CN111754385B - Data point model processing method and system, detection method and system and readable medium - Google Patents
Data point model processing method and system, detection method and system and readable medium Download PDFInfo
- Publication number
- CN111754385B CN111754385B CN201910234448.2A CN201910234448A CN111754385B CN 111754385 B CN111754385 B CN 111754385B CN 201910234448 A CN201910234448 A CN 201910234448A CN 111754385 B CN111754385 B CN 111754385B
- Authority
- CN
- China
- Prior art keywords
- model
- initial
- grid
- image
- data points
- 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
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 79
- 238000003672 processing method Methods 0.000 title claims abstract description 36
- 238000012545 processing Methods 0.000 claims abstract description 64
- 238000000034 method Methods 0.000 claims description 25
- 238000013461 design Methods 0.000 claims description 12
- 230000003287 optical effect Effects 0.000 claims description 12
- 125000004122 cyclic group Chemical group 0.000 claims description 10
- 238000002372 labelling Methods 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims 3
- 230000003252 repetitive effect Effects 0.000 claims 1
- 238000004364 calculation method Methods 0.000 description 3
- 239000006185 dispersion Substances 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000012780 transparent material Substances 0.000 description 2
- 230000001351 cycling effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000003754 machining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/08—Projecting images onto non-planar surfaces, e.g. geodetic screens
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention provides a data point model processing method and a system thereof, a detection method and a system thereof and a readable medium, wherein the data point model processing method comprises the following steps: dividing a model to be processed into a plurality of grids; marking each grid, and marking the model to be processed as a binary image to form a grid image; judging the connected domain of the grid image, and acquiring an additional grid image when the grid image comprises a separated main grid image and the additional grid image; and removing initial data points corresponding to the data points in the additional grid image from the initial model. The data point model processing method can accelerate the processing speed, thereby increasing the detection speed.
Description
Technical Field
The present invention relates to the field of data point model processing, and in particular, to a data point model processing method and system, a detection method and system, and a readable medium.
Background
With the development of modern industry, precision machining is used in more and more fields; meanwhile, the processing precision is also required to be higher and higher. In order to meet the requirement of processing precision and improve the qualification rate of processed samples, the processing process and the processed products need to be tested for morphology distortion frequently so as to ensure that the distortion is within a tolerable range.
The existing distortion detection methods can be classified into a contact detection method and a non-contact detection method. In a contact detection method, for example, three-coordinate detection, a probe needs to be in contact with an object to be detected during detection, so that the object to be detected is easily damaged. The non-contact detection method, including optical detection methods such as binocular vision, chromatic dispersion confocal and structured light detection, is not in contact with the object to be detected, can reduce damage and distortion of the object to be detected, and is increasingly used.
In the process of detecting the distortion of the object to be detected by the optical detection method, the surface of the object to be detected needs to be scanned, and the three-dimensional point cloud is produced. In the scanning process, objects around the object to be detected are easy to detect, so that the acquired three-dimensional point cloud comprises detection information of the object to be detected and additional detection information of the objects around the object to be detected. The additional detection information is easily regarded as a part of the object to be detected, resulting in false detection.
The method for removing the additional detection information in the prior art is complex, so that the detection efficiency is low.
Disclosure of Invention
The invention solves the problem of providing a data point model processing method which can improve the data point model processing speed, thereby improving the detection speed and reducing the false detection rate.
In order to solve the above problems, the present invention provides a data point model processing method, including: providing an initial model, the initial model comprising a main model and an additional model that are separate from each other, the main model and the additional model each comprising initial data points; forming a model to be processed according to an initial model, wherein the initial data points form data points in the model to be processed; dividing a model to be processed into a plurality of grids; marking each grid, and marking the model to be processed as a binary image to form a grid image; judging the connected domain of the grid image, and acquiring an additional grid image when the grid image comprises a separated main grid image and the additional grid image; and removing initial data points corresponding to the data points in the additional grid image from the initial model to form a target model.
Optionally, the initial model is a three-dimensional point cloud; the model to be processed is a two-dimensional image; the step of forming the model to be processed from the initial model includes: providing a projection direction; and carrying out projection processing on the initial model along a projection direction to form the model to be processed.
Optionally, the number of projection directions is multiple, and the data point model processing method further includes: and repeating the cyclic operation processing until additional grid images of all additional models are acquired, wherein the cyclic operation processing comprises the step of judging from the projection processing to the connected domain.
Optionally, the number of the additional models is a plurality, and the repeated cyclic operation processing further includes: removing initial data points corresponding to the data points in the additional grid image from the initial model; or repeating the cyclic operation until additional grid images of all the additional models are acquired, and removing initial data points corresponding to the data points in all the additional grid images from the initial models.
Optionally, the step of forming the model to be processed from the initial model includes: and enabling the model to be processed to be identical to an initial model, wherein the data points are identical to the initial data points.
Optionally, the step of marking includes: setting a first quantity threshold; comparing the number of data points in the grid with the first quantity threshold, and marking the grid as an image point when the number of data points in the grid is larger than or equal to the first quantity threshold; otherwise, the grid is marked as blank.
Optionally, the initial model is an image; the initial data points comprise gray values of pixels in an initial model; the marking process includes the steps of: setting a second number threshold; obtaining a reference number according to the number of data points meeting gray scale conditions in the grid; marking the grid as an image point when the reference number is greater than or equal to a second number threshold; otherwise, marking the grid as blank spots; or the step of marking comprises: setting a first gray threshold; acquiring a grid gray value according to the gray value of each data point in the grid; and comparing the grid gray value with a first gray threshold value, and marking grids meeting different comparison results as image points and blank points respectively.
Correspondingly, the technical scheme of the invention also provides a detection method, which comprises the following steps: detecting an object to be detected to obtain an initial model; and processing the initial model according to a data point model processing method to form a target model.
Optionally, the method further comprises: providing a design model; and comparing the target model with the design model to obtain the distortion of the object to be detected.
Optionally, the object to be detected is detected by a detection device, wherein the detection device comprises an objective lens; the projection direction includes: a first projection direction perpendicular to an optical axis of the objective; and/or a second projection direction parallel to the optical axis of the objective lens.
The technical scheme of the invention also provides a data point model processing system, which comprises: an input system for providing an initial model, the initial model comprising a plurality of initial data points, the initial model comprising a main model and an additional model that are separate from each other; the data processing system is used for forming a to-be-processed model according to an initial model, and the initial data points form data points in the to-be-processed model; the grid dividing system is used for dividing the model to be processed into a plurality of grids; the marking system is used for marking each grid, marking the model to be processed as a binary image and forming a grid image; the connected domain judging system is used for judging the connected domain of the grid image, and acquiring an additional grid image when the grid image comprises a separated main grid image and the additional grid image; and the removing system is used for removing initial data points corresponding to the data points in the additional grid image from the initial model to form a target model.
Optionally, the initial model is a three-dimensional point cloud; the model to be processed is a two-dimensional image; the data processing system includes: and the projection system is used for carrying out projection processing on the initial model along the projection direction to form the model to be processed.
Optionally, the data processing system comprises: and the equivalent system is used for enabling the model to be processed to be identical to an initial model, and the data points are identical to the initial data points.
Optionally, the marking system includes: a setting system for setting a first number threshold; a comparison system for comparing the number of data points in the grid with the first quantity threshold, and marking the grid as an image point when the number of data points in the grid is greater than or equal to the first quantity threshold; otherwise, the grid is marked as blank.
The technical scheme of the invention provides a detection system, which comprises: the model acquisition system is used for detecting the object to be detected and acquiring an initial model; and the data point model processing system is used for carrying out data point model processing on the initial model to form a target model.
Optionally, the input system is further configured to provide a design model; the detection system further comprises: and the distortion detection system is used for comparing the target model with the design model to obtain the distortion of the object to be detected.
The present disclosure also provides a computer readable medium comprising executable instructions that, when executed, cause a processor to perform a data point model processing method.
Compared with the prior art, the technical scheme of the invention has the following advantages:
In the data point model processing method provided by the technical scheme of the invention, the model to be processed is divided into a plurality of grids, each grid is marked, and the initial model is marked as a binary image to form a grid image. After the initial model is marked as a binary image, an additional grid image can be obtained by a connected domain judging method, so that the algorithm of data point model processing can be improved; in addition, as the grids can comprise a plurality of data points, the number of the data points of each grid image can be reduced, so that the complexity of the judgment of the connected domain is simplified, the speed of the judgment of the connected domain is increased, and the processing speed of the data point model is further increased.
Further, the method comprises the steps of providing a projection direction, and carrying out projection processing on the initial model along the projection direction to form the model to be processed. The projection processing can convert the three-dimensional point cloud into a two-dimensional model to be processed, and the operation can be simplified by processing the two-dimensional model to be processed subsequently, so that the processing speed of the data point model is improved.
According to the detection method provided by the technical scheme of the invention, through the data point model processing method, the initial data points corresponding to the data points in the additional grid image are removed from the initial model, so that the calculation speed can be increased, and the false detection can be reduced.
Drawings
FIG. 1 is a flow chart illustrating steps of an embodiment of a data point model processing method according to the present invention;
Fig. 2 to 5 are schematic structural diagrams of steps in an embodiment of a data point model processing method according to the present invention.
Detailed Description
The detection method has a number of problems, such as: the data point model is processed at a slower speed and with lower efficiency.
In order to solve the technical problem, the invention provides a data point model processing method, which comprises the following steps: dividing a model to be processed into a plurality of grids; marking each grid, and marking the model to be processed as a binary image to form a grid image; judging the connected domain of the grid image, and acquiring an additional grid image when the grid image comprises a separated main grid image and the additional grid image; and removing initial data points corresponding to the data points in the additional grid image from the initial model. The data point model processing method can accelerate the processing speed, thereby increasing the detection speed. The method can increase the processing speed, thereby increasing the detection speed.
In order that the above objects, features and advantages of the invention will be readily understood, a more particular description of the invention will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings.
FIG. 1 is a flow chart illustrating steps in an embodiment of a data point model processing method according to the present invention.
Step S01, providing an initial model, wherein the initial model comprises a main model and an additional model which are mutually separated, and the main model and the additional model comprise initial data points;
Step S02, forming a model to be processed according to an initial model, wherein the initial data points form data points in the model to be processed;
step S03, dividing a model to be processed into a plurality of grids;
step S04, marking each grid, and marking the model to be processed as a binary image to form a grid image;
step S05, judging a connected domain of the grid image, and acquiring an additional grid image when the grid image comprises a separated main grid image and the additional grid image;
and step S06, removing initial data points corresponding to the additional grid images from the initial model to form a target model.
The data point model processing method of the invention is described in detail below with reference to the accompanying drawings.
Fig. 2 to 5 are schematic structural diagrams of steps in an embodiment of a data point model processing method according to the present invention.
The following describes a data point model processing method according to the technical scheme of the present invention with reference to fig. 2 to 5.
Referring to fig. 2, step S01 is performed to provide an initial model including a main model 100 and an additional model 110 separated from each other; the main model 100 and the additional model 110 each include initial data points.
In this embodiment, the initial model is a three-dimensional point cloud. Specifically, the initial model is a three-dimensional point cloud obtained by three-dimensionally detecting an object to be detected through a three-dimensional detection device, and the three-dimensional detection device comprises a chromatic dispersion confocal device, a laser triangulation detection device, a laser confocal device or a white light interference device. In other embodiments, the initial model may be a two-dimensional image, in particular a two-dimensional image taken by a camera or a microscope.
In this embodiment, the object to be detected is a transparent object, and the additional model 110 is a tooling image of the bottom of the object to be detected, where the tooling image is detected by the detection light of the detection device through the object to be detected.
The initial model is formed from a large number of initial data points.
In this embodiment, the initial data point is a three-dimensional coordinate point. In other embodiments, when the detection device is an imaging device, the initial data point is a pixel point of a two-dimensional image, which represents a relationship between a two-dimensional position coordinate of a surface point of the object to be detected and light intensity.
In this embodiment, each initial data point has a sequence number.
Referring to fig. 2, step S02 is performed to form a to-be-processed model 130 according to an initial model, wherein the initial data points form data points in the to-be-processed model 130.
It should be noted that, in this embodiment, the initial model is a three-dimensional point cloud; the model 130 to be processed is a two-dimensional image. The step of forming the model 130 to be processed from the initial model includes: providing a projection direction; and performing projection processing on the initial model along a projection direction to form the model 130 to be processed.
And the initial model is subjected to projection processing, so that the three-dimensional point cloud can be converted into a two-dimensional image, the calculation complexity of the subsequent connected domain judgment can be simplified, and the data point model processing speed can be improved.
The detection device comprises an objective lens and a light source, wherein the light source is used for generating detection light, and the objective lens is used for collecting the detection light and enabling the detection light to be incident on the surface of the object to be detected. In this embodiment, the detection device is a dispersive confocal device, and the objective lens is further configured to collect signal light returned from the surface of the object to be detected.
The number of the projection directions may be one or more.
The projection direction includes: a first projection direction perpendicular to an optical axis of the objective; and/or a second projection direction parallel to the optical axis of the objective lens.
When the number of projection directions is a plurality, the data point model processing method further comprises the following steps: repeating the projection processing until the connected domain is judged.
In this embodiment, the number of the additional models 110 is one. The projection direction comprises a first projection direction a, which is perpendicular to the optical axis of the detection device objective.
In this embodiment, the initial model is projected along the first projection direction a, so that the projection of the main model 100 along the first projection direction a and the projection of the additional model 110 along the first projection direction a can be separated from each other.
In other embodiments, when the alignment direction of the additional model and the main model is perpendicular to the optical axis of the detection device objective, the projection direction comprises a second projection direction b, which is parallel to the optical axis of the detection device objective.
It should be noted that, in the process of detecting the object to be detected by the optical detection device to obtain the initial model, the additional model 110 is located at one side of the main model 100, and the additional model 110 and the main model 100 are arranged along the first projection direction a or the second projection direction b. Therefore, projecting the initial model in the first projection direction a and the second projection direction b enables the projections of the main model 100 and the additional model to be separated from each other. Thus, in some embodiments, the steps of projecting the plurality of projection directions including the first projection direction and the second projection direction to the connected domain determination may be repeated until additional grid images of all additional models are acquired.
In other embodiments, when the initial model is a two-dimensional image, the step of forming the model to be processed from the initial model includes: and enabling the model to be processed to be identical to the initial model, wherein the data points are identical to the initial data points. When the initial model is a three-dimensional point cloud, the step of forming a model to be processed according to the initial model comprises the following steps: and enabling the model to be processed to be identical to the initial model, wherein the data points are identical to the initial data points.
In this embodiment, after the model to be processed is formed, the data points are projections of the initial data points.
The projection processing includes the steps of: and removing components of the initial data points along the projection direction to form data points, so that the initial point cloud forms a two-dimensional image to be processed.
Referring to fig. 4, step S03 is performed to divide the model to be processed into a plurality of grids.
The to-be-processed model is divided into a plurality of grids, and the initial model can be converted into the grid image 120 through subsequent marking processing, so that the pixel number of the grid image 120 can be reduced, the calculation amount of subsequent connected domain judgment is reduced, and the data point model processing speed can be improved.
In this embodiment, the grid is square. In other embodiments, the mesh may be rectangular, equilateral hexagonal, or triangular.
In this embodiment, the model to be processed is a two-dimensional image, and the grid is a two-dimensional grid. In other embodiments, the step of forming the model to be processed from the initial model includes: and enabling the model to be processed to be identical to the initial model, wherein the data points are identical to the initial data points. And if the initial model is a three-dimensional point cloud, the model to be processed is the three-dimensional point cloud, and if the grid is a three-dimensional grid.
With continued reference to fig. 4, step S04 is performed to perform a labeling process on each mesh, and label the model to be processed as a binary image to form a mesh image 120.
The initial model is converted into the grid image 120, so that the number of data points of the grid image 120 can be reduced, the calculated amount of judgment of the subsequent connected domain is reduced, and the processing speed of the data point model can be improved.
In this embodiment, the marking process is performed on each grid according to the number of data points in each grid.
Specifically, the marking process includes the steps of: setting a first quantity threshold; comparing the number of data points in the grid with the first quantity threshold, and marking the grid as an image point when the number of data points in the grid is larger than or equal to the first quantity threshold; otherwise, the grid is marked as blank.
In this embodiment, the surface of the object to be detected is scanned and detected to obtain the three-dimensional morphology of the surface of the object to be detected, so as to form an initial model. The initial data points are three-dimensional position coordinate data of the surface of the object to be detected. The additional model 110 is a fixture surface through which the detection light passes to reach the object to be detected, and the detection light is reflected by the fixture surface to return signal light, so as to form initial data of the additional model 110. Other positions do not form signal light because they are outside the field of view of the detection device or the detection light is blocked from reaching, and have no initial data points, i.e. no data points. Therefore, each grid can be subjected to a labeling process according to the number of data points, and a grid image 120 can be formed.
Specifically, in this embodiment, the image point is represented by data 1, and the blank point is represented by data "0".
If the first number threshold is too small, it is not easy to eliminate the influence of noise on the grid image 120, so that it is not easy to acquire the separated main grid image 101 and additional grid image 111 later; if the first number threshold is too large, errors in the grid image 120 from the initial model are easily increased, thereby reducing the accuracy of the data point model processing. In this embodiment, the first number threshold is 1 to 15.
In other embodiments, when the initial data point is a gray value for each pixel of a two-dimensional image. The initial data points comprise gray values of pixels in an initial model; the marking process includes the steps of: setting a second number threshold; obtaining a reference number according to the number of data points meeting the first gray scale condition in the grid; marking the grid as an image point when the reference number is greater than or equal to the second number threshold; otherwise, marking the grid as blank spots; or the step of marking comprises: setting a first gray threshold; acquiring a grid gray value according to the gray value of each data point in the grid; and comparing the grid gray value with a first gray threshold value, and marking grids meeting different comparison results as image points and blank points respectively.
Specifically, the step of marking further includes: a second gray level threshold is set. When the gray values of the main model and the additional model are both larger than the gray value of the background of the initial model, the gray condition is that the gray value in the grid is larger than or equal to a second gray threshold value; when the gray values of the main model and the additional model are smaller than the gray value of the background of the initial model, the gray condition is that the gray value in the grid is smaller than or equal to a second gray threshold value;
When the gray values of the main model and the additional model are both larger than the gray value of the background of the initial model, the step of marking the grids meeting different comparison results as image points and blank points respectively comprises the following steps: marking as an image point when the grid gray is greater than or equal to a first gray threshold; otherwise, the mark is blank. When the gray values of the main model and the additional model are smaller than the gray value of the background of the initial model, the step of marking grids meeting different comparison results as image points and blank points respectively comprises the following steps: marking as an image point when the grid gray is smaller than or equal to the first gray threshold; otherwise, the mark is blank.
The grid gray is the average value of the gray values of the data points in each grid or the sum of the gray values of the data points in each grid.
With continued reference to fig. 4, in step S05, a connected domain judgment is performed on the mesh image 120, and when the mesh image 120 includes the separated main mesh image 101 and the additional mesh image 111, the additional mesh image 111 is acquired.
By the connected domain judgment, it can be judged whether the projection direction can separate the main grid image 101 and the additional grid image 111, and the data point of the additional grid image 111 can be acquired, so that the additional model 110 in the initial model can be removed by the subsequent processing.
In this embodiment, the method for judging the connected domain includes an eight-connected algorithm and a four-connected algorithm.
In this embodiment, the model to be processed is obtained by projecting the initial model along the first projection direction. The additional model 110 is separated from the projection of the main model 100 in the first projection direction, and the main grid image 101 and the additional grid image 111 are separated from each other.
In other embodiments, a plurality of different projection directions are set, and when the obtained mesh images are one connected whole, the replacement projection directions repeat the cyclic operation process to obtain additional mesh images of all additional models, the cyclic operation process including steps S02 to S05. Specifically, when the additional model is one, the steps of S02 to S05 are repeated until additional mesh images are acquired.
Referring to fig. 5, step S06 is performed to remove, from the initial model, initial data points corresponding to data points in the additional grid image 111 (shown in fig. 4).
In this embodiment, the initial data points all have serial numbers; the data points have corresponding serial numbers.
In this embodiment, removing, from the initial model, initial data points corresponding to data points in the additional mesh image 111 includes: acquiring a serial number of data points in the additional grid image 111 as a serial number to be removed; acquiring an initial data point with the serial number to be removed as the initial data point to be removed; and removing the initial data points to be removed from the initial model to obtain a target model.
In this embodiment, the number of the additional models 110 is one, and after the additional grids are acquired, initial data points corresponding to data points in the additional grid image 111 are removed from the initial model to form a target model.
In other implementations, the number of additional models may be multiple. And repeating the cyclic operation processing until additional images of all additional point clouds are acquired, and removing initial data points corresponding to data points in all additional images from the initial point clouds. Or the cycling step further comprises: and removing initial data points corresponding to the data points in the additional grid image from the initial model.
The technical scheme of the invention also provides a detection method which comprises the following steps:
and detecting the object to be detected to obtain an initial model.
In this embodiment, the detection system is a three-dimensional detection device, and the initial model is a three-dimensional point cloud. Specifically, the three-dimensional detection device comprises a chromatic dispersion confocal device, a laser triangulation detection device, a laser confocal device or a white light interference device.
In other embodiments, the detection system may be a two-dimensional detection system, such as a two-dimensional image taken by a camera or microscope. The initial model may be a two-dimensional image.
In this embodiment, the material of the detection object is a transparent material. In other embodiments, the material of the detection object may also be a non-transparent material.
The main model 100 is a point cloud of an object to be detected; the additional model 110 is a point cloud of other objects around the object to be measured.
The model 130 to be processed is processed according to the data point model processing method described above to form the target model 140.
In this embodiment, the data point model processing method is the same as that shown in fig. 1 to 5, and will not be described in detail here.
In this embodiment, the detection method further includes: providing a design model; and comparing the target model 140 with the design model to obtain the distortion of the object to be detected.
The occurrence processing method can increase the processing speed, so that the detection speed of the object to be detected can be increased. In addition, the additional model 110 in the initial model is removed by the data point model processing method, so that the influence of the additional model 110 on the distortion detection result can be avoided, the false detection rate can be reduced, and the accuracy of the detection result can be improved.
The embodiment of the invention also provides a data point model processing system, which comprises: an input system for providing an initial model, the initial model comprising a plurality of initial data points, the initial model comprising a main model 100 and an additional model 110 separated from each other;
A data processing system for forming a model 130 to be processed from an initial model, the initial data points forming data points in the model to be processed;
A grid dividing system for dividing the model 130 to be processed into a plurality of grids;
the marking system is used for carrying out marking processing on each grid, and marking the model 130 to be processed as a binary image to form a grid image 120;
A connected domain judging system for judging the connected domain of the grid image 120, and acquiring an additional grid image 111 when the grid image 120 includes the separated main grid image 101 and the additional grid image 111;
And the removing system is used for removing initial data points corresponding to the data points in the additional grid image 111 from the initial model to form a target model 140.
The data processing system includes: and the projection system is used for carrying out projection processing on the initial model along the projection direction to form the model 130 to be processed.
In other embodiments, the data processing system includes: and the equivalent system is used for enabling the model to be processed to be identical to an initial model, and the data points are identical to the initial data points.
The marking system includes: a setting system for setting a first number threshold; and the comparison system is used for comparing the number of the data points in the grid with the first quantity threshold value, marking the grid as an image point when the number of the data points in the grid is larger than or equal to the first quantity threshold value, otherwise marking the grid as a blank point.
The technical embodiment of the invention also provides a detection system, which comprises: the model acquisition system is used for detecting the object to be detected and acquiring an initial model; a data point model processing system, the data point model processing system being the same as the data point model processing system described in the previous embodiment.
In this embodiment, the input system is further configured to provide a design model;
in this embodiment, the detection system further includes: and the distortion detection system is used for comparing the target model 140 with the design model to obtain the distortion of the object to be detected.
The present disclosure also provides a non-transitory computer readable medium comprising executable instructions that, when executed, cause a processor to perform the data point model processing method shown in fig. 1-5.
Although the present invention is disclosed above, the present invention is not limited thereto. Various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention, and the scope of the invention should be assessed accordingly to that of the appended claims.
Claims (18)
1. A method of data point model processing, comprising:
Providing an initial model, the initial model comprising a main model and an additional model that are separate from each other, the main model and the additional model each comprising initial data points; the additional model is a tool image of the bottom of the object to be detected, wherein the tool image is detected by the detection light of the detection equipment through the object to be detected;
Forming a model to be processed according to an initial model, wherein the initial data points form data points in the model to be processed;
dividing a model to be processed into a plurality of grids;
Marking each grid, and marking the model to be processed as a binary image to form a grid image;
Judging the connected domain of the grid image, and acquiring an additional grid image when the grid image comprises a separated main grid image and the additional grid image;
removing initial data points corresponding to the data points in the additional grid image from the initial model to form a target model;
the removing initial data points corresponding to the data points in the additional grid image from the initial model comprises: acquiring a serial number of a data point in the additional grid image as a serial number to be removed; acquiring an initial data point with the serial number to be removed as the initial data point to be removed; and removing the initial data points to be removed from the initial model to obtain the target model.
2. The data point model processing method of claim 1, wherein the initial model is a three-dimensional point cloud; the model to be processed is a two-dimensional image;
The step of forming the model to be processed from the initial model includes: providing a projection direction; and carrying out projection processing on the initial model along a projection direction to form the model to be processed.
3. The data point model processing method according to claim 2, wherein the number of projection directions is plural, the data point model processing method further comprising: and repeating the cyclic operation processing until additional grid images of all additional models are acquired, wherein the cyclic operation processing comprises the step of judging from the projection processing to the connected domain.
4. The data point model processing method of claim 3, wherein the number of additional models is a plurality, the repetitive cycle operation process further comprising: removing initial data points corresponding to the data points in the additional grid image from the initial model;
Or repeating the cyclic operation until additional grid images of all the additional models are acquired, and removing initial data points corresponding to the data points in all the additional grid images from the initial models.
5. The data point model processing method of claim 1, wherein the step of forming a model to be processed from the initial model comprises: and enabling the model to be processed to be identical to an initial model, wherein the data points are identical to the initial data points.
6. The data point model processing method of claim 1, wherein the step of labeling comprises: setting a first quantity threshold; comparing the number of data points in the grid with the first quantity threshold, and marking the grid as an image point when the number of data points in the grid is larger than or equal to the first quantity threshold; otherwise, the grid is marked as blank.
7. The data point model processing method of claim 1, wherein the initial model is an image; the initial data points comprise gray values of pixels in an initial model;
The marking process includes the steps of: setting a second number threshold; obtaining a reference number according to the number of data points meeting gray scale conditions in the grid; marking the grid as an image point when the reference number is greater than or equal to a second number threshold; otherwise, marking the grid as blank spots;
Or the step of marking comprises: setting a first gray threshold; acquiring a grid gray value according to the gray value of each data point in the grid; and comparing the grid gray value with a first gray threshold value, and marking grids meeting different comparison results as image points and blank points respectively.
8. A method of data point model processing according to any one of claims 2-3, wherein the projection direction comprises: a first projection direction perpendicular to an optical axis of the objective; and/or a second projection direction parallel to the optical axis of the objective lens.
9. A method of detection comprising:
Detecting an object to be detected to obtain an initial model;
The data point model processing method according to any one of claims 1-7, wherein the initial model is processed to form a target model.
10. The method of detection of claim 9, further comprising: providing a design model; and comparing the target model with the design model to obtain the distortion of the object to be detected.
11. The inspection method of claim 9, wherein the object to be inspected is inspected by an inspection apparatus comprising an objective lens.
12. A data point model processing system, comprising:
An input system for providing an initial model, the initial model comprising a plurality of initial data points, the initial model comprising a main model and an additional model that are separate from each other; the additional model is a tool image of the bottom of the object to be detected, wherein the tool image is detected by the detection light of the detection equipment through the object to be detected;
The data processing system is used for forming a to-be-processed model according to an initial model, and the initial data points form data points in the to-be-processed model;
The grid dividing system is used for dividing the model to be processed into a plurality of grids;
The marking system is used for marking each grid, marking the model to be processed as a binary image and forming a grid image;
The connected domain judging system is used for judging the connected domain of the grid image, and acquiring an additional grid image when the grid image comprises a separated main grid image and the additional grid image;
The removing system is used for removing initial data points corresponding to the data points in the additional grid image from the initial model to form a target model;
The removal system is further configured to: acquiring a serial number of a data point in the additional grid image as a serial number to be removed; acquiring an initial data point with the serial number to be removed as the initial data point to be removed; and removing the initial data points to be removed from the initial model to obtain the target model.
13. The data point model processing system of claim 12, wherein the initial model is a three-dimensional point cloud; the model to be processed is a two-dimensional image;
The data processing system includes: and the projection system is used for carrying out projection processing on the initial model along the projection direction to form the model to be processed.
14. The data point model processing system of claim 12, wherein the data processing system comprises: and the equivalent system is used for enabling the model to be processed to be identical to an initial model, and the data points are identical to the initial data points.
15. The data point model processing system of claim 12, wherein the tagging system comprises:
A setting system for setting a first number threshold;
A comparison system for comparing the number of data points in the grid with the first quantity threshold, and marking the grid as an image point when the number of data points in the grid is greater than or equal to the first quantity threshold; otherwise, the grid is marked as blank.
16. A detection system, comprising:
the model acquisition system is used for detecting the object to be detected and acquiring an initial model;
The data point model processing system of any of claims 12-15, configured to perform data point model processing on an initial model to form a target model.
17. The inspection system of claim 16, wherein the input system is further configured to provide a design model;
The detection system further comprises: and the distortion detection system is used for comparing the target model with the design model to obtain the distortion of the object to be detected.
18. A computer readable medium comprising executable instructions that when executed cause a processor to perform the data point model processing method of any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910234448.2A CN111754385B (en) | 2019-03-26 | 2019-03-26 | Data point model processing method and system, detection method and system and readable medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910234448.2A CN111754385B (en) | 2019-03-26 | 2019-03-26 | Data point model processing method and system, detection method and system and readable medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111754385A CN111754385A (en) | 2020-10-09 |
| CN111754385B true CN111754385B (en) | 2024-06-11 |
Family
ID=72671006
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910234448.2A Active CN111754385B (en) | 2019-03-26 | 2019-03-26 | Data point model processing method and system, detection method and system and readable medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111754385B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056614A (en) * | 2016-06-03 | 2016-10-26 | 武汉大学 | Building segmentation and contour line extraction method of ground laser point cloud data |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105898151A (en) * | 2015-11-15 | 2016-08-24 | 乐视移动智能信息技术（北京）有限公司 | Image processing method and device |
| US20170169289A1 (en) * | 2015-12-15 | 2017-06-15 | Le Holdings(Beijing)Co., Ltd. | Hand recognizing method, system, and storage medium |
| CN106548520A (en) * | 2016-11-16 | 2017-03-29 | 湖南拓视觉信息技术有限公司 | A kind of method and system of cloud data denoising |
| CN106815847B (en) * | 2017-01-12 | 2019-11-22 | 非凡智慧（宁夏）科技有限公司 | Trees dividing method and single tree extracting method based on laser radar point cloud |
| CN108961149B (en) * | 2017-05-27 | 2022-01-07 | 北京旷视科技有限公司 | Image processing method, device and system and storage medium |
| CN109547766B (en) * | 2017-08-03 | 2020-08-14 | 杭州海康威视数字技术股份有限公司 | Panoramic image generation method and device |
| CN107578391A (en) * | 2017-09-20 | 2018-01-12 | 广东电网有限责任公司机巡作业中心 | A kind of method that three-dimensional point cloud noise reduction is carried out based on two-dimentional binary Images Processing |
| CN109145969B (en) * | 2018-08-03 | 2020-07-28 | 百度在线网络技术（北京）有限公司 | Method, device, equipment and medium for processing point cloud data of three-dimensional object |
| CN109214982B (en) * | 2018-09-11 | 2022-09-20 | 大连理工大学 | Three-dimensional point cloud imaging method based on double-cylinder projection model |
-
2019
- 2019-03-26 CN CN201910234448.2A patent/CN111754385B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056614A (en) * | 2016-06-03 | 2016-10-26 | 武汉大学 | Building segmentation and contour line extraction method of ground laser point cloud data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111754385A (en) | 2020-10-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9621793B2 (en) | Information processing apparatus, method therefor, and measurement apparatus | |
| KR101910484B1 (en) | A method for three dimensional (3d) vision inspection | |
| US11788973B2 (en) | Defect inspection device and defect inspection method | |
| JP2020008501A (en) | Surface defect detection device and surface defect detection method | |
| CN113781424B (en) | Surface defect detection method, device and equipment | |
| JP2009259036A (en) | Image processing device, image processing method, image processing program, recording medium, and image processing system | |
| JP2018096908A (en) | Inspection device and inspection method | |
| KR20200035320A (en) | Defect detection on transparent or translucent wafers | |
| JP2016205876A (en) | Image recognition device | |
| KR20180115646A (en) | Bead recognition apparatus using vision camera and method thereof | |
| JP5913903B2 (en) | Shape inspection method and apparatus | |
| CN116978834B (en) | Intelligent monitoring and early warning system for wafer production | |
| CN115861351A (en) | Edge detection method, defect detection method and detection device | |
| KR20080056149A (en) | A method and a system for creating a reference image using unknown quality patterns | |
| KR100941978B1 (en) | Method of 3D Inspection of the object using CCD camera laser beam and Apparutus thereof | |
| Li et al. | Stereo vision based automated solder ball height and substrate coplanarity inspection | |
| CN111754385B (en) | Data point model processing method and system, detection method and system and readable medium | |
| CN113039631B (en) | Defect classification by fitting optical signals to point spread functions | |
| CN111833286A (en) | Point cloud processing method and system, detection method and system and readable medium | |
| JP2018155646A (en) | Surface measuring device and surface measurement method | |
| CN103493096A (en) | Analysis of the digital image of the internal surface of a tyre and processing of false measurement points | |
| Kang et al. | Detection of calibration patterns for camera calibration with irregular lighting and complicated backgrounds | |
| CN112285070B (en) | Method and device for detecting bright spots on image and image registration method and device | |
| JP2002277409A (en) | Inspection device for pattern of printed board | |
| CN107449373B (en) | High-speed structured light scanning method and system based on stereoscopic vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| CB02 | Change of applicant information |
Address after: 518110 No. 2 101, 201, 301, Shangyanglang No. 4 Industrial Zone, Tongsheng Community, Dalang Street, Longhua District, Shenzhen City, Guangdong Province Applicant after: Shenzhen Zhongke feice Technology Co.,Ltd. Address before: 518110 No.2, Shanghenglang 4th Industrial Zone, Tongsheng community, Dalang street, Longhua District, Shenzhen City, Guangdong Province 101 201 301 Applicant before: SKYVERSE Ltd. |
|
| CB02 | Change of applicant information | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |