Material screening method based on YOLO algorithm

Abstract

The invention belongs to the technical field of image processing, and particularly relates to a material screening method based on a YOLO algorithm, which comprises the following steps: establishing a three-dimensional material model for warehousing installation materials by using the YOLO algorithm, respectively numbering, establishing a three-dimensional mechanism model for a mechanism lacking the installation materials by using the YOLO algorithm, and performing shape deviation analysis to obtain shape deviation. And calculating the compensation degree of the three-dimensional material model in the database, determining a candidate three-dimensional material model, performing simulation mechanics to determine the installation stability of the three-dimensional material, and outputting the number corresponding to the determined three-dimensional material model to the material library. According to the technical scheme, image acquisition and three-dimensional material model establishment are carried out when the mounting materials which are put in storage are put in storage, so that repeated comparison of each mechanism during mounting is reduced, and the product quality is ensured while the production efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and in particular relates to a material screening method based on the YOLO algorithm. Background Art

[0002] Material screening through image processing methods is widely used in industrial production. It can improve production efficiency, reduce labor costs, and achieve accurate screening and quality control of materials.

[0003] As mentioned in CN118115675A, a material screening method, device and system based on visual detection algorithm, installation materials refer to materials, parts, etc. required for use in the assembly process of various mechanisms; when a mechanism needs to replace the installation materials, it is necessary to screen suitable installation materials for the mechanism. Existing vision-based material screening methods often directly perform visual analysis on the materials to determine whether the materials are of standard size in order to determine the standard materials. However, in the actual production process, there are tolerances for both the mechanism and the installation materials. For example, the tolerance of the mechanism is ±3mm, and the tolerance of the installation material is ±3mm. If the actual size of the mechanism installation position is the standard value +3mm, then according to the existing screening method, the installation material with a standard value of -3mm may be screened out. At this time, the actual size deviation between the two is large, which will reduce the installation stability of the installation material.

[0004] To this end, the technical solution adopts the following screening method:

[0005] S1: Acquire a mechanism image lacking a first installation material and material images of each first installation material;

[0006] S2: Perform visual inspection and analysis on the mechanism image by using a visual inspection algorithm to obtain three-dimensional reconstruction information of the mechanism image, and then establish a three-dimensional mechanism model based on the three-dimensional reconstruction information;

[0007] S3: for any material image, perform visual inspection and analysis on the material image through a visual inspection algorithm to obtain three-dimensional reconstruction information of the material image, and then establish a three-dimensional material model corresponding to the material image according to the three-dimensional reconstruction information;

[0008] S4: determining an installation position of the first installation material on the three-dimensional mechanism model, and performing shape deviation analysis on the installation position, thereby obtaining a first shape deviation corresponding to the installation position;

[0009] S5: calculating the compensation degree of the shape of each three-dimensional material model to the first shape deviation respectively, and taking the three-dimensional material model corresponding to the shape whose compensation degree to the first shape deviation exceeds the preset compensation degree as the candidate three-dimensional material model;

[0010] S6: for any candidate three-dimensional material model, install the candidate three-dimensional material model into the installation position of the three-dimensional mechanism model, and simulate mechanical action on the candidate three-dimensional material model to determine the installation stability of the candidate three-dimensional material model;

[0011] S7: Outputting the material image corresponding to the candidate three-dimensional material model with the highest installation stability.

[0012] Through analysis, it can be seen that this technical solution does solve to a certain extent the problem of large actual size deviation in the original screening method, which will reduce the installation stability of the installed materials. However, this technical solution still has the problem of long screening time and limited increase in production efficiency. Summary of the invention

[0013] The purpose of the present invention is to provide a material screening method based on the YOLO algorithm to solve the problems existing in the above-mentioned background technology.

[0014] To achieve the above objectives, this application is implemented through the following technical solutions:

[0015] A material screening method based on the YOLO algorithm adopts the following steps:

[0016] S1. Collect the image of each installation material in the material library, and perform visual inspection and analysis through the YOLO algorithm to obtain the 3D reconstruction information of the installation material, and then establish the 3D material model corresponding to the installation material image according to the 3D reconstruction information, and store them in the database after numbering them respectively;

[0017] S2, collecting images of the mechanism without installation materials, and performing visual inspection and analysis on the mechanism images through the YOLO algorithm to obtain three-dimensional reconstruction information of the mechanism images, and then establishing a three-dimensional mechanism model based on the three-dimensional reconstruction information;

[0018] S3, determining the installation position of the installation material on the three-dimensional mechanism model, and performing shape deviation analysis on the installation position, thereby obtaining the shape deviation corresponding to the installation position;

[0019] S4, calculating the compensation degree of the three-dimensional material model stored in the database in step S1 for the shape deviation, and taking all corresponding three-dimensional material models whose compensation degree exceeds the set value as candidate three-dimensional material models;

[0020] S5, arranging the candidate three-dimensional material models according to the serial numbers and the order of storage, and sequentially installing the candidate three-dimensional material models into the installation positions of the three-dimensional mechanism model, and performing simulated mechanical actions on the candidate three-dimensional material models to determine the installation stability of the candidate three-dimensional material models;

[0021] S6. Output the number corresponding to the candidate three-dimensional material model with the highest installation stability to the material library.

[0022] Furthermore, in step S6, when there are two or more candidate three-dimensional material models with the highest installation stability, the serial number corresponding to the installation material that first enters the material library is output to the material library.

[0023] Furthermore, step S6 also includes outputting the number corresponding to the first candidate three-dimensional material model whose installation stability reaches the set value to the material library.

[0024] Furthermore, the material screening device includes a first processing module, which is arranged in the installation material library, and is used to collect image information of the installation material, and perform visual detection and analysis on the image of each installation material through the YOLO algorithm to obtain three-dimensional reconstruction information of the installation material, and then establish a three-dimensional material model corresponding to the installation material image based on the three-dimensional reconstruction information, and store it in the database after numbering it respectively.

[0025] Furthermore, the material screening device includes an image acquisition device and a terminal device, wherein the image acquisition device is used to acquire images of the mechanism; and the terminal device is connected to the image acquisition device.

[0026] Furthermore, the terminal device includes a second processing module for determining an installation position of the installation material on the three-dimensional mechanism model, and performing shape deviation analysis on the installation position, thereby obtaining a shape deviation corresponding to the installation position;

[0027] A third processing module is used to calculate the compensation degree of the shape of each 3D material model to the shape deviation, and take the 3D material model corresponding to the shape whose compensation degree to the shape deviation exceeds the preset compensation degree as the candidate 3D material model;

[0028] The fourth processing module uses multiple channels to simultaneously install each candidate three-dimensional material model into the installation position of the three-dimensional mechanism model, and simulates mechanical effects on each candidate three-dimensional material model to determine the installation stability of each candidate three-dimensional material model.

[0029] The beneficial effects of the present invention are:

[0030] This technical solution reduces repeated comparisons during installation for each mechanism by performing image acquisition and establishing a three-dimensional material model when the installation materials enter the warehouse, thereby improving production efficiency while ensuring product quality. DETAILED DESCRIPTION

[0031] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings. The following embodiments are merely exemplary and can only be used to explain and illustrate the technical solution of the present invention, but cannot be construed as limiting the technical solution of the present invention.

[0032] This application provides a material screening method based on the YOLO algorithm, using the following steps:

[0033] S1. Collect the image of each installation material in the material library, and perform visual inspection and analysis through the YOLO algorithm to obtain the 3D reconstruction information of the installation material, and then establish the 3D material model corresponding to the installation material image based on the 3D reconstruction information, and store them in the database after numbering them respectively.

[0034] In this technical feature of the present application, regardless of whether the installation materials are purchased externally or produced by the enterprise itself, the products produced are stored in the material library. Of course, if the installation materials are produced by the enterprise itself, a three-dimensional material model can be established at the same time as the product is offline, and the materials can be numbered and stored in the database. The database of the present application can be a local database, and subsequent terminal devices can call the data in the local database during operation.

[0035] In other embodiments of the present application, the material screening device includes a first processing module, which is arranged in the installation material library, and is used to collect image information of the installation material, and perform visual detection and analysis on the image of each installation material through the YOLO algorithm to obtain three-dimensional reconstruction information of the installation material, and then establish a three-dimensional material model corresponding to the installation material image based on the three-dimensional reconstruction information, and store it in the database after numbering it respectively.

[0036] In this application, the numbering of each installation material is not repeated, so that the data of each installation material can be monitored.

[0037] The key point of this application is that when the installation materials are put into the warehouse, the three-dimensional material model is saved to the database, and there is no need to repeat the modeling for each mechanism that lacks installation materials during the working process. In addition, data is directly retrieved from the database, which is efficient and utilizes the management of the installation material library.

[0038] S2. Collect images of the mechanism without installation materials, and perform visual inspection and analysis on the mechanism images through the YOLO algorithm to obtain three-dimensional reconstruction information of the mechanism images, and then establish a three-dimensional mechanism model based on the three-dimensional reconstruction information.

[0039] In the present application, the scope of the mechanism that lacks installation materials can be a certain mechanical part in the mechanical assembly process. In the next process, the installation material needs to be installed on the mechanical part. It can also be in the field of maintenance, where a material that needs to be replaced is disassembled and replaced with a new installation material. The technical solution of the present application can be applied.

[0040] The material screening device comprises an image acquisition device and a terminal device, wherein the image acquisition device is used to acquire images of the mechanism; and the terminal device is connected to the image acquisition device.

[0041] S3. Determine the installation position of the installation material on the three-dimensional mechanism model, and perform shape deviation analysis on the installation position to obtain the shape deviation corresponding to the installation position. The terminal device includes a second processing module, which is used to determine the installation position of the installation material on the three-dimensional mechanism model, and perform shape deviation analysis on the installation position to obtain the shape deviation corresponding to the installation position.

[0042] S4. Calculate the degree of compensation for shape deviation of the three-dimensional material model stored in the database in step S1, and take all corresponding three-dimensional material models whose compensation degree exceeds the set value as candidate three-dimensional material models; this process involves calling the three-dimensional material models of all installation materials in the database.

[0043] S5. Arrange the candidate three-dimensional material models according to the number and the order of storage, and install the candidate three-dimensional material models into the installation positions of the three-dimensional mechanism model in sequence, and simulate the mechanical action on the candidate three-dimensional material model to determine the installation stability of the candidate three-dimensional material model.

[0044] Specifically, multiple channels are used to simultaneously install each candidate 3D material model into the installation position of the 3D mechanism model, and simulate mechanical action on each candidate 3D material model to determine the installation stability of each candidate 3D material model. Here, it means that during the data retrieval process, the terminal device of the device can retrieve 10 data at the same time. Of course, the specific number can be more or less than 10. This is just an example. Each channel installs a candidate 3D material model into the installation position of the 3D mechanism model, and simulates mechanical action on each candidate 3D material model to determine the installation stability of each candidate 3D material model.

[0045] Therefore, the following situations may occur in the determination of installation stability in this application:

[0046] In one case, the number corresponding to the candidate 3D material model with the highest installation stability is output to the material library. When there are more than two candidate 3D material models with the highest installation stability, the number corresponding to the installation material that enters the material library first is output to the material library.

[0047] In one case, the number corresponding to the first candidate three-dimensional material model whose installation stability reaches the set value is output to the material library. This technology does not pursue the highest installation stability, and if only one installation material reaches the set value among the installation stability calculated at the same time, the number corresponding to the installation material model is directly output to the material library. If there are more than two installation materials that reach the set value at the same time, the installation material number corresponding to the candidate three-dimensional material model with the highest stability is output to the material library.

[0048] The above is only an embodiment of the present invention, and the common sense such as the known specific structure and characteristics in the scheme is not described in detail here. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present invention is defined by the attached claims rather than the above description, so it is intended to include all changes within the meaning and scope of the equivalent elements of the claims. Any figure mark in the claims should not be regarded as limiting the claims involved.

Claims

1. A material screening method based on the YOLO algorithm, characterized in that: Use the following steps: S1. Collect the image of each installation material in the material library, and perform visual inspection and analysis through the YOLO algorithm to obtain the 3D reconstruction information of the installation material, and then establish the 3D material model corresponding to the installation material image according to the 3D reconstruction information, and store them in the database after numbering them respectively; S2, collecting images of the mechanism without installation materials, and performing visual inspection and analysis on the mechanism images through the YOLO algorithm to obtain three-dimensional reconstruction information of the mechanism images, and then establishing a three-dimensional mechanism model based on the three-dimensional reconstruction information; S3, determining the installation position of the installation material on the three-dimensional mechanism model, and performing shape deviation analysis on the installation position, thereby obtaining the shape deviation corresponding to the installation position; S4, calculating the compensation degree of the three-dimensional material model stored in the database in step S1 for the shape deviation, and taking all corresponding three-dimensional material models whose compensation degree exceeds the set value as candidate three-dimensional material models; S5, arranging the candidate three-dimensional material models according to the serial numbers and the order of storage, and sequentially installing the candidate three-dimensional material models into the installation positions of the three-dimensional mechanism model, and performing simulated mechanical actions on the candidate three-dimensional material models to determine the installation stability of the candidate three-dimensional material models; S6. Output the number corresponding to the candidate three-dimensional material model with the highest installation stability to the material library.

2. The material screening method based on the YOLO algorithm according to claim 1, characterized in that: In step S6, when there are two or more candidate three-dimensional material models with the highest installation stability, the serial number corresponding to the installation material that first enters the material library is output to the material library.

3. The material screening method based on the YOLO algorithm according to claim 1, characterized in that: Step S6 also includes outputting the number corresponding to the first candidate three-dimensional material model whose installation stability reaches the set value to the material library.

4. The material screening method based on the YOLO algorithm according to claim 1, characterized in that: The material screening device includes a first processing module, which is arranged in the installation material library, and is used to collect image information of the installation material, and perform visual detection and analysis on the image of each installation material through the YOLO algorithm to obtain three-dimensional reconstruction information of the installation material, and then establish a three-dimensional material model corresponding to the installation material image based on the three-dimensional reconstruction information, and store it in the database after numbering it respectively.

5. The material screening method based on the YOLO algorithm according to claim 4 is characterized in that: The material screening device comprises an image acquisition device and a terminal device, wherein the image acquisition device is used to acquire images of the mechanism; and the terminal device is connected to the image acquisition device.

6. The material screening method based on the YOLO algorithm according to claim 5 is characterized in that: The terminal device includes a second processing module, which is used to determine the installation position of the installation material on the three-dimensional mechanism model, and perform shape deviation analysis on the installation position, thereby obtaining the shape deviation corresponding to the installation position; A third processing module is used to calculate the compensation degree of the shape of each 3D material model to the shape deviation, and take the 3D material model corresponding to the shape whose compensation degree to the shape deviation exceeds the preset compensation degree as the candidate 3D material model; The fourth processing module uses multiple channels to simultaneously install each candidate three-dimensional material model into the installation position of the three-dimensional mechanism model, and simulates mechanical effects on each candidate three-dimensional material model to determine the installation stability of each candidate three-dimensional material model.