Ceramic wine bottle inner wall defect detection method and detection system

Abstract

The invention discloses a ceramic wine bottle inner wall defect detection method and system, and relates to the technical field of automatic detection.The method comprises the steps of A1, inner wall data collection, A2, inner wall defect feature recognition and A3, inner wall defect evaluation.A first image of the inner wall of a ceramic wine bottle is obtained through three-dimensional scanning, and a three-dimensional point cloud picture is generated through denoising processing; dividing the inner wall point cloud into a bottle shoulder stress area, a bottle bottom pressure-bearing area and a bottle neck transition area; respectively calculating Gaussian curvature distribution, Euclidean distance deviation and normal vector dispersion for different areas to identify defect types; and analyzing the risk probability of each region in combination with the use environment data in the transportation process, and quantifying the reliability risk level through the comprehensive evaluation coefficient. Through multi-dimensional defect analysis and environment evaluation, the detection precision and reliability are improved, and the method is suitable for ceramic wine bottle inner wall defect detection in an industrial recovery scene.

Description

Technical Field

[0001] The present invention relates to the field of automated detection technology, and in particular to a method and system for detecting defects on the inner wall of a ceramic wine bottle. Background Art

[0002] In the field of automated inspection technology, the importance of inspecting the inner walls of ceramic wine bottles for defects is growing. As a key packaging container for alcoholic beverages, the quality of ceramic bottles directly impacts the storage safety and brand image of the product. However, due to the inherent brittleness of ceramic materials and the complexities of the production process, various defects such as cracks, glaze peeling, glaze collapse, and wall thinning are common on the inner walls of ceramic bottles. These defects not only affect the aesthetics of the bottles but, more importantly, can reduce their compressive and impact resistance, increasing the risk of breakage during transportation and storage, posing a potential threat to the safety and quality of the alcoholic beverages.

[0003] When dealing with defects on the inner walls of ceramic wine bottles, existing technologies often ignore the combination of defect characteristics and the actual use environment of ceramic wine bottles. Ceramic wine bottles will experience various physical stresses during transportation and use, such as vibration, impact, pressure, etc. The impact of these factors on the defects on the inner walls of the bottles cannot be ignored. However, existing detection technologies lack comprehensive consideration of these environmental factors, resulting in insufficient accuracy and reliability of defect assessment results, making it difficult to meet the strict requirements for accuracy and reliability in industrial recycling detection. Regarding the technical background of defect detection on the inner walls of ceramic wine bottles, there are problems such as a lack of systematic analysis of the structure of the inner walls of the bottles, a lack of differentiated detection solutions for different areas, and a disconnect between defect assessment and the actual use environment. Therefore, it is particularly important to develop a technical solution that can comprehensively consider the structural characteristics of the inner walls of the bottles, the use environment factors, and achieve high-precision and high-reliability defect detection. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for detecting defects on the inner wall of a ceramic wine bottle, which solves the problems existing in the background technology.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions: The first aspect of the present invention provides a method for detecting defects on the inner wall of a ceramic wine bottle, comprising: A1, inner wall data acquisition: performing a three-dimensional scan on the inner wall of a recycled ceramic wine bottle to obtain a first image; performing denoising processing on the first image to obtain a second image; generating a three-dimensional point cloud map of the inner wall according to the second image, and performing point cloud region segmentation according to the three-dimensional point cloud map of the inner wall, wherein the point cloud region segmentation divides the point cloud of the inner wall of the wine bottle into a bottle shoulder stress zone, a bottle bottom pressure zone, and a bottleneck transition zone.

[0006] A2, inner wall defect feature recognition: calculate the Gaussian curvature distribution of the shoulder stress zone and identify the defect type; calculate the Euclidean distance deviation of the bottom pressure bearing zone and identify the defect type; analyze the normal vector dispersion of the bottleneck transition zone and identify the defect type; summarize the defect types of the shoulder stress zone, bottom pressure bearing zone and bottleneck transition zone.

[0007] A3, inner wall defect evaluation: according to the use environment data of the recovered ceramic wine bottle during transportation, analyze the risk probability of the shoulder stress zone, bottom pressure bearing zone and bottleneck transition zone, and then analyze the comprehensive evaluation coefficient to quantify the reliability risk level.

[0008] Preferably, the Gaussian curvature distribution of the shoulder stress zone is calculated and the defect type is identified, and the specific method is: estimating the initial point number according to the geometric size of the shoulder stress zone, discretizing the inner wall surface of the shoulder stress zone into triangular mesh point cloud to obtain a representative sample point set.

[0009] For each vertex in the triangular mesh, calculate the discrete Gaussian curvature approximation value at the vertex Where i represents the unique identification of the target vertex, i = 1, 2,..., M, M represents the total number of vertices of the triangular mesh of the shoulder stress zone, j represents the traversal index of the adjacent triangle, j = 1, 2,..., g, g represents the number of adjacent triangles of the vertex, θ j represents the internal angle of the jth adjacent triangle at the vertex, A i represents the area of the vertex, and D i is taken as the Gaussian curvature value of the point; after calculating the Gaussian curvature values of all vertices of the entire shoulder stress zone, the curvature maximum value and the curvature minimum value are extracted, and the spatial distribution pattern is identified.

[0010] The curvature maximum value, curvature minimum value and spatial distribution pattern of the shoulder stress zone are compared with the corresponding curvature maximum value interval and curvature minimum value interval and spatial distribution pattern of each type of defect stored in the database. If the curvature maximum value of the shoulder stress zone is within the curvature maximum value interval of a certain defect type, the curvature minimum value is within the curvature minimum value interval of a certain defect type, and the spatial distribution pattern is the same as that of the defect type, then the defect type is recorded as the defect type of the shoulder stress zone.

[0011] The defect type includes cracks and glaze peeling.

[0012] Preferably, the Euclidean distance deviation of the bottle bottom pressure area is calculated and the defect type is identified, and the specific method is: the measured point cloud data of the inner wall surface of the bottle bottom pressure area is recorded as the source point cloud P, the target point cloud Q corresponding to the bottle bottom pressure area is called from the pre-stored standard point cloud database, the target point cloud Q represents a standard three-dimensional model without defects, the source point cloud P is rigidly transformed by calculating the optimal rotation matrix R and the translation vector t, so that the distance between the source point cloud P and the target point cloud Q is minimized, and the accurate matching of the two point clouds is realized, the accurate matching is that the source point cloud P is mapped to another target point cloud Q, so that the points in the two groups of point clouds correspond to each other in space; after the registration is completed, an allowed surface set deviation threshold L is set; for the point pair formed after the registration, the point pair is the point in the source point cloud P and its nearest neighbor point in the target point cloud Q, the Euclidean distance of each point pair in the three-dimensional space is calculated by the Euclidean distance calculation formula The Euclidean distance of each point pair in the three-dimensional space is calculated, wherein (x 11 ,x 12 ,x 13 ) and (x 21 ,x 22 ,x 23 ) represent the three-dimensional coordinates of a point pair in the source point cloud P and the target point cloud Q respectively; compare each calculated Euclidean distance G value with the set threshold L, and judge whether the point position has a defect feature: if G>L, it indicates that the measured surface geometry at the point has a significant deviation from the standard model stored in the database, and the point is determined as a surface defect feature point; if G≤L, it indicates that the measured surface geometry at the point is within the allowed deviation range, and the point is determined as a normal point.

[0013] The Euclidean distance deviation maximum value, the Euclidean distance deviation minimum value and the spatial distribution mode of the bottle bottom pressure area are compared with the corresponding Euclidean distance deviation maximum value interval and the Euclidean distance deviation minimum value interval and the spatial distribution mode of each type of defect stored in the database, if the Euclidean distance deviation maximum value of the bottle bottom pressure area is within the Euclidean distance deviation maximum value interval of a certain defect type, the Euclidean distance deviation minimum value is within the Euclidean distance deviation minimum value interval of a certain defect type, and the spatial distribution mode is the same as that of the defect type, the defect type is recorded as the defect type of the bottle bottom pressure area.

[0014] The defect type includes cracks and glaze crushing.

[0015] Preferably, the method for analyzing the normal vector discreteness of the bottleneck transition zone and identifying the defect type is as follows: obtaining the three-dimensional point cloud data of the bottleneck transition zone, estimating the number of initial sampling points according to the geometric dimensions of the bottleneck transition zone, discretizing the inner wall surface of the bottleneck transition zone into a dense point cloud sample set, constructing a spherical domain with each sample point as the center of mass, fitting the local tangent plane of the point in the spherical domain by the least squares method, calculating the local normal vector of the sample point, and calculating the discreteness index of all normal vectors in the spherical domain. Where N represents the number of normal vectors in the spherical domain, θ r Represents the angle between the rth normal vector and the average normal vector of the spherical field, For all θ r The arithmetic mean of σ is calculated; all sample points are traversed to generate a dispersion distribution map, and the distribution map is processed using an adaptive threshold segmentation algorithm. The area that satisfies σ>T is marked as a defect area, where T is a preset dispersion threshold.

[0016] The maximum value, minimum value and distribution diagram of the normal vector dispersion in the bottleneck transition zone are compared with the maximum value interval, minimum value interval and distribution diagram of the normal vector dispersion corresponding to each defect type in the bottleneck transition zone stored in the database. If the maximum value of the normal vector dispersion in the bottleneck transition zone is within the maximum value interval of the normal vector dispersion of a certain defect type, the minimum value of the normal vector dispersion is within the minimum value interval of the normal vector dispersion of a certain defect type, and the dispersion distribution diagram is the same as that of the defect type, then the defect type is recorded as the defect type of the bottleneck transition zone.

[0017] The defect types include cracks and wall thinning.

[0018] Preferably, the risk probability of the bottle shoulder stress zone, the bottle bottom pressure zone and the bottle neck transition zone is analyzed based on the usage environment data of the recycled ceramic wine bottles during transportation, and the specific method is: obtaining defect detection data of the bottle shoulder stress zone, the bottle bottom pressure zone and the bottle neck transition zone on the inner wall of the ceramic wine bottle; the defect detection data includes: defect characteristic data of the bottle shoulder stress zone, defect characteristic data of the bottle bottom pressure zone, and defect characteristic data of the bottle neck transition zone.

[0019] A defect probability prediction model for the bottle shoulder stress area is established, and the first risk probability P1 of the bottle shoulder stress area is calculated based on the usage environment data and historical defect data of the bottle shoulder stress area.

[0020] A defect probability prediction model for the bottom pressure-bearing area of ​​the bottle is established, and the second risk probability P2 of the bottom pressure-bearing area of ​​the bottle is calculated based on the usage environment data and historical defect data of the bottom pressure-bearing area of ​​the bottle.

[0021] A bottleneck transition zone defect probability prediction model is established, and the third risk probability P3 of the bottleneck transition zone is calculated based on the usage environment data and historical defect data of the bottleneck transition zone.

[0022] Preferably, the first risk probability P1 of the bottle shoulder stress zone is calculated by a specific analysis method as follows: obtaining defect detection data of the bottle shoulder stress zone on the inner wall of the ceramic wine bottle, wherein the defect detection data includes the number of microcracks, the length of microcracks, and the proportion of glaze peeling area; normalizing the number of microcracks n, the average length of microcracks l, and the proportion of glaze peeling area s to where n 标准 、l 标准 and s 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum length of microcracks allowed, and the maximum glaze peeling area ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the vibration frequency f, acceleration a, and impact frequency F during transportation. c , the vibration frequency f, acceleration a, impact frequency F c Compared with the corresponding standard thresholds, different weight coefficients are assigned, and the stress influence factor F corresponding to the bottle shoulder stress area is obtained by weighted summation, considering the fatigue strength S of the ceramic material f , based on defect detection data, stress influence factor F and fatigue strength S of ceramic materials f , get the first risk probability of the bottle shoulder stress area

[0023] Preferably, the second risk probability P2 of the bottom pressure-bearing area of ​​the bottle is calculated by: obtaining defect detection data of the bottom pressure-bearing area of ​​the inner wall of the ceramic wine bottle, wherein the defect detection data includes the number of microcracks, the depth of microcracks, and the proportion of glaze crushing area; normalizing the number of microcracks m, the depth of microcracks d, and the proportion of glaze crushing area p to where m 标准 d 标准 and p 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum microcrack depth allowed, and the maximum glaze crushing area ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the number of wine bottle stacking layers N, the weight of a single bottle W, the dynamic impact peak pressure P during loading and unloading max , convert the number of stacked layers N into static pressure P tm , the static pressure P tm , dynamic impact peak pressure P max Compare with the corresponding standard thresholds, assign different weight coefficients, and obtain the stress influence factor G corresponding to the pressure-bearing area of ​​the bottle bottom through weighted summation to determine the compressive strength C of the ceramic material. p, based on defect detection data, stress influence factor G, compressive strength C of ceramic materials p , get the second risk probability of the pressure-bearing area at the bottom of the bottle

[0024]

[0025] Preferably, the third risk probability P3 of the bottleneck transition zone is calculated by: obtaining defect detection data of the bottleneck transition zone of the inner wall of the ceramic wine bottle, wherein the defect detection data includes the number of microcracks, the width of microcracks, and the proportion of wall thickness reduction; normalizing the number of microcracks n1, the width of microcracks w, and the proportion of wall thickness reduction z to The n1 standard, d standard, and p standard represent the maximum number of microcracks allowed, the maximum microcrack width allowed, and the maximum wall thickness reduction ratio allowed, respectively, stored in the database; according to the use environment data during transportation, the use environment data includes the maximum bending angle θ during transportation. max , transport vibration frequency f1, collision number C, the maximum bending angle θ max Convert it into bending stress value, combine the vibration frequency f1, the number of collisions C and their corresponding standard thresholds, assign different weight coefficients, and obtain the stress influence factor H corresponding to the bottleneck transition zone through weighted summation to determine the fatigue limit S of the ceramic material. e Based on the defect detection data, stress influence factor H, fatigue limit S of ceramic material e , and obtain the third risk probability of the bottleneck transition zone

[0026] Preferably, the analysis comprehensively evaluates the coefficient and quantifies the reliability risk level, and the specific method is: according to the first risk probability P1 of the bottle shoulder stress zone, the second risk probability P2 of the bottle bottom bearing zone, and the third risk probability P3 of the bottleneck transition zone, the comprehensive risk coefficient ψ=w1*P1+w2*P2+w3*P3 of the inner wall of the ceramic wine bottle is evaluated, wherein w1, w2, and w3 are the regional weight coefficients of the bottle shoulder stress zone, the bottle bottom bearing zone, and the bottleneck transition zone, respectively.

[0027] According to the comprehensive assessment coefficient of the inner wall of the ceramic wine bottle and the numerical range of the comprehensive assessment coefficient corresponding to each level of reliability risk stored in the data warehouse, the reliability risk level of the inner wall of the ceramic wine bottle is identified. The reliability risk levels include level I safety status, level II warning status, and level III high-risk status.

[0028] The second aspect of the present application provides a system for performing the ceramic wine bottle inner wall defect detection method described in the present application, comprising: an inner wall data acquisition module: for three-dimensional scanning of the inner wall of the recycled ceramic wine bottle to obtain a first image; denoising the first image to obtain a second image; generating an inner wall three-dimensional point cloud image according to the second image, and performing point cloud region segmentation according to the inner wall three-dimensional point cloud image, which divides the wine bottle inner wall point cloud into a shoulder stress area, a bottom pressure bearing area and a neck transition area.

[0029] An inner wall defect feature recognition module: for calculating the Gaussian curvature distribution of the shoulder stress area and identifying the defect type; calculating the Euclidean distance deviation of the bottom pressure bearing area and identifying the defect type; analyzing the normal vector dispersion of the neck transition area and identifying the defect type; and summarizing the defect types of the shoulder stress area, the bottom pressure bearing area and the neck transition area.

[0030] An inner wall defect evaluation module: for analyzing the risk probability of the shoulder stress area, the bottom pressure bearing area and the neck transition area according to the usage environment data of the recycled ceramic wine bottle during transportation, and then analyzing the comprehensive evaluation coefficient and quantifying the reliability risk level.

[0031] The beneficial effects of the present application are as follows: (1) A1, inner wall data acquisition of the present application, by three-dimensional scanning, obtains a first image of the inner wall of the ceramic wine bottle, generates a three-dimensional point cloud image after denoising, and divides the inner wall point cloud into a shoulder stress area, a bottom pressure bearing area and a neck transition area;

[0032] (2) A2, inner wall defect feature recognition of the present application, for different areas, respectively calculates the Gaussian curvature distribution, the Euclidean distance deviation and the normal vector dispersion to identify the defect type;

[0033] (3) A3, inner wall defect evaluation of the present application, combined with the usage environment data during transportation, analyzes the risk probability of each area, and quantifies the reliability risk level through the comprehensive evaluation coefficient. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0035] Figure 1 The present application is a method flowchart.

[0036] Figure 2 The present application is a system module schematic diagram. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present application will be apparently and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0038] Referring to Figure 1 The first aspect of the present application provides a ceramic wine bottle inner wall defect detection method, comprising: a ceramic wine bottle inner wall defect detection method, comprising: A1, inner wall data acquisition: three-dimensional scanning is performed on the inner wall of the recycled ceramic wine bottle to obtain a first image; denoising processing is performed on the first image to obtain a second image; an inner wall three-dimensional point cloud image is generated according to the second image, and point cloud region segmentation is performed according to the inner wall three-dimensional point cloud image, wherein the point cloud region segmentation divides the wine bottle inner wall point cloud into a shoulder stress area, a bottom pressure bearing area and a neck transition area.

[0039] It should be noted that the specific operation process of the inner wall data acquisition is as follows: first, place the ceramic wine bottle on the sampling table, and then use the endoscope assembly as the execution terminal of the wine bottle inner wall defect automatic detection equipment to complete the collection of the inner wall data of the ceramic wine bottle.

[0040] A2, inner wall defect feature recognition of the present application: calculating the Gaussian curvature distribution of the shoulder stress area and identifying the defect type; calculating the Euclidean distance deviation of the bottom pressure bearing area and identifying the defect type; analyzing the normal vector dispersion of the neck transition area and identifying the defect type; and summarizing the defect types of the shoulder stress area, the bottom pressure bearing area and the neck transition area;

[0041] In one specific embodiment, the calculation of the Gaussian curvature distribution of the shoulder stress area and the identification of the defect type are specifically as follows: estimating the initial point number according to the geometric size of the shoulder stress area, discretizing the inner wall surface of the shoulder stress area into a triangular mesh point cloud, and obtaining a representative sample point set;

[0042] For each vertex in the triangular mesh, the discrete Gaussian curvature approximation value at the vertex is calculated Where i represents the unique identification of the target vertex, i=1, 2,..., M, M represents the total number of vertices of the triangular mesh of the shoulder stress area, j represents the traversal index of the adjacent triangle, j=1, 2,..., g, g represents the number of adjacent triangles of the vertex, and represents the inner angle of the jth adjacent triangle at the vertex. j A i represents the area of the region of the vertex, and D i is taken as the Gaussian curvature value of the point; after calculating the Gaussian curvature values of all the vertices of the entire shoulder stress area, the curvature maximum value and the curvature minimum value are extracted, and the spatial distribution pattern is identified.

[0043] The curvature maximum value, curvature minimum value and spatial distribution mode of the bottle shoulder stress zone are compared with the curvature maximum value interval, curvature minimum value interval and spatial distribution mode of each defect type corresponding to the bottle shoulder stress zone in the database. If the curvature maximum value of the bottle shoulder stress zone is within the curvature maximum value interval of a certain defect type, the curvature minimum value is within the curvature minimum value interval of a certain defect type, and the spatial distribution mode is the same as the defect type, the defect type is recorded as the defect type of the bottle shoulder stress zone.

[0044] The defect type includes a crack and a glaze peeling.

[0045] It should be noted that the spatial distribution mode of the crack defect is arranged continuously along the spatial path, and the point group is in an elongated strip shape. The spatial distribution mode of the glaze peeling defect is in an irregular block shape.

[0046] It should be noted that the database is used to store the curvature maximum value interval, curvature minimum value interval and spatial distribution mode of each defect type corresponding to the bottle shoulder stress zone, the Euclidean distance deviation maximum value interval, Euclidean distance deviation minimum value interval and spatial distribution mode of each defect type corresponding to the bottle bottom pressure area, the normal vector dispersion maximum value interval, normal vector dispersion minimum value interval and dispersion distribution diagram of each defect type corresponding to the bottle neck transition area, the target point cloud Q of the bottle bottom pressure area, the surface geometric shape standard model, the allowed maximum number of microcracks, the allowed maximum microcrack length, the allowed maximum glaze peeling area ratio, the allowed maximum number of microcracks, the allowed maximum microcrack depth, the allowed maximum glaze crushing area ratio, the allowed maximum number of microcracks, the allowed maximum microcrack width, and the allowed maximum wall thickness thinning ratio.

[0047] In one specific embodiment, the Euclidean distance deviation of the bottle bottom pressure area is calculated and the defect type is identified. The specific method is as follows: the measured point cloud data of the inner wall surface of the bottle bottom pressure area is recorded as the source point cloud P, the target point cloud Q corresponding to the bottle bottom pressure area is retrieved from the pre-stored standard point cloud database, the target point cloud Q represents a standard three-dimensional model without defects, the source point cloud P is rigidly transformed by calculating the optimal rotation matrix R and translation vector t, so that the distance between the source point cloud P and the target point cloud Q is minimized, and the accurate matching of the two point clouds is realized. The accurate matching is that the source point cloud P is mapped to another target point cloud Q, so that the points in the two point clouds correspond to each other in space; after registration, an allowed surface set deviation threshold L is set; for the point pair formed after registration, the point pair is the point in the source point cloud P and its nearest neighbor point in the target point cloud Q, the Euclidean distance of each point pair in the three-dimensional space is calculated by the Euclidean distance calculation formula 11 12 13 ​​​) and (x 21 ,x 22 ,x 23 ) respectively represent the three-dimensional coordinates of a point pair in the source point cloud P and the target point cloud Q; each calculated Euclidean distance G value is compared with a set threshold L to determine whether the point position has a defect feature: if G>L, it indicates that the measured surface geometry at the point has a significant deviation from the standard model stored in the database, and the point is determined as a surface defect feature point; if G≤L, it indicates that the measured surface geometry at the point is within the allowable deviation range, and the point is determined as a normal point;

[0048] The Euclidean distance deviation maximum value, the Euclidean distance deviation minimum value and the spatial distribution pattern of the bottle bottom pressure area are compared with the corresponding Euclidean distance deviation maximum value interval and the Euclidean distance deviation minimum value interval and the spatial distribution pattern of each defect type stored in the database in the bottle bottom pressure area. If the Euclidean distance deviation maximum value of the bottle bottom pressure area is within the Euclidean distance deviation maximum value interval of a certain defect type, the Euclidean distance deviation minimum value is within the Euclidean distance deviation minimum value interval of a certain defect type, and the spatial distribution pattern is the same as that of the defect type, the defect type is recorded as the defect type of the bottle bottom pressure area.

[0049] The defect type includes a crack and a glaze crushing.

[0050] It should be noted that the spatial distribution pattern of the crack defect is arranged in an elongated band along the principal stress direction; and the spatial distribution pattern of the glaze crushing defect is in a central radiation circular shape.

[0051] In one specific embodiment, the normal vector dispersion of the bottle neck transition area is analyzed and the defect type is identified, and the specific method is as follows: three-dimensional point cloud data of the bottle neck transition area is obtained, the initial sampling point number is estimated according to the geometric size of the bottle neck transition area, the inner wall surface of the bottle neck transition area is discretized into a dense point cloud sample set, for each sample point, a spherical field is constructed with the sample point as the center of mass, a local tangent plane of the points in the spherical field is fitted by the least square method, and the local normal vector of the sample point is calculated, and the dispersion index of all normal vectors in the spherical field is calculated where N represents the number of normal vectors in the spherical field, θ r represents the included angle between the rth normal vector and the average normal vector of the spherical field, is the arithmetic mean of all θ r ; all sample points are traversed to generate a dispersion distribution map, an adaptive threshold segmentation algorithm is used to process the distribution map, and the region satisfying σ>T is marked as a defect region, where T is a preset dispersion threshold;

[0052] It should be noted that the dispersion index of the normal vector refers to quantifying the abnormality of the surface geometry of the bottleneck transition zone, if the surface of the inner wall of the ceramic wine bottle is smooth, the direction of the normal vector changes gently, if the surface of the inner wall of the ceramic wine bottle has defects, the direction of the normal vector changes sharply.

[0053] The maximum value of the normal vector dispersion of the bottleneck transition zone, the minimum value of the normal vector dispersion and the dispersion distribution graph are compared with the corresponding maximum value interval and minimum value interval of the normal vector dispersion and the dispersion distribution graph of each type of defect stored in the database, if the maximum value of the normal vector dispersion of the bottleneck transition zone is in the maximum value interval of the normal vector dispersion of a certain defect type, the minimum value of the normal vector dispersion is in the minimum value interval of the normal vector dispersion of a certain defect type, and the dispersion distribution graph is the same as the defect type, the defect type is recorded as the defect type of the bottleneck transition zone.

[0054] The defect type includes cracks and wall thickness reduction.

[0055] It should be noted that the dispersion distribution graph of the crack defect is linear along the curved stress path, and the dispersion distribution graph of the wall thickness reduction is uniformly distributed in a large area.

[0056] A3, inner wall defect evaluation of the application: according to the use environment data of the recycled ceramic wine bottle in the transportation process, the risk probability of the shoulder stress area, the bottom pressure bearing area and the bottleneck transition area is analyzed, and then the comprehensive evaluation coefficient is analyzed, and the reliability risk level is quantified.

[0057] In one specific embodiment, the use environment data of the recycled ceramic wine bottle in the transportation process is analyzed, and the risk probability of the shoulder stress area, the bottom pressure bearing area and the bottleneck transition area is analyzed.

[0058] Obtain the defect detection data of the shoulder stress area, the bottom pressure bearing area and the bottleneck transition area of the inner wall of the ceramic wine bottle; the defect detection data includes: defect feature data of the shoulder stress area, defect feature data of the bottom pressure bearing area, defect feature data of the bottleneck transition area;

[0059] A defect probability prediction model of the shoulder stress area is established, and the first risk probability P1 of the shoulder stress area is calculated according to the use environment data and historical defect data of the shoulder stress area;

[0060] A defect probability prediction model of the bottom bearing area is established, and the second risk probability P2 of the bottom bearing area is calculated according to the use environment data and historical defect data of the bottom bearing area;

[0061] A defect probability prediction model of the bottleneck transition area is established, and the third risk probability P3 of the bottleneck transition area is calculated according to the use environment data and historical defect data of the bottleneck transition area.

[0062] In a specific embodiment, the first risk probability P1 of the bottle shoulder stress zone is calculated, and its specific analysis method is as follows: obtaining defect detection data of the bottle shoulder stress zone on the inner wall of the ceramic wine bottle, wherein the defect detection data includes the number of microcracks, the length of microcracks, and the proportion of glaze peeling area; the number of microcracks n, the average length of microcracks l, and the proportion of glaze peeling area s are normalized to where n 标准 、l 标准 and s 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum length of microcracks allowed, and the maximum glaze peeling area ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the vibration frequency f, acceleration a, and impact frequency F during transportation. c , the vibration frequency f, acceleration a, impact frequency F c Compared with the corresponding standard thresholds, different weight coefficients are assigned, and the stress influence factor F corresponding to the bottle shoulder stress area is obtained by weighted summation, considering the fatigue strength S of the ceramic material f , based on defect detection data, stress influence factor F and fatigue strength S of ceramic materials f , get the first risk probability of the bottle shoulder stress area

[0063] It should be noted that the fatigue strength of the ceramic material refers to the stress value of fatigue failure of the ceramic material under cyclic load.

[0064] In a specific embodiment, the second risk probability P2 of the bottom pressure-bearing area of ​​the bottle is calculated by: obtaining defect detection data of the bottom pressure-bearing area of ​​the inner wall of the ceramic wine bottle, the defect detection data including the number of microcracks, the depth of microcracks, and the proportion of glaze crushing area, normalizing the number of microcracks m, the depth of microcracks d, and the proportion of glaze crushing area p to where m 标准 d 标准 and p 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum microcrack depth allowed, and the maximum glaze crushing area ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the number of wine bottle stacking layers N, the weight of a single bottle W, the dynamic impact peak pressure P during loading and unloading max , convert the number of stacked layers N into static pressure P tm , the static pressure P tm , dynamic impact peak pressure P max Compare with the corresponding standard thresholds, assign different weight coefficients, and obtain the stress influence factor G corresponding to the pressure-bearing area of ​​the bottle bottom through weighted summation to determine the compressive strength C of the ceramic material. p, based on defect detection data, stress influence factor G, compressive strength C of ceramic materials p , get the second risk probability of the pressure-bearing area at the bottom of the bottle

[0065] It should be noted that the compressive strength of the ceramic material refers to the stress value that the ceramic material can withstand under a compressive load without being damaged.

[0066] In a specific embodiment, the third risk probability P3 of the bottleneck transition zone is calculated by: obtaining defect detection data of the bottleneck transition zone of the inner wall of the ceramic wine bottle, wherein the defect detection data includes the number of microcracks, the width of microcracks, and the proportion of wall thickness reduction; the number of microcracks n1, the width of microcracks w, and the proportion of wall thickness reduction z are normalized to where n1 标准 d 标准 and p 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum microcrack width allowed, and the maximum wall thickness reduction ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the maximum bending angle θ during transportation max , transport vibration frequency f1, collision number C, the maximum bending angle θ max Convert it into bending stress value, combine the vibration frequency f1, the number of collisions C and their corresponding standard thresholds, assign different weight coefficients, and obtain the stress influence factor H corresponding to the bottleneck transition zone through weighted summation to determine the fatigue limit S of the ceramic material. e Based on the defect detection data, stress influence factor H, fatigue limit S of ceramic material e , and obtain the third risk probability of the bottleneck transition zone

[0067] It should be noted that the fatigue limit of the ceramic material refers to the stress value at which the ceramic material can withstand repeated forces without breaking.

[0068] In a specific embodiment, the comprehensive evaluation coefficient is analyzed to quantify the reliability risk level. The specific method is as follows: based on the first risk probability P1 of the bottle shoulder stress zone, the second risk probability P2 of the bottle bottom bearing zone, and the third risk probability P3 of the bottle neck transition zone, the comprehensive risk coefficient ψ=w1*P1+w2*P2+w3*P3 of the inner wall of the ceramic wine bottle is evaluated, wherein w1, w2, and w3 are the regional weight coefficients of the bottle shoulder stress zone, the bottle bottom bearing zone, and the bottle neck transition zone, respectively;

[0069] According to the comprehensive assessment coefficient of the inner wall of the ceramic wine bottle and the numerical range of the comprehensive assessment coefficient corresponding to each level of reliability risk stored in the data warehouse, the reliability risk level of the inner wall of the ceramic wine bottle is identified. The reliability risk levels include level I safety status, level II warning status, and level III high-risk status.

[0070] Reference Figure 2 As shown, the second aspect of the present invention provides a system for executing the ceramic wine bottle inner wall defect detection method of the present invention, comprising: an inner wall data acquisition module, an inner wall defect feature recognition module, an inner wall defect evaluation module and a database.

[0071] It should be noted that the inner wall data acquisition module is connected to the inner wall defect feature recognition module, the inner wall defect feature recognition module is connected to the inner wall defect evaluation module, and the database is connected to the inner wall data acquisition module, the inner wall defect feature recognition module and the inner wall defect evaluation module.

[0072] The inner wall data acquisition module is used to perform a three-dimensional scan on the inner wall of the recycled ceramic wine bottle to obtain a first image; perform denoising processing on the first image to obtain a second image; generate a three-dimensional point cloud map of the inner wall based on the second image, and perform point cloud region segmentation based on the three-dimensional point cloud map of the inner wall. The point cloud region segmentation divides the point cloud of the inner wall of the wine bottle into a bottle shoulder stress area, a bottle bottom pressure-bearing area, and a bottleneck transition area.

[0073] The inner wall defect feature recognition module is used to calculate the Gaussian curvature distribution of the bottle shoulder stress zone and identify the defect type; calculate the Euclidean distance deviation of the bottle bottom pressure zone and identify the defect type; analyze the normal vector discreteness of the bottleneck transition zone and identify the defect type; and summarize the defect types of the bottle shoulder stress zone, bottle bottom pressure zone, and bottleneck transition zone.

[0074] The inner wall defect assessment module is used to analyze the risk probabilities of the bottle shoulder stress area, the bottle bottom pressure area, and the bottleneck transition area based on the usage environment data of the recycled ceramic wine bottles during transportation, and then analyze the comprehensive assessment coefficient to quantify the reliability risk level.

[0075] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all fall within the scope of protection of the present invention.

Claims

1. A method for detecting defects on the inner wall of a ceramic wine bottle, characterized in that: include: A1. Inner wall data acquisition: Perform a three-dimensional scan of the inner wall of the recycled ceramic wine bottle to obtain a first image; De-noising the first image to obtain a second image; generating a three-dimensional point cloud image of the inner wall based on the second image; and performing point cloud region segmentation based on the three-dimensional point cloud image of the inner wall. The point cloud region segmentation divides the point cloud of the inner wall of the wine bottle into a shoulder stress area, a bottom pressure-bearing area, and a neck transition area. A2. Inner wall defect feature identification: Calculate the Gaussian curvature distribution of the bottle shoulder stress zone and identify the defect type; calculate the Euclidean distance deviation of the bottle bottom pressure zone and identify the defect type; analyze the normal vector dispersion of the bottle neck transition zone and identify the defect type; summarize the defect types of the bottle shoulder stress zone, bottle bottom pressure zone, and bottle neck transition zone; A3. Inner wall defect assessment: Based on the usage environment data of the recycled ceramic wine bottles during transportation, the risk probability of the bottle shoulder stress area, bottle bottom pressure area, and bottleneck transition area is analyzed, and then the comprehensive assessment coefficient is analyzed to quantify the reliability risk level.

2. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 1, characterized in that: The specific method for calculating the Gaussian curvature distribution of the bottle shoulder stress area and identifying the defect type is as follows: The number of initial points is estimated based on the geometric dimensions of the bottle shoulder stress zone, and the inner wall surface of the bottle shoulder stress zone is discretized into a triangular grid point cloud to obtain a representative sample point set. For each vertex in the triangle mesh, calculate the discrete Gaussian curvature approximation at the vertex Where i represents the unique identifier of the target vertex, i = 1, 2, ..., M, M represents the total number of vertices of the triangle mesh of the shoulder stress area, j represents the traversal index of the adjacent triangle, j = 1, 2, ..., g, g represents the number of adjacent triangles of the vertex, θ j represents the interior angle of the jth adjacent triangle at the vertex, A i Represents the area of ​​the vertex, D i As the Gaussian curvature value of the point; After calculating the Gaussian curvature values ​​of all vertices in the entire bottle shoulder stress zone, the maximum and minimum curvature values ​​are extracted and the spatial distribution pattern is identified; Compare the curvature maximum, curvature minimum, and spatial distribution pattern of the bottle shoulder stress zone with the curvature maximum interval, curvature minimum interval, and spatial distribution pattern corresponding to various defect types in the bottle shoulder stress zone stored in the database. If the curvature maximum of the bottle shoulder stress zone is within the curvature maximum interval of a certain defect type, the curvature minimum is within the curvature minimum interval of a certain defect type, and the spatial distribution pattern is the same as that of the defect type, then record the defect type as the defect type of the bottle shoulder stress zone; The defect types mentioned include cracks and glaze peeling.

3. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 1, characterized in that: The specific method for calculating the Euclidean distance deviation of the pressure-bearing area of ​​the bottle bottom and identifying the defect type is as follows: The measured point cloud data of the inner wall surface of the pressure-bearing area of ​​the bottle bottom is recorded as the source point cloud P, and the target point cloud Q corresponding to the pressure-bearing area of ​​the bottle bottom is retrieved from the pre-stored standard point cloud database. The target point cloud Q represents a standard three-dimensional model without defects. By calculating the optimal rotation matrix R and translation vector t, the source point cloud P is rigidly transformed to minimize the sum of the distances between it and the target point cloud Q, thereby achieving precise matching of the two point clouds. The precise matching is to map the source point cloud P to another target point cloud Q so that the points in the two sets of point clouds correspond one to one in space. After the registration is completed, an allowable surface set deviation threshold L is set. For the point pairs formed after the registration, the point pairs are points in the source point cloud P and their nearest neighbor points in the target point cloud Q. The Euclidean distance calculation formula is used. Calculate the Euclidean distance of each point pair in three-dimensional space, where (x 11 ,x 12 ,x 13 ) and (x 21 ,x 22 ,x 23 ) represent the three-dimensional coordinates of a point pair in the source point cloud P and the target point cloud Q respectively; compare each calculated Euclidean distance G value with the set threshold L to determine whether there is a defect feature at the position of the point: if G>L, it indicates that there is a significant deviation between the measured surface geometry at the point and the standard model stored in the database, and the point is determined to be a surface defect feature point; if G≤L, it indicates that the measured surface geometry at the point is within the allowable deviation range, and the point is determined to be a normal point; Compare the maximum Euclidean distance deviation, minimum Euclidean distance deviation, and spatial distribution pattern of the bottom pressure zone with the maximum Euclidean distance deviation interval, minimum Euclidean distance deviation interval, and spatial distribution pattern corresponding to various defect types in the bottom pressure zone stored in the database. If the maximum Euclidean distance deviation of the bottom pressure zone is within the maximum Euclidean distance deviation interval of a certain defect type, the minimum Euclidean distance deviation is within the minimum Euclidean distance deviation interval of a certain defect type, and the spatial distribution pattern is the same as that of the defect type, then record the defect type as the defect type of the bottom pressure zone; The defect types include cracks and glaze collapse.

4. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 1, characterized in that: The specific method of analyzing the normal vector discreteness of the bottleneck transition zone and identifying the defect type is as follows: Obtain the three-dimensional point cloud data of the bottleneck transition zone, estimate the number of initial sampling points based on the geometric dimensions of the bottleneck transition zone, discretize the inner wall surface of the bottleneck transition zone into a dense point cloud sample set, and construct a spherical domain with each sample point as the center of mass. Fit the local tangent plane of the point in the spherical domain by the least squares method, calculate the local normal vector of the sample point, and calculate the discreteness index of all normal vectors in the spherical domain. Where N represents the number of normal vectors in the spherical domain, θ r Represents the angle between the rth normal vector and the average normal vector of the spherical field, For all θ r The arithmetic mean of σ is calculated; all sample points are traversed to generate a dispersion distribution map, and the distribution map is processed using an adaptive threshold segmentation algorithm. The area that satisfies σ>T is marked as a defect area, where T is a preset dispersion threshold; Compare the maximum value, minimum value, and dispersion distribution diagram of the normal vector dispersion in the bottleneck transition zone with the maximum value interval, minimum value interval, and dispersion distribution diagram of the normal vector dispersion corresponding to each defect type in the bottleneck transition zone stored in the database. If the maximum value of the normal vector dispersion in the bottleneck transition zone is within the maximum value interval of the normal vector dispersion of a certain defect type, the minimum value of the normal vector dispersion is within the minimum value interval of the normal vector dispersion of a certain defect type, and the dispersion distribution diagram is the same as that of the defect type, then record the defect type as the defect type of the bottleneck transition zone; The defect types include cracks and wall thinning.

5. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 1, characterized in that: The risk probability of the bottle shoulder stress zone, bottle bottom pressure zone, and bottle neck transition zone is analyzed based on the usage environment data of the recycled ceramic wine bottles during transportation. The specific method is as follows: Obtaining defect detection data for the shoulder stress zone, the bottom pressure zone, and the neck transition zone of the inner wall of the ceramic wine bottle; the defect detection data includes defect characteristic data for the shoulder stress zone, the bottom pressure zone, and the neck transition zone; Establish a defect probability prediction model for the bottle shoulder stress area, and calculate the first risk probability P1 of the bottle shoulder stress area based on the usage environment data and historical defect data of the bottle shoulder stress area; Establish a defect probability prediction model for the bottom pressure-bearing area of ​​the bottle. Calculate the second risk probability P2 of the bottom pressure-bearing area based on the usage environment data and historical defect data of the bottom pressure-bearing area of ​​the bottle. A bottleneck transition zone defect probability prediction model is established, and the third risk probability P3 of the bottleneck transition zone is calculated based on the usage environment data and historical defect data of the bottleneck transition zone.

6. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 1, characterized in that: The specific analysis method for calculating the first risk probability P1 of the bottle shoulder stress area is as follows: Obtain defect detection data of the shoulder stress zone of the inner wall of the ceramic wine bottle, which includes the number of microcracks, the length of microcracks, and the proportion of glaze peeling area. The number of microcracks n, the average length of microcracks l, and the proportion of glaze peeling area s are normalized to where n 标准 、l 标准 and s 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum length of microcracks allowed, and the maximum glaze peeling area ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the vibration frequency f, acceleration a, and impact frequency F during transportation. c , the vibration frequency f, acceleration a, impact frequency F c Compared with the corresponding standard thresholds, different weight coefficients are assigned, and the stress influence factor F corresponding to the bottle shoulder stress area is obtained by weighted summation, considering the fatigue strength S of the ceramic material f , based on defect detection data, stress influence factor F and fatigue strength S of ceramic materials f , get the first risk probability of the bottle shoulder stress area .

7. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 6, characterized in that: The specific method for calculating the second risk probability P2 of the bottle bottom pressure-bearing area is as follows: Obtain the defect detection data of the bottom pressure zone of the inner wall of the ceramic wine bottle, which includes the number of microcracks, the depth of microcracks, and the proportion of glaze crushing area. The number of microcracks m, the depth of microcracks d, and the proportion of glaze crushing area p are normalized to where m 标准 d 标准 and p 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum microcrack depth allowed, and the maximum glaze crushing area ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the number of wine bottle stacking layers N, the weight of a single bottle W, the dynamic impact peak pressure P during loading and unloading max , convert the number of stacked layers N into static pressure P tm , the static pressure P tm , dynamic impact peak pressure P max Compare with the corresponding standard thresholds, assign different weight coefficients, and obtain the stress influence factor G corresponding to the pressure-bearing area of ​​the bottle bottom through weighted summation to determine the compressive strength C of the ceramic material. p , based on defect detection data, stress influence factor G, compressive strength C of ceramic materials p , get the second risk probability of the pressure-bearing area at the bottom of the bottle 8. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 1, characterized in that: The specific method for calculating the third risk probability P3 of the bottleneck transition zone is as follows: Obtain defect detection data of the bottleneck transition zone of the inner wall of the ceramic wine bottle, the defect detection data includes the number of microcracks, the width of microcracks, and the proportion of wall thickness thinning. The number of microcracks n1, the width of microcracks w, and the proportion of wall thickness thinning z are normalized to where n1 标准 d 标准 and p 标准 They are respectively represented by the maximum number of microcracks allowed, the maximum microcrack width allowed, and the maximum wall thickness reduction ratio allowed stored in the database; according to the use environment data during transportation, the use environment data includes the maximum bending angle θ during transportation max , transport vibration frequency f1, collision number C, the maximum bending angle θ max Convert it into bending stress value, combine the vibration frequency f1, the number of collisions C and their corresponding standard thresholds, assign different weight coefficients, and obtain the stress influence factor H corresponding to the bottleneck transition zone through weighted summation to determine the fatigue limit S of the ceramic material. e Based on the defect detection data, stress influence factor H, fatigue limit S of ceramic material e , and get the third risk probability of the bottleneck transition zone 9. The method for detecting defects on the inner wall of a ceramic wine bottle according to claim 1, characterized in that: The specific method for analyzing the comprehensive evaluation coefficient and quantifying the reliability risk level is as follows: According to the first risk probability P1 of the bottle shoulder stress zone, the second risk probability P2 of the bottle bottom bearing zone, and the third risk probability P3 of the bottle neck transition zone, the comprehensive risk coefficient ψ=w1*P1+w2*P2+w3*P3 of the inner wall of the ceramic wine bottle is evaluated, where w1, w2, and w3 are the regional weight coefficients of the bottle shoulder stress zone, the bottle bottom bearing zone, and the bottle neck transition zone, respectively. According to the comprehensive assessment coefficient of the inner wall of the ceramic wine bottle and the numerical range of the comprehensive assessment coefficient corresponding to each level of reliability risk stored in the data warehouse, the reliability risk level of the inner wall of the ceramic wine bottle is identified. The reliability risk levels include level I safety status, level II warning status, and level III high-risk status.

10. A system for executing the method for detecting defects on the inner wall of a ceramic wine bottle according to any one of claims 1 to 9, comprising: Inner wall data acquisition module: performs three-dimensional scanning on the inner wall of the recycled ceramic wine bottle to obtain a first image; De-noising the first image to obtain a second image; generating a three-dimensional point cloud image of the inner wall based on the second image; and performing point cloud region segmentation based on the three-dimensional point cloud image of the inner wall. The point cloud region segmentation divides the point cloud of the inner wall of the wine bottle into a shoulder stress area, a bottom pressure-bearing area, and a neck transition area. Inner wall defect feature recognition module: Calculates the Gaussian curvature distribution of the bottle shoulder stress zone and identifies the defect type; calculates the Euclidean distance deviation of the bottle bottom pressure zone and identifies the defect type; analyzes the normal vector discreteness of the bottle neck transition zone and identifies the defect type; summarizes the defect types of the bottle shoulder stress zone, bottle bottom pressure zone, and bottle neck transition zone; Inner wall defect assessment module: Based on the usage environment data of recycled ceramic wine bottles during transportation, the module analyzes the risk probability of the bottle shoulder stress area, the bottle bottom pressure area, and the bottleneck transition area, and then analyzes the comprehensive assessment coefficient to quantify the reliability risk level.

Images