CN115456964A

CN115456964A - Product internal flow passage molded surface extraction method based on industrial CT

Info

Publication number: CN115456964A
Application number: CN202211026883.4A
Authority: CN
Inventors: 凌文辉; 马兆光; 闫熙; 张宏; 康达; 王梓鑫
Original assignee: Beijing Power Machinery Institute
Current assignee: Beijing Power Machinery Institute
Priority date: 2022-08-25
Filing date: 2022-08-25
Publication date: 2022-12-09

Abstract

The application provides a product internal flow passage profile extraction method based on industrial CT. The method comprises the following steps: the method comprises the following steps: acquiring an industrial CT three-dimensional image of a product to be detected; slicing the industrial CT three-dimensional image to obtain a plurality of two-dimensional slices; obtaining a plurality of runner contours in the plurality of two-dimensional slices; splicing the plurality of flow channel outlines to form a three-dimensional model; and extracting the runner profile based on the three-dimensional model. According to the method, the flow channel structure is directly extracted by adopting a segmentation technology in the three-dimensional CT volume data, and is reconstructed into the three-dimensional point cloud model, so that the visual and rapid analysis of parameters such as the wall thickness, the size, the shape, the directionality and the like of the flow channel is realized. The problem that the blockage, the size and the integrity of a flow channel cannot be evaluated directly through volume data in the prior art of the current industrial CT is solved.

Description

Product internal flow passage profile extraction method based on industrial CT

Technical Field

The application relates to the technical field of nondestructive testing, in particular to a product internal flow passage profile extraction method based on industrial CT.

Background

Industrial CT (Computed Tomography) is one of nondestructive testing technologies, and uses X-ray beams to penetrate a detected workpiece in a transmission manner from a plurality of equiangular directions distributed circumferentially, and uses a detector to collect and record the attenuated X-ray information after penetrating the detected workpiece, and displays the structure of the detected workpiece in the form of a three-dimensional image by using an image reconstruction algorithm through a computer, and the intuitive result of the detection is the information of the internal and external three-dimensional structures, defects and the like of the detected workpiece, and can completely acquire the three-dimensional volume data of the detected workpiece. The industrial CT detection result has the advantages of intuition, clearness, accuracy, no damage to the detected workpiece, no limitation by the shape of the detected workpiece material, capability of directly obtaining three-dimensional data, strong longitudinal resolution, high detection efficiency and the like, is one of advanced detection equipment in the world at present, and is widely applied to the application fields of additive manufacturing, composite materials and the like.

When the industrial CT performs complex structure analysis, rendering analysis is usually performed on different internal tissues, different structures and different flow channel directions, the structure size and the flow channel profile are analyzed, information such as the position and the geometric size of the structure is determined, and data is provided for product manufacturing conformity.

In the prior art, the structure of the molded surface of the flow channel is observed layer by layer through a CT slice image, the blockage, the size and the integrity of the flow channel of the layer by layer are checked, the condition of the geometric size of the flow channel is indirectly analyzed, and the problems of non-visual detection result, long consumed time and the like exist.

Disclosure of Invention

The application provides a product internal flow passage molded surface extraction method based on industrial CT, which is used for at least solving the problem A in the related art. The technical scheme of the application is as follows:

in a first aspect, an embodiment of the present application provides a method for extracting a product internal flow channel profile based on industrial CT, including:

acquiring an industrial CT three-dimensional image of a product to be detected;

slicing the industrial CT three-dimensional image to obtain a plurality of two-dimensional slices;

obtaining a plurality of runner contours in the plurality of two-dimensional slices;

splicing the plurality of flow channel outlines to form a three-dimensional model;

and extracting the runner profile based on the three-dimensional model.

In some embodiments, slicing the industrial CT three-dimensional image comprises:

preprocessing the industrial CT three-dimensional image, and determining the shape characteristics of a product to be detected in the industrial CT three-dimensional image;

determining a cutting direction based on the shape feature;

and slicing the industrial CT three-dimensional image based on the cutting direction.

In some embodiments, the acquiring a plurality of flow channel profiles in the plurality of two-dimensional slices comprises:

determining an analysis area of the runner information to be extracted of the plurality of two-dimensional slices;

determining a first area where a flow channel is located in the analysis area based on an image recognition algorithm;

and adjusting search parameters based on an image search algorithm, and acquiring a flow channel contour in a first area corresponding to each of the plurality of two-dimensional slices so as to acquire a plurality of flow channel contours in the plurality of two-dimensional slices.

In some embodiments, the determining the analysis region of the plurality of two-dimensional slices from which the flow channel information is to be extracted includes:

and determining an analysis area of the runner information to be extracted of the plurality of two-dimensional slices based on a three-view method.

In some embodiments, the determining a first region in which the flow channel is located in the analysis region based on an image recognition algorithm includes:

and determining a first area where the flow channel is located based on the gray scale characteristics of the image in the analysis area.

In some embodiments, the adjusting the search parameter based on the image search algorithm to obtain the flow channel contour in the first area corresponding to each of the plurality of two-dimensional slices to obtain the plurality of flow channel contours in the plurality of two-dimensional slices includes:

determining a first search parameter for a first two-dimensional slice in the plurality of two-dimensional slices to obtain a flow channel contour;

determining a preset number of first slice groups from the plurality of two-dimensional slices, and acquiring a flow channel contour corresponding to each two-dimensional slice in the first slice group based on the first search parameter and an image search algorithm;

adjusting the first search parameter, determining a preset number of second slice groups from the remaining two-dimensional slices except the first slice group from the plurality of two-dimensional slices, and acquiring a runner profile corresponding to each two-dimensional slice in the second slice groups based on the adjusted first search parameter and an image search algorithm; until a plurality of flow channel profiles in the plurality of two-dimensional slices is acquired.

In some embodiments, after the stitching the plurality of runner profiles to form the three-dimensional model, the stitching further comprises:

judging whether the flow channel model accords with the flow channel characteristics;

and when the runner model does not accord with the runner characteristics, repeating the step of obtaining the plurality of runner outlines in the plurality of two-dimensional slices until the runner model accords with the runner characteristics.

In some embodiments, said extracting said flowpath profile based on said three-dimensional model comprises:

and removing redundant parts in the three-dimensional model, and extracting the runner profile.

In some embodiments, the search parameters include an inner contour weight, an outer contour weight, a canonical term system, a length term system, a time step, a delta (x) parameter, a gaussian kernel parameter, an iteration number, and a parametric variable step.

The technical scheme provided by the embodiment of the application at least has the following beneficial effects:

by adopting a segmentation technology in three-dimensional CT volume data, a flow channel structure is directly extracted and reconstructed into a three-dimensional point cloud model, so that the visual and rapid analysis of parameters such as wall thickness, size, shape, directionality and the like of the flow channel is realized. The problem that the blockage, the size and the integrity of a flow channel cannot be evaluated directly through volume data in the prior art of the current industrial CT is solved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application and are not to be construed as limiting the application.

Fig. 1 is a flow chart illustrating a method for industrial CT based product internal flow channel profile extraction in accordance with an exemplary embodiment.

Fig. 2 is a flow chart illustrating a method for industrial CT based product internal flow channel profile extraction in accordance with another exemplary embodiment.

Fig. 3 is a flow chart illustrating a method for industrial CT based product internal flow channel profile extraction in accordance with yet another exemplary embodiment.

Fig. 4 is a flowchart illustrating a method for extracting a product internal flow channel profile based on industrial CT according to a specific example 1.

Fig. 5 is an industrial CT three-dimensional image of a part product of example 1 of the present invention.

Fig. 6 is a schematic view of the shape characteristics of the product of parts of example 1 of the present invention.

Fig. 7 is a setting interface of the cutting direction of example 1 of the present invention.

FIG. 8 is a schematic illustration of a determined analysis area of a part product of example 1 of the present invention.

Fig. 9 is a first region selection setting interface of example 1 of the present invention.

FIG. 10 is a schematic illustration of a defined first region of a part product of example 1 of the present invention.

Fig. 11 is a schematic diagram of setting of search parameters of example 1 of the present invention.

Fig. 12 is a schematic view of a flow channel profile obtained in example 1 of the present invention.

Fig. 13 is a schematic view of a three-dimensional model of example 1 of the present invention.

Fig. 14 (a) and (b) are schematic views of the flow channel profile of example 1 of the present invention.

Detailed Description

In order to make the technical solutions of the present application better understood by those of ordinary skill in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

It should be noted that the terms "first," "second," and the like in this application are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the application, as detailed in the appended claims.

Fig. 1 is a flowchart of a method for extracting a product internal flow channel profile based on industrial CT according to an embodiment of the present application. As shown in fig. 1, the method for extracting the internal flow channel profile of the industrial CT-based product may include the following steps.

And S101, acquiring an industrial CT three-dimensional image of a product to be detected.

In this embodiment, the industrial CT detection device is used to obtain CT volume data of the product to be detected, where the CT volume data is an industrial CT three-dimensional image.

And S102, slicing the industrial CT three-dimensional image to obtain a plurality of two-dimensional slices.

In order to analyze the internal structure of the product to be detected, the acquired industrial CT three-dimensional image needs to be sliced to obtain a two-dimensional slice image, which is convenient for analyzing the internal structure of the product to be detected.

The industrial CT three-dimensional image can be sliced through software carried by an industrial CT detection device to obtain a plurality of two-dimensional slices.

Step S103, acquiring a plurality of flow channel contours in the plurality of two-dimensional slices.

In order to extract the flow channel profile inside the product, the flow channel contour in multiple two-dimensional slice images needs to be identified.

And step S104, splicing the plurality of flow channel outlines to form a three-dimensional model.

It is to be understood that after acquiring the plurality of flow channel profiles in the plurality of two-dimensional slices, a three-dimensional reconstruction is performed. Optionally, after identifying the runner outlines from the multiple two-dimensional slices, splicing the multiple runner outlines according to the slice sequence of the original industrial CT three-dimensional picture, and obtaining the runner profiles.

The three-dimensional model is a three-dimensional point cloud model.

And S105, extracting the runner molded surface based on the three-dimensional model.

And extracting the runner molded surface according to the obtained three-dimensional model spliced by the runner contour.

After extracting the channel profiles, the channel information can be directly extracted from the original CT volume data, including: the parameters of the flow channel such as wall thickness, size, shape, directionality and the like are convenient for observing the shape of the flow channel, measuring and analyzing the size characteristics of the flow channel, detecting the defects of the flow channel and the like.

In the specific implementation formula, after the three-dimensional model is confirmed, CT volume data can be loaded to generate an STL file.

According to the method for extracting the molded surface of the internal flow channel of the product based on the industrial CT, the flow channel structure is directly extracted by adopting a segmentation technology in the obtained three-dimensional image of the industrial CT and is reconstructed into a three-dimensional model, so that the intuitive and rapid analysis of a plurality of parameters of the flow channel is realized. Namely, the segmentation technology is utilized to segment and reconstruct the interested region, the specific information of the internal flow channel of the product is extracted on the premise of not damaging the detected product, the manufacturing conformance inspection of the internal flow channel of the product is convenient, the operation is simple and rapid, and the calculated amount is small. Based on the existing industrial CT collected image, the shape of the flow channel can be restored with high precision, the shape and the size characteristics of the flow channel can be observed clearly, and the position and the shape of the defect can be determined. In the prior art, the size of the flow channel can be measured only layer by layer according to the content of the CT image, so that the technology has great significance for improving the accuracy of detecting the internal structure of the product and the detection efficiency.

In some embodiments, slicing the industrial CT three-dimensional image, as shown in fig. 2, comprises the steps of:

step S201, preprocessing the industrial CT three-dimensional image, and determining the shape characteristics of the product to be detected in the industrial CT three-dimensional image.

It can be understood that, before slicing, the cutting direction needs to be determined, and then the overall shape characteristics of the product to be detected need to be obtained first to determine the cutting direction, and only if the morphological characteristics of the product to be detected are known, the proper cutting direction can be determined.

By preprocessing the industrial CT image to be detected and searching the area where the boundary of the product to be detected is located, the shape characteristic of the product to be detected can be obtained so as to observe the shape characteristic and determine the cutting direction.

Step S202, determining the cutting direction based on the shape characteristics.

And searching a proper cutting direction (x, y, z) according to the shape characteristics of the product so as to ensure that the two-dimensional slice has a clear outline and make the internal and external characteristics of the product easy to distinguish.

Optionally, the industrial CT three-dimensional image is two-dimensionally sliced along a specific axis to ensure that the slices of the product have clear outlines and internal and external features.

And S203, slicing the industrial CT three-dimensional image based on the cutting direction.

In the embodiment, the shape characteristics of the product to be detected are firstly identified, the cutting direction is determined according to the shape characteristics, and the two-dimensional slice is obtained by cutting in the proper cutting direction, so that the two-dimensional slice is ensured to have clear outline and internal and external characteristics.

In some embodiments, acquiring a plurality of flow channel profiles in a plurality of two-dimensional slices, as shown in fig. 3, comprises the steps of:

step S301, determining an analysis region of the flow channel information to be extracted of the multiple two-dimensional slices.

In order to reduce the burden of a computer and improve the modeling speed, a region which is as small and accurate as possible is selected for slice analysis. Namely, selecting the product part of the flow channel information to be extracted.

As a possible implementation manner, based on a three-view method, an analysis area of the runner information to be extracted of the multiple two-dimensional slices is determined. Namely, the region to be analyzed in the industrial CT three-dimensional image is divided by a three-view method.

And a three-view method is adopted, so that the method is visual and simple.

Step S302, based on an image recognition algorithm, determining a first area where the flow channel is located in the analysis area.

As a possible implementation manner, the first area where the flow channel is located is determined based on the gray scale features of the image in the analysis area.

And carrying out initial flow channel contour setting through an image recognition algorithm, and roughly selecting the area where the flow channel is located.

In this embodiment, the position of the flow channel is determined by using the image gray scale information of the two-dimensional slice image, the two-dimensional slice image is subjected to preliminary processing according to the gray scale characteristics, and an initial range is divided for the search of the flow channel profile, so that the success rate of the flow channel profile search can be greatly improved, the accuracy of the flow channel profile identification is improved, the time required by the search is reduced, and the efficiency of establishing the three-dimensional model of the flow channel is improved.

Step S303, based on an image search algorithm, adjusting search parameters, and obtaining a flow channel contour in a first area corresponding to each of the plurality of two-dimensional slices, so as to obtain a plurality of flow channel contours in the plurality of two-dimensional slices.

It should be noted that, when the flow channel contour in each two-dimensional slice is obtained by using the image search algorithm for a plurality of two-dimensional slices, not all the two-dimensional slices use the same search parameter to perform the image search algorithm. Because the shape of the whole product is irregular, and the shape and position of the flow channel are different for each two-dimensional slice, the search parameters need to be adjusted for different two-dimensional slices to obtain the flow channel profile more accurately.

As a possible implementation, acquiring a plurality of flow channel profiles in the plurality of two-dimensional slices includes:

adjusting the first search parameter, determining a preset number of second slice groups from the remaining two-dimensional slices except the first slice group from the plurality of two-dimensional slices, and acquiring a flow channel contour corresponding to each two-dimensional slice in the second slice groups based on the adjusted first search parameter and an image search algorithm; until a plurality of flow channel profiles in the plurality of two-dimensional slices is acquired.

That is, firstly, for a certain single two-dimensional slice, using an image search algorithm to adjust search parameters, and selecting a clear flow channel contour in a first area of the two-dimensional slice; and then extracting a proper number of two-dimensional slices, and performing contour search on other two-dimensional slices by using the current search parameters. Then, aiming at the next two-dimensional slice, the applicability of the current search parameter is confirmed, if the applicability is not suitable, the search parameter is adjusted, the new search parameter suitable for the current two-dimensional slice is confirmed, then the proper number of two-dimensional slices are extracted, and the new search parameter is adopted to carry out flow channel contour search; until the flow channel contours in all two-dimensional slices are confirmed. Namely, on the basis of the search parameters of the current two-dimensional slice, selecting other two-dimensional slices for contour search, and repeatedly adjusting the search parameters to confirm the stability of the runner contour search result. And the two-dimensional slice data extracted each time can be the same or different, and the accuracy of flow channel contour division is ensured by a method for continuously confirming the applicability of the current search parameter.

It should be noted that the search parameters may include, but are not limited to, an inner contour weight, an outer contour weight, a regular term system, a length term system, a time step, a δ (x) parameter, a gaussian kernel parameter, an iteration number, and a parameter-varying step.

In the embodiment, the selection of the slice region of the industrial CT three-dimensional image, the determination of the position range of the flow channel and the selection of the search parameter of the flow channel contour are realized through the steps of analyzing the region determination, the initial contour setting and the flow channel contour search, so that the calculation amount is effectively reduced, the processing speed is improved, and the accuracy of flow channel contour identification is improved, thereby improving the efficiency of establishing the flow channel three-dimensional model.

In some cases, for example, if there is an abnormal feature caused by an image recognition error, the obtained three-dimensional model will be deviated to some extent, in this case, the obtained three-dimensional model needs to be determined, and if the determined result is that the currently obtained three-dimensional model is inaccurate, parameters of each step for obtaining the flow channel contour need to be adjusted to obtain the flow channel contour again. Therefore, in some embodiments, after the step S104 of stitching the plurality of flow channel profiles to form the three-dimensional model, the method further includes:

judging whether the flow channel model accords with the flow channel characteristics or not;

In this embodiment, whether the three-dimensional model is accurate is directly determined by observing whether the obtained three-dimensional model conforms to the characteristics of the flow channel.

And finally, the molded surface of the runner can be correctly extracted through the accuracy judgment of the obtained three-dimensional model.

Since the first area includes not only all the areas of the flow channels but also a partial area outside the product when determining the first area where the flow channel is located in step S302, the parts other than the flow channel need to be removed. Therefore, in step S105, extracting the runner profile based on the three-dimensional model, further comprises: and removing redundant parts in the three-dimensional model, and extracting the runner profile.

During specific implementation, redundant parts except for the runner structure in the three-dimensional model in the STL file are removed, the runner model is reserved, and finally the runner molded surface is extracted.

The method for extracting the internal flow channel profile of the product based on the industrial CT according to the embodiment of the present invention is described below by taking a certain part product having a flow channel structure as an example.

Fig. 4 is a flow chart of a method for extracting a product internal flow channel profile based on industrial CT according to an embodiment of the present application. As shown in fig. 4, the flow channel profile extraction method may include the steps of:

step S401, acquiring an industrial CT three-dimensional image of the part product, as shown in fig. 5.

Step S402, preprocessing the industrial CT three-dimensional image, and obtaining shape characteristics of the part product, as shown in fig. 6.

And S403, selecting a method for transversely cutting along the Z axis based on the shape characteristics, and slicing the industrial CT three-dimensional image to obtain a plurality of two-dimensional slices.

In particular, the Z-axis transection is selected in the interface shown in fig. 7.

Step S404, determining an analysis area of the flow channel information to be extracted of the plurality of two-dimensional slices based on a three-view method, where the analysis area is shown as an area pointed to by a in fig. 8.

Step S405, determining a first area where the flow channel is located based on the gray scale characteristics of the image in the analysis area.

For example, the contour of the flow channel to be selected is a black area in the two-dimensional slice, so that the upper limit value of the gray scale is reduced to about 47, and the first area is selected at the interface shown in fig. 9, so that the first area where the flow channel is located can be determined, as shown in fig. 10.

It should be noted that the first region shown in fig. 10 includes the positions of all the flow channels, and similarly, the selected region also includes the product exterior, which needs to be further processed after modeling.

Step S406, selecting a suitable search parameter for one of the two-dimensional slices to obtain a suitable flow channel profile.

The appropriate flow channel profile is obtained using the search parameters shown in fig. 11, and the flow channel profile is shown in fig. 12.

Step S407, determining the applicability of the current search parameter every preset number of two-dimensional slices until a plurality of flow channel profiles corresponding to all two-dimensional slices in the plurality of two-dimensional slices are obtained.

As an example, every 10 two-dimensional slices confirm the applicability of the current search parameters.

And step S408, splicing the plurality of flow channel outlines to form a three-dimensional model.

Step S409, the accuracy of the generated three-dimensional model is confirmed.

The resulting three-dimensional model, as shown in fig. 13, includes the flow channel structure and other structures of a portion of the product.

And S410, removing redundant parts of the three-dimensional model, reserving the runner model, and extracting the runner molded surface.

In specific implementation, after removing the redundant part of the three-dimensional model in the STL file, the obtained runner model is shown in fig. 14, where (a) and (b) in fig. 14 are schematic diagrams of two directions of the runner profile, respectively.

It should also be noted that the exemplary embodiments mentioned in this patent describe some methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the above-described steps, that is, the steps may be performed in the order mentioned in the embodiments, may be performed in an order different from the order in the embodiments, or may be performed simultaneously.

Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only.

It will be understood that the present application is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the appended claims.

Claims

1. A product internal flow passage molded surface extraction method based on industrial CT is characterized by comprising the following steps:

acquiring an industrial CT three-dimensional image of a product to be detected;

and extracting the runner profile based on the three-dimensional model.

2. The method of claim 1, wherein slicing the industrial CT three-dimensional image comprises:

determining a cutting direction based on the shape feature;

3. The method of claim 1, wherein said obtaining a plurality of flow channel profiles in said plurality of two-dimensional slices comprises:

4. The method of claim 3, wherein the determining an analysis region of the plurality of two-dimensional slices from which flow channel information is to be extracted comprises:

5. The method of claim 3, wherein determining a first region in which the flow channel is located within the analysis region based on an image recognition algorithm comprises:

6. The method of claim 3, wherein the adjusting search parameters based on the image search algorithm to obtain the flow channel contours in the first area corresponding to each of the plurality of two-dimensional slices to obtain a plurality of flow channel contours in the plurality of two-dimensional slices comprises:

7. The method of claim 1, wherein said stitching said plurality of runner profiles, after forming a three-dimensional model, further comprises:

8. The method of claim 1, wherein said extracting the runner profile based on the three-dimensional model comprises:

9. The method of claim 3, wherein the search parameters comprise an inner contour weight, an outer contour weight, a regularization term, a length term, a time step, a delta (x) parameter, a Gaussian kernel parameter, an iteration number, and a parameter variation step.