CN115236092B - Guqin surface material damage detection and identification method based on optical means - Google Patents

Guqin surface material damage detection and identification method based on optical means Download PDF

Info

Publication number
CN115236092B
CN115236092B CN202211151374.4A CN202211151374A CN115236092B CN 115236092 B CN115236092 B CN 115236092B CN 202211151374 A CN202211151374 A CN 202211151374A CN 115236092 B CN115236092 B CN 115236092B
Authority
CN
China
Prior art keywords
point
target
dimensional image
grid
discrete index
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202211151374.4A
Other languages
Chinese (zh)
Other versions
CN115236092A (en
Inventor
熊立群
薛磊
吴开平
熊颖
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yangzhou Jinyun Musical Instruments Co ltd
Original Assignee
Yangzhou Jinyun Musical Instruments Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yangzhou Jinyun Musical Instruments Co ltd filed Critical Yangzhou Jinyun Musical Instruments Co ltd
Priority to CN202211151374.4A priority Critical patent/CN115236092B/en
Publication of CN115236092A publication Critical patent/CN115236092A/en
Application granted granted Critical
Publication of CN115236092B publication Critical patent/CN115236092B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G06T17/30—Polynomial surface description
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10028—Range image; Depth image; 3D point clouds

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Geometry (AREA)
  • Pathology (AREA)
  • Computer Graphics (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Immunology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Analytical Chemistry (AREA)
  • Quality & Reliability (AREA)
  • Signal Processing (AREA)
  • Algebra (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Image Analysis (AREA)

Abstract

The invention relates to a detection and identification method for surface material damage of a guqin based on an optical means, belonging to the technical field of material testing and analysis. The method comprises the steps of obtaining target three-dimensional images of all surfaces of a guqin to be detected by utilizing line structured light; calculating the distance from each point in the point cloud to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud; then constructing a triangular network of point clouds corresponding to the target three-dimensional image; obtaining a second discrete index of each point in the point cloud according to the area of each triangle in the triangular net taking each point as a common vertex; screening each point in the point cloud corresponding to the target three-dimensional image according to the first discrete index and the second discrete index of each point to obtain each target point in the point cloud corresponding to the target three-dimensional image; constructing and obtaining a feature vector corresponding to each target point according to the coordinates of each target point; and judging whether the damage of each surface of the seven-stringed plucked instrument to be detected occurs or not according to the characteristic vector. The invention can reliably detect and identify the damage of the surface material of the guqin.

Description

Guqin surface material damage detection and identification method based on optical means
Technical Field
The invention relates to the technical field of material testing and analysis, in particular to a detection and identification method for damage of a surface material of a guqin based on an optical means.
Background
The guqin is an ancient chinese traditional musical instrument, is longer tile arc panel and a trapezoidal floor by the one and constitutes, and the guqin can lead to the guqin surface to appear abnormal phenomena because external force in production process or the use, and then can influence the tone quality, the tone quality etc. of guqin, for example damaged phenomenon appears in the guqin surface.
The existing method for detecting and identifying the damage or surface abnormality of the surface material of the guqin is generally based on a manual mode, the manual mode is generally obtained by depending on a pianist or a user through tone rhythm debugging and type quality observation, the detection method is greatly influenced by human subjectivity, the detection efficiency is lower, the detection accuracy cannot be guaranteed, and small damages or abnormalities are difficult to detect, so that the reliability of detecting and identifying the damage of the surface material of the guqin based on manual work is lower.
Disclosure of Invention
The invention provides a detection and identification method for surface material damage of a guqin based on an optical means, which is used for solving the problem that the existing method can not reliably detect and identify the surface material damage of the guqin, and adopts the following technical scheme:
in a first aspect, an embodiment of the present invention provides a method for detecting and identifying surface material breakage of a guqin based on an optical means, including the following steps:
acquiring target three-dimensional images of all surfaces of the guqin to be detected; converting the target three-dimensional images corresponding to the surfaces into point clouds to obtain point clouds corresponding to the target three-dimensional images;
fitting the point cloud with a three-dimensional curved surface to obtain a fitted curved surface corresponding to the target three-dimensional image; calculating the distance from each point in the point cloud corresponding to the target three-dimensional image to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image;
constructing a triangulation of the point cloud; obtaining second discrete indexes of each point in the point cloud corresponding to the target three-dimensional image according to the area of each triangle taking each point as a common vertex in the triangular net;
screening each point in the point cloud corresponding to the target three-dimensional image according to the first discrete index of each point and the second discrete index of each point to obtain each target point in the point cloud corresponding to the target three-dimensional image;
constructing and obtaining a characteristic vector corresponding to each target point according to the coordinates of each target point; and judging whether the damage occurs on each surface of the guqin to be detected according to the characteristic vector.
Has the advantages that: the method comprises the steps of obtaining a target three-dimensional image of each surface of the guqin to be detected by utilizing line structured light; fitting the three-dimensional curved surface of the point cloud corresponding to the target three-dimensional image corresponding to each surface to obtain the fitted curved surface of the target three-dimensional image corresponding to each surface; then calculating the distance from each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface; then constructing a triangular network of point clouds corresponding to the target three-dimensional images corresponding to the surfaces; according to the area of each triangle in the triangular net which takes each point as a common peak, obtaining a second discrete index of each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface; then screening each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface according to the first discrete index and the second discrete index of each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface to obtain each target point in the point cloud corresponding to the target three-dimensional image corresponding to each surface; finally, according to the coordinates of each target point, constructing and obtaining a feature vector corresponding to each target point; and judging whether the damage of each surface of the seven-stringed plucked instrument to be detected occurs or not according to the characteristic vector. The detection method for detecting the damage of the surface of the guqin has high efficiency and accuracy, and is a detection and identification method with high intellectualization and automation.
Preferably, fitting the point cloud with a three-dimensional curved surface to obtain a fitted curved surface corresponding to the target three-dimensional image; the method for calculating the distance from each point in the point cloud corresponding to the target three-dimensional image to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image comprises the following steps:
fitting the point cloud corresponding to the target three-dimensional image by using a least square method to obtain a fitting curved surface corresponding to the target three-dimensional image;
calculating the shortest distance from each point in the point cloud corresponding to the target three-dimensional image to the corresponding curved surface;
and recording the shortest distance as a first discrete index of each point in the point cloud corresponding to the target three-dimensional image.
Preferably, constructing a triangulation network of the point cloud; the method for obtaining the second discrete indexes of each point in the point cloud corresponding to the target three-dimensional image according to the area of each triangle taking each point as a common vertex in the triangulation network comprises the following steps:
constructing a triangular network of point clouds corresponding to the target three-dimensional image by using a ball pivot algorithm;
acquiring the area of each triangle taking each point as a common vertex in the triangular net;
for any point in the point cloud corresponding to the target three-dimensional image:
and acquiring the area sum of each triangle taking the point as a common vertex, and acquiring a second discrete index of the point according to the area sum of each triangle taking the point as the common vertex.
Preferably, the second discrete indicator of the point is calculated according to the following formula:
W＝e -S1
where W is the second discrete index of the point, and S1 is the sum of areas of triangles having the point as a common vertex.
Preferably, the method for obtaining each target point in the point cloud corresponding to the target three-dimensional image by screening each point in the point cloud corresponding to the target three-dimensional image according to the first discrete index of each point and the second discrete index of each point includes:
carrying out mesh division on the point cloud corresponding to the target three-dimensional image to obtain each point in each mesh corresponding to the target three-dimensional image; for each point in any grid corresponding to the target three-dimensional image:
acquiring the central position of the grid, calculating the distance from each point in the grid to the corresponding central position, and selecting the point with the minimum distance as a central point; marking other points except the central point in the grid as characteristic points in the grid;
calculating the distance from each characteristic point in the grid to the corresponding central point, and arranging the characteristic points in the grid from small to large to obtain a distance sequence corresponding to the grid;
obtaining the sequence of each feature point in the grid according to the distance sequence corresponding to the grid;
calculating the square of the depth value difference between two adjacent order feature points in the grid;
obtaining a third discrete index of each feature point in the grid according to the square of the depth value difference between two adjacent feature points in the grid;
obtaining the importance degree of the grid according to the first discrete index, the second discrete index and the third discrete index of each feature point in the grid;
judging whether the importance degree of the grid is greater than a preset threshold value, if so, marking each point in the grid as a target point; otherwise, the center point in the mesh is marked as the target point.
Preferably, the importance of the grid is calculated according to the following formula:
Figure GDA0003937517190000031
wherein B is the importance of the grid, J is the number of feature points in the grid, D j Is the first discrete index of the jth characteristic point in the grid, W j A second discrete index, S, for the jth feature point in the grid j And a is a third discrete index of the jth characteristic point in the grid, alpha is the weight of the first discrete index, beta is the weight of the second discrete index, and gamma is the weight of the third discrete index.
Preferably, a feature vector corresponding to each target point is constructed and obtained according to the coordinates of each target point; according to the characteristic vector, the method for judging whether the damage occurs to each surface of the guqin to be detected comprises the following steps:
for any target point in the point cloud corresponding to the target three-dimensional image:
acquiring a three-dimensional coordinate of the target point;
projecting the three-dimensional coordinates of the target point on three planes respectively to obtain the projection coordinates of the target point on the three planes respectively;
then, fourier transformation is carried out on the projection coordinates of the target point on the three surfaces to obtain three Fourier descriptors corresponding to the target point;
constructing and obtaining a feature vector corresponding to the target point according to the three Fourier descriptors corresponding to the target point;
acquiring a standard characteristic vector of the target point corresponding to the standard guqin;
calculating cosine similarity between the characteristic vector corresponding to the target point and the corresponding standard characteristic vector; and judging whether the cosine similarity is smaller than a preset similarity threshold value, if so, judging that the possibility that the surface area of the guqin to be detected corresponding to the target point is damaged is higher, and otherwise, judging that the possibility that the surface area of the guqin to be detected corresponding to the target point is damaged is lower.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flow chart of a method for detecting and identifying damaged surface material of a guqin based on optical means.
Detailed Description
In the following, the technical solutions in the embodiments of the present invention will be clearly and completely described with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments, and all other embodiments obtained by a person skilled in the art based on the embodiments of the present invention belong to the protection scope of the embodiments of the present invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment provides a method for detecting and identifying damage of a surface material of a guqin based on an optical means, which is described in detail as follows:
as shown in fig. 1, the method for detecting and identifying the damage of the surface material of the guqin based on the optical means comprises the following steps:
s001, acquiring a target three-dimensional image of each surface of the guqin to be detected; and converting the target three-dimensional images corresponding to the surfaces into point clouds to obtain the point clouds corresponding to the target three-dimensional images.
Because the traditional detection means is difficult to detect the damage which is difficult to observe or small, the embodiment utilizes laser to accurately construct three-dimensional images of the surfaces of the guqin, then utilizes filtering to remove disordered noise points, namely useless points caused by objective conditions, then screens point clouds corresponding to the three-dimensional images based on a plurality of angles, selects points with larger description importance degree on the surfaces of the guqin, and removes some repeated or less important points, so that the three-dimensional outline of the outer surface of the guqin can be more accurately described, and the data volume for analyzing whether the damage occurs on the surface of the guqin in the follow-up process can be reduced; finally, the characteristics of the points with larger importance degree are used as a basis for judging whether the surfaces of the guqin are damaged or not; the method for detecting the damage of the surface of the guqin is high in efficiency and accuracy, and is an intelligent and automatic detection and identification method.
In the embodiment, a line structure light detection device is used for detecting and obtaining three-dimensional images of all surfaces of the guqin to be detected, and the three-dimensional images can reflect depth information of the surface of the guqin; the line-structured light detection device mainly comprises a structured light projector, a CCD (charge coupled device) camera and a storage platform, wherein the line-structured light projector is used for projecting surface structured light to the surface of the guqin, the surface structured light can form a line-structured light bar on the surface of the guqin, and the CCD camera is used for capturing the line-structured light bar on the surface of the guqin; and when detecting, the guqin is placed on the object placing platform, and then the scanning of the line structured light to the surface of the guqin is realized through the movement of the object placing platform.
When the three-dimensional image is obtained by line structured light scanning, the influence of external factors such as human factors can be caused, so that the obtained three-dimensional image has noise and data redundancy, and the subsequent processing analysis can be influenced; therefore, in the embodiment, the double filtering processing is performed on the three-dimensional image of each surface of the guqin to be detected, and because the line structured light is used for scanning the outer surface of the guqin, noise only appears on the outer surface, the noise is removed by using a semicircular edge detection algorithm, and because the method cannot completely detect large-scale noise points, the large-scale noise is removed by using a point cloud fuzzy C-means clustering method; and recording the three-dimensional image after the double filtering processing as a target three-dimensional image of each surface of the guqin to be detected. Converting the target three-dimensional images of the surfaces of the guqin to be detected into point clouds to obtain the point clouds corresponding to the target three-dimensional images of the surfaces of the guqin to be detected; the point cloud corresponding to the three-dimensional image is a known technology, and therefore the detailed description of the embodiment is omitted.
In this embodiment, in order to reduce the amount of calculation and describe the three-dimensional profile of each surface of the guqin to be detected more accurately, a point with a greater degree of importance for the description of each surface of the guqin needs to be selected for analysis, and since a single analysis angle cannot better reflect the degree of importance of each point in the point cloud corresponding to the target three-dimensional image of each surface, a plurality of angles need to be selected for measurement and analysis.
S002, fitting the point cloud with a three-dimensional curved surface to obtain a fitted curved surface corresponding to the target three-dimensional image; and calculating the distance from each point in the point cloud corresponding to the target three-dimensional image to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image.
In the embodiment, the distance from each point cloud corresponding to the target three-dimensional image of each surface to the corresponding fitting curved surface is analyzed to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image of each surface; the specific process is as follows:
fitting the point clouds corresponding to the target three-dimensional images of the surfaces by using a least square method to obtain a fitting curved surface corresponding to the target three-dimensional images of the surfaces; then calculating to obtain the shortest distance from each point in the point cloud corresponding to the target three-dimensional image of each surface to the corresponding curved surface; recording the shortest distance from each point in the point cloud corresponding to the target three-dimensional image of each surface to the corresponding curved surface as a first discrete index of each point in the point cloud corresponding to the target three-dimensional image of each surface; the larger the shortest distance from any point in the point cloud corresponding to the target three-dimensional image to the corresponding curved surface is, the higher the possibility that the corresponding point to the surface area of the guqin to be detected is broken is, namely the more important the point is, otherwise, the less important the point is.
In the embodiment, the method for obtaining the curved surface by fitting by using the least square method is the prior art, and therefore, detailed description is omitted; the method for calculating the shortest distance from a point to a plane in the present embodiment is a lagrangian multiplier method, which is a prior art, and therefore, will not be described in detail.
S003, constructing a triangular net of the point cloud; and obtaining second discrete indexes of each point in the point cloud corresponding to the target three-dimensional image according to the area of each triangle taking each point as a common vertex in the triangular net.
In this embodiment, a triangulation network of point clouds corresponding to the target three-dimensional images of the surfaces is analyzed to obtain second discrete indexes of each point in the point clouds corresponding to the target three-dimensional images of the surfaces; the method comprises the following specific steps:
constructing a triangulation network of point clouds corresponding to the target three-dimensional images of the surfaces by using a Ball Pivot Algorithm (BPA), wherein the ball pivot algorithm is a surface reconstruction method related to an alpha shape; acquiring the area of each triangle in the triangular net by taking each point as a common vertex; for any point in the point cloud corresponding to the target three-dimensional image of each surface, acquiring the area sum of each triangle taking the point as a common vertex in the corresponding triangular net, and acquiring a second discrete index of the point according to the area sum of each triangle taking the point as a common vertex in the corresponding triangular net; calculating a second discrete index for the point according to the following formula:
W＝e -S1
wherein, W is a second discrete index of the point, and S1 is the area sum of each triangle with the point as a common vertex; the larger W is, the higher the possibility that the corresponding point has damage to the surface area of the guqin to be detected is, namely, the more important the point is, otherwise, the less important the point is; because the triangular net can present the characteristics of dense triangles and small area of a single triangle at the position where the surface structure changes, the surface structure changes comprise surface protrusions, depressions or damages, and the characteristics of large area and dispersed distribution of triangles are presented at the position where the surface structure changes, the smaller S1 is, and the larger W is.
Therefore, the second discrete indexes of each point in the point cloud corresponding to the target three-dimensional image of each surface can be obtained through the method.
And step S004, screening each point in the point cloud corresponding to the target three-dimensional image according to the first discrete index of each point and the second discrete index of each point to obtain each target point in the point cloud corresponding to the target three-dimensional image.
In this embodiment, each target point in the point cloud corresponding to the target three-dimensional image of each surface is obtained by analyzing the first discrete index and the second discrete index of each point in the point cloud corresponding to the target three-dimensional image of each surface; the specific process is as follows:
firstly, carrying out mesh division on point clouds corresponding to target three-dimensional images of all surfaces to obtain points in all meshes corresponding to the target three-dimensional images of all surfaces; the mesh division is carried out for screening each point in the point cloud in the follow-up process; the size of the grid needs to be set according to the size of the surface of the guqin to be detected and the screening precision. For each point in any grid corresponding to the target three-dimensional image of each surface:
acquiring the central position of the grid, calculating the distance from each point in the grid to the corresponding central position, and selecting the point with the minimum distance as a central point; marking other points except the central point in the grid as characteristic points in the grid; then calculating the distance from each characteristic point in the grid to the corresponding central point, and arranging the characteristic points in the grid from small to large to obtain a distance sequence corresponding to the grid; marking each feature point in the grid according to the distance sequence corresponding to the grid, setting the sequence of the feature point corresponding to the first parameter in the distance sequence as 1, and setting the sequence of the feature point corresponding to the second parameter in the distance sequence as 2, and so on to obtain the sequence of each feature point in the grid; then, the square of the difference in depth values between two feature points in the adjacent order in the mesh is calculated, and the square of the difference in depth values between two feature points in the adjacent order is recorded as the third discrete index of the feature point in the later order, for example, the square of the difference between the depth value of the feature point in the mesh with order 3 and the depth value of the feature point in the order 2 is recorded as the third discrete index of the feature point in the mesh with order 3, and then the square of the difference in depth values between the depth value of the feature point in the mesh with order 1 and the depth value of the center point is recorded as the third discrete index of the point in the mesh with order 1.
Therefore, the third discrete index of each feature point in the grid can be obtained through the process; the larger the third discrete index is, the higher the possibility that the corresponding point is damaged in the surface area of the guqin to be detected is, namely, the greater the importance degree of the point is, otherwise, the smaller the importance degree of the point is; then normalizing the first discrete index, the second discrete index and the third discrete index of each feature point in the grid; obtaining the importance degree of the grid according to the first discrete index, the second discrete index and the third discrete index of each feature point in the grid after normalization; the importance of the grid is calculated according to the following formula:
Figure GDA0003937517190000081
wherein B is the importance of the grid, J is the number of feature points in the grid, D j Is a first discrete index, W, of the jth characteristic point in the grid after normalization j Is a second discrete index of the j-th characteristic point in the grid after normalization, S j After normalization, a is a third discrete index of the jth characteristic point in the grid, alpha is the weight of the first discrete index, beta is the weight of the second discrete index, and gamma is the weight of the third discrete index; the larger the B is, the higher the importance degree of the grid is, and the higher the possibility that the corresponding surface of the guqin to be detected is damaged is; and the values of α, β, and γ need to be set according to actual conditions, but the sum of α, β, and γ is 1.
In this embodiment, the importance degree of each grid corresponding to the target three-dimensional image of each surface can be obtained through the above process; and then, carrying out normalization processing on the importance degree of each grid corresponding to the target three-dimensional image of each surface to obtain the target importance degree of each grid corresponding to the target three-dimensional image of each surface.
The target importance degree of any grid corresponding to the target three-dimensional image of each surface is as follows: judging whether the target importance degree of the grid is greater than a preset threshold value, if so, recording each point in the grid as a target point; otherwise, recording the central point in the grid as a target point, and eliminating other points except the central point. The preset threshold needs to be set according to actual conditions, and may be, for example, 0.27; and when the target importance degree of the grid is greater than a preset threshold value, points in the grid are discrete and fluctuated.
Therefore, the present embodiment can obtain each target point in the point cloud corresponding to the target three-dimensional image of each surface through the above process; and only the target point is analyzed subsequently, so that the calculation amount can be reduced.
Step S005, constructing and obtaining a characteristic vector corresponding to each target point according to the coordinates of each target point; and judging whether the damage occurs on each surface of the guqin to be detected according to the characteristic vector.
In this embodiment, feature vectors corresponding to target points in the point clouds corresponding to the target three-dimensional images of the surfaces are constructed by analyzing the target points in the point clouds corresponding to the target three-dimensional images of the surfaces; judging whether the surfaces of the guqin to be detected are damaged or not based on the characteristic vectors corresponding to the target points in the point cloud corresponding to the target three-dimensional images of the surfaces; the specific process is as follows:
because the Fourier transform has rotation invariance and size invariance, the damage of the surface of the guqin can be accurately detected when the scanned three-dimensional model of the surface of the guqin has position difference; thus for any target point in the point cloud corresponding to the target three-dimensional image for each surface:
acquiring a three-dimensional coordinate A (x 1, y1, z 1) of the target point, wherein x1 is an abscissa of the target point, y1 is a ordinate of the target point, and z1 is a vertical coordinate of the target point; projecting the three-dimensional coordinates of the target point to an xy plane, a yz plane and an xz plane respectively to obtain the projection coordinates of the target point on three surfaces respectively, and then performing Fourier transform on the projection coordinates of the target point on the three surfaces to obtain three Fourier descriptors corresponding to the target point; constructing and obtaining a characteristic vector corresponding to the target point according to the three Fourier descriptors corresponding to the target point; obtaining a standard characteristic vector of the target point corresponding to the standard guqin according to the mode; calculating cosine similarity between the characteristic vector corresponding to the target point and the corresponding standard characteristic vector; and judging whether the cosine similarity is smaller than a preset similarity threshold value, if so, judging that the possibility that the surface area of the guqin to be detected corresponding to the target point is damaged is higher, otherwise, judging that the possibility that the surface area of the guqin to be detected corresponding to the target point is damaged is lower.
Has the beneficial effects that: the method comprises the steps of obtaining a target three-dimensional image of each surface of the guqin to be detected by utilizing line structured light; fitting the three-dimensional curved surface of the point cloud corresponding to the target three-dimensional image corresponding to each surface to obtain the fitted curved surface of the target three-dimensional image corresponding to each surface; then calculating the distance from each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface; then constructing a triangular network of point clouds corresponding to the target three-dimensional images corresponding to the surfaces; according to the area of each triangle in the triangular net with each point as a common vertex, obtaining a second discrete index of each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface; then screening each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface according to the first discrete index and the second discrete index of each point in the point cloud corresponding to the target three-dimensional image corresponding to each surface to obtain each target point in the point cloud corresponding to the target three-dimensional image corresponding to each surface; finally, according to the coordinates of each target point, constructing and obtaining a feature vector corresponding to each target point; and judging whether the damage of each surface of the seven-stringed plucked instrument to be detected occurs or not according to the characteristic vector. The detection method for detecting the damage to the surface of the guqin is high in efficiency and accuracy, and is intelligent and high in automation.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present application and are intended to be included within the scope of the present application.

Claims (3)

1. A detection and identification method for surface material breakage of a guqin based on an optical means is characterized by comprising the following steps:
acquiring target three-dimensional images of all surfaces of the guqin to be detected; converting the target three-dimensional images corresponding to the surfaces into point clouds to obtain point clouds corresponding to the target three-dimensional images;
fitting the point cloud with a three-dimensional curved surface to obtain a fitted curved surface corresponding to the target three-dimensional image; calculating the distance from each point in the point cloud corresponding to the target three-dimensional image to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image;
constructing a triangulation of the point cloud; obtaining second discrete indexes of each point in the point cloud corresponding to the target three-dimensional image according to the area of each triangle taking each point as a common vertex in the triangular net;
screening each point in the point cloud corresponding to the target three-dimensional image according to the first discrete index of each point and the second discrete index of each point to obtain each target point in the point cloud corresponding to the target three-dimensional image;
constructing and obtaining a characteristic vector corresponding to each target point according to the coordinates of each target point; judging whether the surface of the seven-stringed plucked instrument to be detected is damaged or not according to the characteristic vector;
fitting the point cloud with a three-dimensional curved surface to obtain a fitted curved surface corresponding to the target three-dimensional image; the method for calculating the distance from each point in the point cloud corresponding to the target three-dimensional image to the corresponding fitting curved surface to obtain a first discrete index of each point in the point cloud corresponding to the target three-dimensional image comprises the following steps:
fitting the point cloud corresponding to the target three-dimensional image by using a least square method to obtain a fitting curved surface corresponding to the target three-dimensional image;
calculating the shortest distance from each point in the point cloud corresponding to the target three-dimensional image to the corresponding curved surface;
recording the shortest distance as a first discrete index of each point in the point cloud corresponding to the target three-dimensional image;
the construction of a triangulation of the point cloud; the method for obtaining the second discrete indexes of each point in the point cloud corresponding to the target three-dimensional image according to the area of each triangle taking each point as a common vertex in the triangulation network comprises the following steps:
constructing a triangular network of point clouds corresponding to the target three-dimensional image by using a ball pivot algorithm;
acquiring the area of each triangle in the triangular net by taking each point as a common vertex;
for any point in the point cloud corresponding to the target three-dimensional image:
acquiring the area sum of each triangle taking the point as a common vertex, and acquiring a second discrete index of the point according to the area sum of each triangle taking the point as the common vertex;
the method for screening each point in the point cloud corresponding to the target three-dimensional image according to the first discrete index of each point and the second discrete index of each point to obtain each target point in the point cloud corresponding to the target three-dimensional image comprises the following steps:
carrying out mesh division on the point cloud corresponding to the target three-dimensional image to obtain each point in each mesh corresponding to the target three-dimensional image; for each point in any grid corresponding to the target three-dimensional image:
acquiring the central position of the grid, calculating the distance from each point in the grid to the corresponding central position, and selecting the point with the minimum distance as a central point; marking other points except the central point in the grid as characteristic points in the grid;
calculating the distance from each characteristic point in the grid to the corresponding central point, and arranging the characteristic points in the grid from small to large to obtain a distance sequence corresponding to the grid;
obtaining the sequence of each feature point in the grid according to the distance sequence corresponding to the grid;
calculating the square of the depth value difference between two adjacent sequence feature points in the grid;
obtaining a third discrete index of each feature point in the grid according to the square of the depth value difference between two adjacent feature points in the grid;
obtaining the importance degree of the grid according to the first discrete index, the second discrete index and the third discrete index of each feature point in the grid;
judging whether the importance degree of the grid is greater than a preset threshold value, if so, recording each point in the grid as a target point; otherwise, recording the central point in the grid as a target point;
constructing and obtaining a characteristic vector corresponding to each target point according to the coordinates of each target point; according to the characteristic vector, the method for judging whether the damage occurs to each surface of the guqin to be detected comprises the following steps:
for any target point in the point cloud corresponding to the target three-dimensional image:
acquiring a three-dimensional coordinate of the target point;
projecting the three-dimensional coordinates of the target point on three planes respectively to obtain the projection coordinates of the target point on the three planes respectively;
then, fourier transformation is carried out on the projection coordinates of the target point on the three surfaces to obtain three Fourier descriptors corresponding to the target point;
constructing and obtaining a characteristic vector corresponding to the target point according to the three Fourier descriptors corresponding to the target point;
acquiring a standard characteristic vector of the target point corresponding to the standard guqin;
calculating cosine similarity between the characteristic vector corresponding to the target point and the corresponding standard characteristic vector; and judging whether the cosine similarity is smaller than a preset similarity threshold value, if so, judging that the possibility that the surface area of the guqin to be detected corresponding to the target point is damaged is higher, and otherwise, judging that the possibility that the surface area of the guqin to be detected corresponding to the target point is damaged is lower.
2. The method of claim 1, wherein the second discrete index of the point is calculated according to the following formula:
W＝e -S1
where W is the second discrete index of the point, and S1 is the sum of areas of triangles having the point as a common vertex.
3. The method of claim 1, wherein the importance of the grid is calculated according to the following formula:
Figure FDA0003937517180000031
wherein B is the importance of the grid, J is the number of feature points in the grid, D j Is the first discrete index of the jth characteristic point in the grid, W j A second discrete index, S, for the jth feature point in the grid j And a is a third discrete index of the jth characteristic point in the grid, alpha is the weight of the first discrete index, beta is the weight of the second discrete index, and gamma is the weight of the third discrete index.
CN202211151374.4A 2022-09-21 2022-09-21 Guqin surface material damage detection and identification method based on optical means Active CN115236092B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211151374.4A CN115236092B (en) 2022-09-21 2022-09-21 Guqin surface material damage detection and identification method based on optical means

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211151374.4A CN115236092B (en) 2022-09-21 2022-09-21 Guqin surface material damage detection and identification method based on optical means

Publications (2)

Publication Number Publication Date
CN115236092A CN115236092A (en) 2022-10-25
CN115236092B true CN115236092B (en) 2022-12-13

Family

ID=83681982

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211151374.4A Active CN115236092B (en) 2022-09-21 2022-09-21 Guqin surface material damage detection and identification method based on optical means

Country Status (1)

Country Link
CN (1) CN115236092B (en)

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102853803B (en) * 2012-08-08 2015-05-27 北京建筑大学 Testing method of damaged area of cultural relic
CN107862745B (en) * 2017-10-25 2021-04-09 武汉楚锐视觉检测科技有限公司 Reflective curved surface three-dimensional reconstruction labeling method and device
CN113008195B (en) * 2021-02-01 2022-04-22 厦门大学 Three-dimensional curved surface distance measuring method and system based on space point cloud
CN113916130B (en) * 2021-12-15 2022-02-25 天津风霖物联网科技有限公司 Building position measuring method based on least square method

Also Published As

Publication number Publication date
CN115236092A (en) 2022-10-25

Similar Documents

Publication Publication Date Title
CN113705478B (en) Mangrove single wood target detection method based on improved YOLOv5
CN112818988B (en) Automatic identification reading method and system for pointer instrument
CN110544233B (en) Depth image quality evaluation method based on face recognition application
CN109190473A (en) The application of a kind of " machine vision understanding " in remote monitoriong of electric power
CN108711148B (en) Tire defect intelligent detection method based on deep learning
CN106023185A (en) Power transmission equipment fault diagnosis method
CN111144325A (en) Fault identification and positioning method, device and equipment for power equipment of transformer substation
CN114749342B (en) Lithium battery pole piece coating defect identification method, device and medium
CN110057828A (en) A kind of surface inspecting method and system based on machine vision
CN113688817A (en) Instrument identification method and system for automatic inspection
CN110634137A (en) Bridge deformation monitoring method, device and equipment based on visual perception
CN115239625B (en) Top cover welding spot cloud defect detection method, device, equipment and storage medium
CN103884294B (en) The method and its device of a kind of infrared light measuring three-dimensional morphology of wide visual field
CN113313107A (en) Intelligent detection and identification method for multiple types of diseases on cable surface of cable-stayed bridge
CN114419140A (en) Positioning algorithm for light spot center of track laser measuring device
CN113705564A (en) Pointer type instrument identification reading method
CN115236092B (en) Guqin surface material damage detection and identification method based on optical means
CN115597494A (en) Precision detection method and system for prefabricated part preformed hole based on point cloud
CN115690468A (en) Face image similarity comparison optimization method and system
CN113029103B (en) Inclination measuring method and system for foundation ring of wind turbine tower and storage medium
CN115375875A (en) Grain pile concave-convex shape feature recognition method, equipment and medium based on three-dimensional shape
CN112361989B (en) Method for calibrating parameters of measurement system through point cloud uniformity consideration
CN114092415A (en) Large-size light guide plate defect visual detection method
CN115330832A (en) Computer vision-based transmission tower full-freedom displacement monitoring system and method
CN115393387A (en) Building displacement monitoring method and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant