the difference between the height value of the current pixel and the height average value is calculated, and the difference result is updated as the height value of the current pixel.

[0131] Specifically, the arithmetic mean of the global pixel height values in the height image is calculated to quantify the overall height offset of the workpiece. A pixel-by-pixel subtraction operation is then performed to eliminate systematic deviations introduced by fixture tilt or z-axis translation, forcing the height distribution reference plane to align with the zero plane. This step ensures objectivity and consistency in surface topography analysis by unifying the calculated reference parameters through global height correction.

[0132] Step S3, calculating and generating a basic topography parameter group based on the regular point cloud height image through integral operation and normalized moment operation;

[0133] In a preferred embodiment, the calculation and generation of the basic topographic parameter group based on the regular point cloud height image by integral operation and normalized moment operation includes:

[0134] Performing a global integration operation on the regular point cloud height image to generate an arithmetic mean height;

[0135] Performing a root mean square operation on the regular point cloud height image to generate a root mean square height;

[0136] Calculating the third-order moment of the regular point cloud height image, and performing normalization processing on the calculation result based on the root mean square height to generate skewness;

[0137] Calculating the fourth-order moment of the regular point cloud height image, and performing normalization processing on the calculation result based on the root mean square height to generate kurtosis;

[0138] The arithmetic mean height, the root mean square height, the skewness and the kurtosis are used as a basic morphological parameter group;

[0139] The arithmetic mean height is generated by the following formula:

[0140]

[0141] Where S a is the arithmetic mean height; A is the effective area of the regular point cloud height image; z(x, y) is the height deviation value at the coordinate (x, y) in the regular point cloud height image;

[0142] The root mean square height is generated by the following formula:

[0143]

[0144] Where S q is the root mean square height;

[0145] The skewness is generated by the following formula:

[0146]

[0147] Where S sk is the skewness;

[0148] The kurtosis is generated by the following formula:

[0149]

[0150] Where S su is the kurtosis.

[0151] Specifically, the arithmetic mean height of the height distribution is obtained through global integration to characterize the overall surface roughness level. The root mean square height calculation reflects the degree of height dispersion as a homogenization indicator. The third-order moment of the height distribution and the root mean square height normalization process are used to quantify surface asymmetry, and the fourth-order moment normalization process characterizes the sharpness of the height distribution. These four parameters are integrated into a basic topographic parameter group, comprehensively covering the surface macro-roughness, distribution symmetry, and peak-to-valley extreme characteristics, providing a basic quantitative basis for surface quality analysis.

[0152] Step S4: generating a composite parameter group through gradient calculation and curvature extreme value calculation according to the regular point cloud height image;

[0153] In a preferred embodiment, the generating of a composite parameter group based on the regular point cloud height image by gradient calculation and curvature extreme value calculation includes:

[0154] Performing a gradient operation on the regular point cloud height image to generate a root mean square inclination;

[0155] performing a surface integration operation on the regular point cloud height image to generate a surface development area ratio;

[0156] Performing curvature extreme value detection on the regular point cloud height image to generate an arithmetic mean curvature of the peak vertex;

[0157] The root mean square slope, the surface area ratio and the arithmetic mean curvature of the peak apex are used as a composite parameter group;

[0158] where the root mean square tilt is generated by the following formula:

[0159]

[0160] Where S dq is the root mean square tilt;

[0161] The surface area ratio is generated by the following formula:

[0162]

[0163] Where S dr is the surface area ratio;

[0164] The arithmetic mean curvature of the peak apex is generated by the following formula:

[0165]

[0166] Where S pc is the arithmetic mean curvature of the peak apex; n is the number of maximum curvature points.

[0167] Specifically, the root mean square value of the surface slope is calculated through gradient field statistics to quantify the steepness of the microtopography and evaluate friction performance. The surface expansion rate is analyzed based on the area ratio of the three-dimensional surface to the projected surface to characterize the impact of complex structures on sealing or lubrication functions. The extreme curvature of the peak area is extracted and the arithmetic mean is calculated to reflect the potential risk of local sharpness to wear resistance and stress distribution. The three types of parameters are integrated into a composite parameter group to characterize the surface functional characteristics from multiple dimensions such as slope, complexity, and stress concentration. This step significantly improves the correlation between surface quality assessment and actual engineering needs through function-oriented composite parameter design.

[0168] Step S5: Counting the pixel distribution of each height interval in the regular point cloud height image to generate a height distribution histogram, and generating a load curve with the vertical axis being the height value and the horizontal axis being the cumulative probability percentage based on the height distribution histogram;

[0169] Specifically, the height image pixels are divided into height intervals and their frequencies are counted to generate a histogram representing the height distribution density. Based on the histogram data, the probability percentages are accumulated step by step from the lowest to the highest height, forming a load curve with cumulative probability as the horizontal axis and height as the vertical axis. This intuitively reflects the cumulative proportion characteristics of the surface height distribution from valley to peak. This step provides basic data support for the calculation of material volume parameters through probabilistic statistical modeling of height distribution.

[0170] Step S6: Calculate and generate a material volume parameter group based on the load curve and a preset load area ratio threshold;

[0171] In a preferred embodiment, the load area ratio threshold includes a first load area ratio threshold and a second load area ratio threshold; the first load area ratio threshold is greater than the second load area ratio threshold;

[0172] The calculation and generation of the material volume parameter group according to the load curve and a preset load area ratio threshold includes:

[0173] performing an integration operation on the load curve in an area above a height corresponding to a first load area ratio threshold to generate a peak material volume;

[0174] performing an integration operation on a region of the load curve between a first load area ratio threshold and a second load area ratio threshold to generate a core material volume;

[0175] performing a reverse integration operation on a region of the load curve between a first load area ratio threshold and a second load area ratio threshold to generate a core layer dead volume;

[0176] performing an integration operation on the load curve in an area below a height corresponding to a second load area ratio threshold to generate a valley void volume;

[0177] The peak material volume, the core material volume, the core invalid volume and the valley invalid volume are used as a material volume parameter group;

[0178] The peak material volume is generated by the following formula:

[0179]

[0180] Where V mp is the volume of peak material; z p1 is the height value corresponding to the cumulative probability percentage on the load curve being equal to the first load area ratio threshold p1; max is the maximum height of the regular point cloud height image; C(z) is the cumulative probability percentage of the load curve;

[0181] The core material volume is generated by the following formula:

[0182]

[0183] Where V mc is the volume of the core material; z p2 is the height value corresponding to when the cumulative probability percentage on the load curve is equal to the second load area ratio threshold value p2;

[0184] The core void volume is generated by the following formula:

[0185]

[0186] Where V vc is the core layer invalid volume;

[0187] The valley void volume is generated by the following formula:

[0188]

[0189] Where V vv is the ineffective volume of the valley; z min is the minimum height of the regular point cloud height image.

[0190] Specifically, the method divides height intervals based on preset thresholds, and quantifies the volume distribution of different surface regions through integral calculations. The peak material volume reflects the material distribution density in high-load-bearing areas, the core material volume and the void volume characterize the effective material ratio and defect void characteristics of the secondary load-bearing layer, respectively, and the valley void volume describes the volume distribution of low-lying areas. This parameter set analyzes surface characteristics from multiple perspectives, including load-bearing capacity, material uniformity, and potential defects, providing multidimensional data support for functional quality assessment.

[0191] Step S7: Generate a spatial parameter group based on the regular point cloud height image through spatial domain autocorrelation calculation and frequency domain angular spectrum analysis;

[0192] In a preferred embodiment, generating a spatial parameter group based on the regular point cloud height image by spatial domain autocorrelation operation and frequency domain angular spectrum analysis includes:

[0193] Performing a spatial autocorrelation function operation on the regular point cloud height image to generate an anisotropy index and a minimum autocorrelation length;

[0194] Performing frequency domain angular spectrum energy integration on the regular point cloud height image to generate a texture main direction deviation angle;

[0195] Taking the anisotropy index, the minimum autocorrelation length and the texture main direction deviation angle as a spatial parameter group;

[0196] The anisotropy index is generated by the following formula:

[0197]

[0198] Where S tr is the anisotropy index; S al_long S is the longitudinal offset distance when the autocorrelation function decays to the preset autocorrelation attenuation threshold along the main direction of the texture; al_trans The lateral offset distance when the autocorrelation function decays perpendicular to the main direction of the texture to a preset autocorrelation attenuation threshold;

[0199] The minimum autocorrelation length is generated by the following formula:

[0200] S al =min(S al_long , S al_trans )

[0201] Where S al is the minimum autocorrelation length;

[0202] The texture main direction deviation angle is generated by the following formula:

[0203]

[0204] Where θ is the main direction deviation angle of the texture; F(r, φ) is the frequency domain amplitude of the regular point cloud height image in the polar coordinate system; r is the radial frequency; φ is the frequency domain direction angle; r min is the radial minimum frequency of the spectrum integration; r max is the radial maximum frequency of the spectrum integration; φ design The theoretical design direction angle of the workpiece surface texture.

[0205] Specifically, the ratio of the longitudinal and transverse autocorrelation lengths of the surface texture is analyzed through spatial autocorrelation operations to quantify the degree of anisotropy and characterize the directional consistency of the texture. The minimum separation distance at which the autocorrelation function decays to a preset threshold is calculated, reflecting the microscopic repetition period of the surface texture. The main texture direction is extracted based on the frequency domain angular spectrum energy distribution, and the deviation angle from the theoretical ideal direction is calculated to evaluate the overall orientation consistency of the processed texture. The spatial parameter group describes the surface texture characteristics from multiple dimensions, including directionality, repetition period, and orientation deviation, providing a quantitative basis for evaluating processing stability and texture functional properties.

[0206] Step S8: generating a comprehensive surface quality score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model;

[0207] In a preferred embodiment, generating a comprehensive surface quality score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model includes:

[0208] The basic morphological parameter group is normalized to generate a morphological feature vector:

[0209] Perform a logarithmic transformation on the composite parameter set to generate a functional feature vector:

[0210] A proportional coupling operation is performed on the material volume parameter group to generate a material distribution vector:

[0211] Nonlinear gain processing is performed on the spatial parameter group to generate a texture feature vector:

[0212] Input the morphological feature vector, functional feature vector, material distribution vector and texture feature vector into the preset multi-dimensional weighted fusion model to generate a comprehensive surface quality score;

[0213] The generated morphological feature vector is specifically:

[0214]

[0215] Where F1 is the morphological feature vector; σ a is the standard deviation of the arithmetic mean height; σ q is the root mean square height standard deviation; μ sk is the mean skewness; σ sk is the standard deviation of skewness; μ ku is the mean kurtosis; σ ku is the standard deviation of kurtosis;

[0216] The functional feature vector is specifically:

[0217] F2=[ln(S dq +1),ln(S dr +1),ln(|S pc |+1)]

[0218] Where F2 is the functional feature vector;

[0219] The material distribution vector is specifically:

[0220]

[0221] Where F3 is the material distribution vector; ∈ is the stabilization constant;

[0222] The texture feature vector is specifically:

[0223]

[0224] Where F4 is the texture feature vector; K θ is the direction deviation gain factor, K θ =1+0.5sin 2 θ; α is the texture normalization coefficient; β is the directional sensitivity coefficient.

[0225] The multi-dimensional weighted fusion model is specifically:

[0226] Q=||w1·F1+w2·F2+w3·F3+w4·F4||

[0227] Where Q is the comprehensive score of surface quality; w1 is the weight coefficient of the morphological feature vector; w2 is the weight coefficient of the functional feature vector; w3 is the weight coefficient of the material distribution vector; and w4 is the weight coefficient of the texture feature vector.

[0228] Specifically, the basic morphological parameters (such as arithmetic mean height and root mean square height) are mapped to a unified dimension through mean-standard deviation standardization to eliminate the magnitude differences between parameters; natural logarithm transformation is applied to composite parameters (such as root mean square inclination and surface area ratio) to compress the numerical range and weaken the interference of outliers; the material volume parameters (peak, core layer, valley volume) are coupled according to the proportion of adjacent layers (such as peak volume / core layer volume, core layer volume / valley volume) to form a chain correlation relationship; directional sensitivity coefficients (such as 1+0.5*sin 2 θ), amplifying the impact of texture orientation deviation on the score. Finally, the four eigenvectors are fed into a fusion model using a linear weighting formula (weights summing to 1) and a nonlinear product term to output a comprehensive score. This step uses mathematical operations to unify feature scales and enhance the contributions of key parameters, directly linking the score results to the actual functional properties of the surface.

[0229] Step S9: Determine the quality grade of the workpiece to be inspected based on the comprehensive surface quality score.

[0230] Specifically, the comprehensive surface quality score is converted into a 0-100 point system through a linear normalization method. The normalization parameters are dynamically calculated based on the maximum and minimum scores of similar workpieces in the historical inspection data to ensure that all scoring results are unified into a fixed range. The grade judgment threshold is set according to the general industry standards: superior products (≥80 points), qualified products (60-79 points), and unqualified products (<60 points). At the same time, it supports fine-tuning the threshold interval according to different workpiece types (such as sealing surfaces and bearing tracks) (for example, the threshold for superior sealing surfaces is set to 85 points). The judgment results are transmitted to the production line sorting system in real time through the industrial communication interface, driving the robotic arm to send superior products to the finishing line, qualified products are diverted to the packaging line, and unqualified products trigger sound and light alarms and are pushed to the rework station. The system updates the normalization parameters and thresholds based on the latest inspection data every month to ensure that the rules adapt to changes in production line processes.

[0231] Based on the above method embodiments, the present invention provides corresponding device embodiments.

[0232] like Figure 2 As shown, an embodiment of the present invention provides a workpiece surface quality detection device based on point cloud analysis, comprising: a three-dimensional point cloud data acquisition module, a point cloud gridding processing module, a basic morphology parameter group generation module, a composite parameter group generation module, a material volume parameter group generation module, a space parameter group generation module, a surface quality comprehensive score generation module, and a workpiece quality grade determination module;

[0233] The three-dimensional point cloud data acquisition module is used to acquire three-dimensional point cloud data of the workpiece to be inspected;

[0234] The point cloud gridding processing module is used to perform gridding processing on the three-dimensional point cloud data to generate a regular point cloud height image;

[0235] The basic morphology parameter group generating module is used to calculate and generate the basic morphology parameter group according to the regular point cloud height image through integral operation and normalized moment operation;

[0236] The composite parameter group generating module is used to generate a composite parameter group through gradient calculation and curvature extreme value calculation according to the regular point cloud height image;

[0237] The material volume parameter group generation module is used to count the pixel distribution of each height interval in the regular point cloud height image to generate a height distribution histogram, and generate a load curve with the vertical axis as the height value and the horizontal axis as the cumulative probability percentage based on the height distribution histogram; based on the load curve, calculate and generate the material volume parameter group through a preset load area ratio threshold;

[0238] The spatial parameter group generating module is used to generate a spatial parameter group according to the regular point cloud height image through spatial domain autocorrelation operation and frequency domain angular spectrum analysis;

[0239] The surface quality comprehensive score generating module is configured to generate a surface quality comprehensive score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model;

[0240] The workpiece quality grade determination module is used to determine the quality grade of the workpiece to be inspected based on the comprehensive surface quality score.

[0241] It should be noted that the embodiment of the device described above corresponds to the above-mentioned embodiment of the present invention, and it can implement any one of the above-mentioned methods for detecting the surface quality of workpieces based on point cloud analysis of the present invention. In addition, the embodiment of the above-mentioned device is merely schematic, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the embodiment of the device provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0242] Based on the above method embodiment of the present invention, a corresponding electronic device embodiment is provided.

[0243] One embodiment of the present invention provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the workpiece surface quality detection method based on point cloud analysis described in any one of the present invention is implemented, or when the processor executes the computer program, the functions of each module in the above-mentioned device embodiments are implemented.

[0244] Exemplarily, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the terminal device.

[0245] The terminal device may be a computing device such as a desktop computer, a notebook computer, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0246] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, connecting various parts of the entire terminal device using various interfaces and lines.

[0247] The memory can be used to store the computer programs and / or modules, and the processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created based on the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0248] Based on the above method embodiment, the present invention provides a corresponding storage medium embodiment;

[0249] Another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute any one of the above-mentioned workpiece surface quality detection methods based on point cloud analysis of the present invention.

[0250] The above-mentioned storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0251] In the description of this specification, the reference terms "one embodiment," "some embodiments," "example," "specific example," or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0252] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for detecting workpiece surface quality based on point cloud analysis, characterized in that: include: Obtain three-dimensional point cloud data of the workpiece to be inspected; Performing grid processing on the three-dimensional point cloud data to generate a regular point cloud height image; According to the regular point cloud height image, a basic topography parameter group is calculated and generated through integral operation and normalized moment operation; According to the regular point cloud height image, a composite parameter group is generated through gradient calculation and curvature extreme value calculation; Counting the pixel distribution of each height interval in the regular point cloud height image to generate a height distribution histogram, and generating a load curve with the vertical axis being the height value and the horizontal axis being the cumulative probability percentage based on the height distribution histogram; According to the load curve, a material volume parameter group is calculated and generated by using a preset load area ratio threshold; According to the regular point cloud height image, a spatial parameter group is generated through spatial domain autocorrelation operation and frequency domain angular spectrum analysis; Generate a comprehensive surface quality score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model; The quality grade of the workpiece to be inspected is determined based on the comprehensive surface quality score.

2. The workpiece surface quality detection method based on point cloud analysis according to claim 1, characterized in that: The basic morphological parameter group is calculated and generated by integrating operation and normalized moment operation according to the regular point cloud height image, including: Performing a global integration operation on the regular point cloud height image to generate an arithmetic mean height; Performing a root mean square operation on the regular point cloud height image to generate a root mean square height; Calculating the third-order moment of the regular point cloud height image, and performing normalization processing on the calculation result based on the root mean square height to generate skewness; Calculating the fourth-order moment of the regular point cloud height image, and performing normalization processing on the calculation result based on the root mean square height to generate kurtosis; The arithmetic mean height, the root mean square height, the skewness and the kurtosis are used as a basic morphological parameter group; The arithmetic mean height is generated by the following formula: Where S a is the arithmetic mean height; A is the effective area of the regular point cloud height image; z(x, y) is the height deviation value at the coordinate (x, y) in the regular point cloud height image; The root mean square height is generated by the following formula: Where S q is the root mean square height; The skewness is generated by the following formula: Where S sk is the skewness; The kurtosis is generated by the following formula: Where S su is the kurtosis.

3. The workpiece surface quality detection method based on point cloud analysis according to claim 2, characterized in that: The method of generating a composite parameter group based on the regular point cloud height image by gradient calculation and curvature extreme value calculation includes: Performing a gradient operation on the regular point cloud height image to generate a root mean square inclination; performing a surface integration operation on the regular point cloud height image to generate a surface development area ratio; Performing curvature extreme value detection on the regular point cloud height image to generate an arithmetic mean curvature of the peak vertex; The root mean square slope, the surface area ratio and the arithmetic mean curvature of the peak apex are used as a composite parameter group; where the root mean square tilt is generated by the following formula: Where S dq is the root mean square tilt; The surface area ratio is generated by the following formula: Where S dr is the surface area ratio; The arithmetic mean curvature of the peak apex is generated by the following formula: Where S pc is the arithmetic mean curvature of the peak apex; n is the number of maximum curvature points.

4. The workpiece surface quality detection method based on point cloud analysis according to claim 3, characterized in that: The load area ratio threshold includes a first load area ratio threshold and a second load area ratio threshold; The first load area ratio threshold is greater than the second load area ratio threshold; The calculation and generation of the material volume parameter group according to the load curve and a preset load area ratio threshold includes: performing an integration operation on the load curve in an area above a height corresponding to a first load area ratio threshold to generate a peak material volume; performing an integration operation on a region of the load curve between a first load area ratio threshold and a second load area ratio threshold to generate a core material volume; performing a reverse integration operation on a region of the load curve between a first load area ratio threshold and a second load area ratio threshold to generate a core layer dead volume; performing an integration operation on the load curve in an area below a height corresponding to a second load area ratio threshold to generate a valley void volume; The peak material volume, the core material volume, the core invalid volume and the valley invalid volume are used as a material volume parameter group; The peak material volume is generated by the following formula: Where V mp is the volume of peak material; z p1 is the height value corresponding to the cumulative probability percentage on the load curve being equal to the first load area ratio threshold p1; max is the maximum height of the regular point cloud height image; C(z) is the cumulative probability percentage of the load curve; The core material volume is generated by the following formula: Where V mc is the volume of the core material; z p2 is the height value corresponding to when the cumulative probability percentage on the load curve is equal to the second load area ratio threshold value p2; The core void volume is generated by the following formula: Where V vc is the core layer invalid volume; The valley void volume is generated by the following formula: Where V vv is the ineffective volume of the valley; z min is the minimum height of the regular point cloud height image.

5. The workpiece surface quality detection method based on point cloud analysis according to claim 4, characterized in that: The method of generating a spatial parameter group based on the regular point cloud height image by performing a spatial autocorrelation operation and a frequency domain angular spectrum analysis includes: Performing a spatial autocorrelation function operation on the regular point cloud height image to generate an anisotropy index and a minimum autocorrelation length; Performing frequency domain angular spectrum energy integration on the regular point cloud height image to generate a texture main direction deviation angle; Taking the anisotropy index, the minimum autocorrelation length and the texture main direction deviation angle as a spatial parameter group; The anisotropy index is generated by the following formula: Where S tr is the anisotropy index; S al_long S is the longitudinal offset distance when the autocorrelation function decays to the preset autocorrelation attenuation threshold along the main direction of the texture; al_trans The lateral offset distance when the autocorrelation function decays perpendicular to the main direction of the texture to a preset autocorrelation attenuation threshold; The minimum autocorrelation length is generated by the following formula: S al =min(S al_long ,S al_trans ) Where S al is the minimum autocorrelation length; The texture main direction deviation angle is generated by the following formula: Where θ is the main direction deviation angle of the texture; F(r, φ) is the frequency domain amplitude of the regular point cloud height image in the polar coordinate system; r is the radial frequency; φ is the frequency domain direction angle; r min is the radial minimum frequency of the spectrum integration; r max is the radial maximum frequency of the spectrum integration; φ design The theoretical design direction angle of the workpiece surface texture.

6. The workpiece surface quality detection method based on point cloud analysis according to claim 5, characterized in that: The step of generating a comprehensive surface quality score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model includes: The basic morphological parameter group is normalized to generate a morphological feature vector: Perform a logarithmic transformation on the composite parameter set to generate a functional feature vector: A proportional coupling operation is performed on the material volume parameter group to generate a material distribution vector: Nonlinear gain processing is performed on the spatial parameter group to generate a texture feature vector: Input the morphological feature vector, functional feature vector, material distribution vector and texture feature vector into the preset multi-dimensional weighted fusion model to generate a comprehensive surface quality score; The generated morphological feature vector is specifically: Where F1 is the morphological feature vector; σ a is the standard deviation of the arithmetic mean height; σ q is the root mean square height standard deviation; μ sk is the mean skewness; σ sk is the standard deviation of skewness; μ ku is the mean kurtosis; σ ku is the standard deviation of kurtosis; The functional feature vector is specifically: F2=[ln(S dq +1),ln(S dr +1),ln(|S pc |+1)] Where F2 is the functional feature vector; The material distribution vector is specifically: Where F3 is the material distribution vector; ∈ is the stabilization constant; The texture feature vector is specifically: Where F4 is the texture feature vector; K θ is the direction deviation gain factor, K θ =1+0.5sin 2 θ; α is the texture normalization coefficient; β is the directional sensitivity coefficient; The multi-dimensional weighted fusion model is specifically: <h2 style=";text-align:left;direction:ltr">Q=||w1·F1+w2·F2+w3·F3+w4<h2 style=";text-align:left;direction:ltr"> · <h2 style=";text-align:left;direction:ltr"> F4|| Where Q is the comprehensive score of surface quality; w1 is the weight coefficient of the morphological feature vector; w2 is the weight coefficient of the functional feature vector; w3 is the weight coefficient of the material distribution vector; and w4 is the weight coefficient of the texture feature vector.

7. The workpiece surface quality detection method based on point cloud analysis according to claim 6, characterized in that: After the three-dimensional point cloud data is gridded to generate a regular point cloud height image, the method further includes: Calculating the arithmetic mean of the heights of all pixels in the regular point cloud height image to generate a height average; For each pixel in the regular point cloud height image, the difference between the height value of the current pixel and the height average value is calculated, and the difference result is updated as the height value of the current pixel.

8. A workpiece surface quality detection device based on point cloud analysis, characterized in that: include: 3D point cloud data acquisition module, point cloud gridding processing module, basic morphology parameter group generation module, composite parameter group generation module, material volume parameter group generation module, spatial parameter group generation module, surface quality comprehensive score generation module and workpiece quality grade determination module; The three-dimensional point cloud data acquisition module is used to acquire three-dimensional point cloud data of the workpiece to be inspected; The point cloud gridding processing module is used to perform gridding processing on the three-dimensional point cloud data to generate a regular point cloud height image; The basic morphology parameter group generating module is used to calculate and generate the basic morphology parameter group according to the regular point cloud height image through integral operation and normalized moment operation; The composite parameter group generating module is used to generate a composite parameter group through gradient calculation and curvature extreme value calculation according to the regular point cloud height image; The material volume parameter group generation module is used to count the pixel distribution of each height interval in the regular point cloud height image to generate a height distribution histogram, and generate a load curve with the vertical axis as the height value and the horizontal axis as the cumulative probability percentage based on the height distribution histogram; based on the load curve, calculate and generate the material volume parameter group through a preset load area ratio threshold; The spatial parameter group generating module is used to generate a spatial parameter group according to the regular point cloud height image through spatial domain autocorrelation operation and frequency domain angular spectrum analysis; The surface quality comprehensive score generating module is configured to generate a surface quality comprehensive score based on the basic morphology parameter group, the composite parameter group, the volume parameter group, and the space parameter group through a preset multi-dimensional weighted fusion model; The workpiece quality grade determination module is used to determine the quality grade of the workpiece to be inspected based on the comprehensive surface quality score.

9. An electronic device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for detecting the surface quality of a workpiece based on point cloud analysis according to any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the workpiece surface quality detection method based on point cloud analysis according to any one of claims 1 to 7.

Images