Mobile phone mainboard defect visual detection method and system based on AI technology
Through the visual detection method of mobile phone motherboard defects based on AI technology, combined with component division and image processing, the problem of difficulty in identifying internal defects in traditional detection is solved, and accurate defect identification and production process adjustment of mobile phone motherboards are achieved, and production quality is improved.
Patent Information
- Application Number
- CN202410994517.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional mobile phone motherboard detection methods are difficult to accurately identify internal defects, resulting in the inability to adjust the production process in a targeted manner, affecting production quality.
Using the visual detection method of mobile phone motherboard defects based on AI technology, the mobile phone motherboard is divided and image acquisition, and the external and internal defect images are screened, and the external defects are removed to determine internal defects, and the production process is adjusted.
It realizes accurate identification of internal defects of mobile phone motherboard and removal of external defects, ensures targeted adjustment of production processes, and improves production quality.
Smart Images

Figure CN118937335A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of visual inspection technology, and in particular to a method and system for visual inspection of mobile phone motherboard defects based on AI technology. Background Art
[0002] With the rapid development of the smartphone industry, the quality control of mobile phone motherboards has become increasingly important. As the core component of mobile phones, the performance of mobile phone motherboards directly affects the stability and performance of the entire mobile phone. Traditional mobile phone motherboard inspection methods usually rely on manual inspection or automated systems based on simple rules. These methods may not be able to effectively identify complex defects such as tiny cracks, poor micro solder joints, etc., especially in high productivity and precision manufacturing environments. The above shortcomings can be overcome to a certain extent through intelligent recognition technology.
[0003] However, the defects of mobile phone motherboards are not only reflected on the outside, but also on the inside of its parts. Due to the interference of superimposed external defects, internal defects cannot be accurately identified, resulting in the inability to make targeted adjustments to the parameters of the production process based on the identification results. Summary of the invention
[0004] In order to solve the above technical problems, a method and system for visual detection of mobile phone motherboard defects based on AI technology are provided. This technical solution solves the problem that the defects of the mobile phone motherboard mentioned in the above background technology are not only reflected on the outside, but also on the inside of its parts. Due to the interference of superimposed external defects, internal defects cannot be accurately identified, resulting in the inability to make targeted adjustments to the parameters of the production process according to the identification results.
[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:
[0006] A method for visually detecting defects of mobile phone motherboards based on AI technology, comprising:
[0007] Dividing at least one mobile phone mainboard into parts to obtain at least one basic parts structure;
[0008] Acquire at least one external defect and an external defect image of a basic part structure;
[0009] Based on at least one external defect image, at least one basic external defect image is obtained by screening, and an external defect corresponding to the basic external defect image is used as a basic external defect;
[0010] Under the condition that there are no defects inside the basic part structure, an image with basic external defects is obtained as a basic external defect infrared image;
[0011] Acquire at least one internal defect of a basic part structure and an infrared image of the internal defect;
[0012] Based on at least one internal defect infrared image, at least one basic internal defect infrared image is obtained by screening, and the internal defect corresponding to the basic internal defect infrared image is used as the basic internal defect;
[0013] Segment the tested mobile phone motherboard into components to obtain at least one component to be tested, and pair the component to be tested with the basic part structure;
[0014] Perform surface identification on the tested component to obtain at least one actual basic external defect;
[0015] Perform infrared recognition on the parts to be tested to obtain the actual infrared image;
[0016] Based on the actual basic external defects, the actual infrared image is preprocessed to obtain the actual internal infrared image, wherein the actual internal infrared image is an infrared image of the interior of the component to be tested;
[0017] Based on the actual internal infrared image, at least one actual basic internal defect is obtained;
[0018] Based on the actual basic external defects and the actual basic internal defects, the production process of the parts to be tested is adjusted.
[0019] Preferably, said dividing at least one mobile phone mainboard into parts to obtain at least one basic parts structure comprises the following steps:
[0020] Divide the components in the mobile phone mainboard into minimum units according to the functions to be implemented, and obtain at least one preliminary component structure;
[0021] Repetitive preliminary part structures in at least one preliminary part structure are removed to obtain at least one basic part structure.
[0022] Preferably, the screening and obtaining of at least one basic external defect image based on at least one external defect image comprises the following steps:
[0023] Based on the at least one external defect image, forming at least one external defect image combination, wherein the at least one external defect image combination includes all possible combinations of the at least one external defect image;
[0024] Acquire a characteristic external defect image combination, wherein at least one external defect image can be obtained by superimposing at least one external defect image in the characteristic external defect image combination;
[0025] At least one external defect image in the characteristic external defect image combination is used as at least one basic external defect image.
[0026] Preferably, the screening and obtaining of at least one basic internal defect infrared image based on at least one internal defect infrared image comprises the following steps:
[0027] Based on the at least one internal defect infrared image, forming at least one internal defect infrared image combination, wherein the at least one internal defect infrared image combination includes all possible combinations of the at least one internal defect infrared image;
[0028] Acquire a characteristic internal defect infrared image combination, wherein at least one internal defect infrared image can be obtained by superimposing at least one internal defect infrared image in the characteristic internal defect infrared image combination;
[0029] At least one internal defect infrared image in the characteristic internal defect infrared image combination is used as at least one basic internal defect infrared image.
[0030] Preferably, obtaining at least one component to be tested and pairing the component to be tested with the basic part structure comprises the following steps:
[0031] Divide the components in the tested mobile phone motherboard into at least one component area, where different component areas are not connected in the tested mobile phone motherboard;
[0032] Acquire a basic part structure combination, wherein the total area of the basic part structures in the basic part structure combination is equal to the area of the component region;
[0033] Forming at least one plane arrangement combination for the basic part structures in the basic part structure combination, wherein the at least one plane arrangement combination includes all arrangement situations of the basic part structures in the basic part structure combination;
[0034] Obtain a plane arrangement combination that is consistent with the shape of the component region as a feature plane arrangement combination;
[0035] The basic part structure in the feature plane arrangement combination is used as the part to be tested, and the basic part structure with the same shape is paired with the part to be tested.
[0036] Preferably, the surface identification of the component to be tested to obtain at least one actual basic external defect comprises the following steps:
[0037] Acquire at least one external defect combination, where the at least one external defect combination includes all combinations of at least one basic external defect;
[0038] Superimposing the basic external defect images of the basic external defects in the external defect combination onto the surface of the basic part structure corresponding to the component to be tested, to obtain an external image to be judged;
[0039] Acquire an external image to be judged that is consistent with the surface of the component to be tested as a feature external image to be judged;
[0040] The basic external defects in the external defect combination corresponding to the feature external image to be judged are taken as the actual basic external defects.
[0041] Preferably, the preprocessing of the actual infrared image to obtain the actual internal infrared image comprises the following steps:
[0042] Acquire a basic external defect infrared image corresponding to an actual basic external defect as a characteristic basic external defect infrared image;
[0043] All characteristic basic external defect infrared images are removed from the actual infrared image to obtain the actual internal infrared image, wherein the result of superimposing the actual internal infrared image and all characteristic basic external defect infrared images is the actual infrared image.
[0044] Preferably, obtaining at least one actual basic internal defect based on the actual internal infrared image comprises the following steps:
[0045] Acquire at least one internal defect combination, where the at least one internal defect combination includes all combinations of at least one basic internal defect;
[0046] Superimposing the basic internal defect infrared images of the basic internal defects in the internal defect combination to obtain an internal image to be judged;
[0047] Acquire an internal image to be judged that is consistent with the surface of the component to be tested as a feature internal image to be judged;
[0048] The basic internal defect in the internal defect combination corresponding to the image to be judged inside the feature is taken as the actual basic internal defect.
[0049] A mobile phone motherboard defect visual detection system based on AI technology, used to implement the above-mentioned mobile phone motherboard defect visual detection method based on AI technology, comprising:
[0050] A parts division module, wherein the parts division module divides at least one mobile phone mainboard into parts to obtain at least one basic parts structure;
[0051] An image acquisition module, wherein the image acquisition module acquires at least one external defect of a basic part structure and an external defect image;
[0052] An external defect acquisition module, wherein the external defect acquisition module obtains at least one basic external defect image based on at least one external defect image, takes the external defect corresponding to the basic external defect image as the basic external defect, and obtains an image with the basic external defect as the basic external defect infrared image under the condition that there is no defect inside the basic part structure;
[0053] An internal defect acquisition module, wherein the internal defect acquisition module acquires at least one internal defect of a basic part structure and an internal defect infrared image, and based on the at least one internal defect infrared image, screens and obtains at least one basic internal defect infrared image, and uses the internal defect corresponding to the basic internal defect infrared image as the basic internal defect;
[0054] A parts segmentation module, wherein the parts segmentation module performs parts segmentation on the motherboard of the mobile phone to be tested, obtains at least one part to be tested, and pairs the part to be tested with the basic part structure;
[0055] A defect recognition module, wherein the defect recognition module performs surface recognition on the component to be tested to obtain at least one actual basic external defect, performs infrared recognition on the component to be tested to obtain an actual infrared image, and pre-processes the actual infrared image based on the actual basic external defect to obtain an actual internal infrared image, wherein the actual internal infrared image is an infrared image of the interior of the component to be tested, and obtains at least one actual basic internal defect based on the actual internal infrared image;
[0056] A parameter adjustment module is used to adjust the production process of the component to be tested based on the actual basic external defects and the actual basic internal defects.
[0057] Compared with the prior art, the present invention has the following beneficial effects:
[0058] By setting up an external defect acquisition module, an internal defect acquisition module, a part segmentation module and a defect recognition module, it is possible to use limited basic defects to identify internal and external defects. During the recognition, ordinary images are used to identify external defects, and infrared images of external defects are obtained. Based on this, the overall actual infrared image is processed to remove the infrared image of external defects, thereby obtaining an infrared image containing only internal defects, and then the composition of internal defects can be determined. Therefore, according to the actual basic external defects and the actual basic internal defects, the production process of the parts to be tested can be adjusted in a targeted manner to eliminate errors in the production process, thereby ensuring the production quality of mobile phone motherboards. BRIEF DESCRIPTION OF THE DRAWINGS
[0059] Figure 1 It is a flow chart of the method for visually detecting defects of mobile phone motherboards based on AI technology of the present invention;
[0060] Figure 2 A schematic diagram of a process of dividing at least one mobile phone mainboard into parts to obtain at least one basic parts structure according to the present invention;
[0061] Figure 3 A schematic diagram of a process of screening and obtaining at least one basic external defect image based on at least one external defect image according to the present invention;
[0062] Figure 4 A schematic diagram of a process of screening and obtaining at least one basic internal defect infrared image based on at least one internal defect infrared image according to the present invention;
[0063] Figure 5 A schematic diagram of a process of obtaining at least one component to be tested and pairing the component to be tested with a basic part structure according to the present invention;
[0064] Figure 6 A schematic diagram of a process of performing surface recognition on a component to be tested to obtain at least one actual basic external defect according to the present invention;
[0065] Figure 7 A schematic diagram of the process of preprocessing an actual infrared image to obtain an actual internal infrared image according to the present invention;
[0066] Figure 8 The present invention is a schematic diagram of a process of obtaining at least one actual basic internal defect based on an actual internal infrared image. DETAILED DESCRIPTION
[0067] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.
[0068] Reference Figure 1 As shown, a method for visually detecting defects of mobile phone motherboards based on AI technology includes:
[0069] Dividing at least one mobile phone mainboard into parts to obtain at least one basic parts structure;
[0070] Acquire at least one external defect and an external defect image of a basic part structure;
[0071] Based on at least one external defect image, at least one basic external defect image is obtained by screening, and an external defect corresponding to the basic external defect image is used as a basic external defect;
[0072] Under the condition that there are no defects inside the basic part structure, an image with basic external defects is obtained as a basic external defect infrared image;
[0073] Acquire at least one internal defect of a basic part structure and an infrared image of the internal defect;
[0074] Based on at least one internal defect infrared image, at least one basic internal defect infrared image is obtained by screening, and the internal defect corresponding to the basic internal defect infrared image is used as the basic internal defect;
[0075] Segment the tested mobile phone motherboard into components to obtain at least one component to be tested, and pair the component to be tested with the basic part structure;
[0076] Perform surface identification on the tested component to obtain at least one actual basic external defect;
[0077] Perform infrared recognition on the parts to be tested to obtain the actual infrared image;
[0078] Based on the actual basic external defects, the actual infrared image is preprocessed to obtain the actual internal infrared image, wherein the actual internal infrared image is an infrared image of the interior of the component to be tested;
[0079] Based on the actual internal infrared image, at least one actual basic internal defect is obtained;
[0080] Based on the actual basic external defects and the actual basic internal defects, the production process of the parts to be tested is adjusted.
[0081] Reference Figure 2 As shown, dividing at least one mobile phone mainboard into parts to obtain at least one basic parts structure includes the following steps:
[0082] Divide the components in the mobile phone mainboard into minimum units according to the functions to be implemented, and obtain at least one preliminary component structure;
[0083] Repetitive preliminary part structures in at least one preliminary part structure are removed to obtain at least one basic part structure.
[0084] There are various electronic components installed on the mobile phone motherboard, and the main defects come from these electronic components. It is difficult to identify all the electronic components as a whole. Therefore, the electronic components are identified separately. There are a limited number of electronic components installed on the mobile phone motherboard. Therefore, they can be compared one by one through exhaustive method to complete the identification of defects.
[0085] Reference Figure 3 As shown, based on at least one external defect image, screening to obtain at least one basic external defect image includes the following steps:
[0086] Based on the at least one external defect image, forming at least one external defect image combination, wherein the at least one external defect image combination includes all possible combinations of the at least one external defect image;
[0087] Acquire a characteristic external defect image combination, wherein at least one external defect image can be obtained by superimposing at least one external defect image in the characteristic external defect image combination;
[0088] At least one external defect image in the characteristic external defect image combination is used as at least one basic external defect image.
[0089] The external defect is a single defect or a superposition of multiple defects, so at least one external defect image obtained contains an image of a single defect or a superposition of multiple defects. Therefore, at least one basic external defect image can be screened therefrom, and all external defect images can be obtained by combining several of the at least one basic external defect image. Therefore, when identifying external defects, a combination of basic external defect images can be used for identification.
[0090] Reference Figure 4 As shown, based on at least one internal defect infrared image, screening to obtain at least one basic internal defect infrared image includes the following steps:
[0091] Based on the at least one internal defect infrared image, forming at least one internal defect infrared image combination, wherein the at least one internal defect infrared image combination includes all possible combinations of the at least one internal defect infrared image;
[0092] Acquire a characteristic internal defect infrared image combination, wherein at least one internal defect infrared image can be obtained by superimposing at least one internal defect infrared image in the characteristic internal defect infrared image combination;
[0093] At least one internal defect infrared image in the characteristic internal defect infrared image combination is used as at least one basic internal defect infrared image.
[0094] The internal defect is a single defect or a superposition of multiple defects, so at least one internal defect infrared image obtained contains an image of a single defect or a superposition of multiple defects. Therefore, at least one basic internal defect infrared image can be screened therefrom, and all internal defect infrared images can be obtained by combining several of the at least one basic internal defect infrared image. Therefore, when identifying internal defects, a combination of basic internal defect infrared images can be used for identification.
[0095] Reference Figure 5 As shown, obtaining at least one component to be tested and pairing the component to be tested with a basic part structure includes the following steps:
[0096] Divide the components in the tested mobile phone motherboard into at least one component area, where different component areas are not connected in the tested mobile phone motherboard;
[0097] Acquire a basic part structure combination, wherein the total area of the basic part structures in the basic part structure combination is equal to the area of the component region;
[0098] Forming at least one plane arrangement combination for the basic part structures in the basic part structure combination, wherein the at least one plane arrangement combination includes all arrangement situations of the basic part structures in the basic part structure combination;
[0099] Obtain a plane arrangement combination that is consistent with the shape of the component region as a feature plane arrangement combination;
[0100] The basic part structure in the feature plane arrangement combination is used as the part to be tested, and the basic part structure with the same shape is paired with the part to be tested.
[0101] The workload of overall identification of components in the mobile phone motherboard to be tested is relatively huge, and it involves many combinations. Therefore, the components in the mobile phone motherboard to be tested are divided into at least one component area. Since at least one component area has no connection relationship with each other, they are independent of each other, and the identification between them will not interfere with each other. Therefore, they can be performed simultaneously, which can speed up the identification efficiency. When identifying, it is first necessary to identify the basic part structure contained in the part area, and the parts to be tested that constitute the part area are determined by exhaustive arrangement, wherein the parts to be tested are actually basic part structures.
[0102] Reference Figure 6 As shown, performing surface recognition on the component to be tested to obtain at least one actual basic external defect includes the following steps:
[0103] Acquire at least one external defect combination, where the at least one external defect combination includes all combinations of at least one basic external defect;
[0104] Superimposing the basic external defect images of the basic external defects in the external defect combination onto the surface of the basic part structure corresponding to the component to be tested, to obtain an external image to be judged;
[0105] Acquire an external image to be judged that is consistent with the surface of the component to be tested as a feature external image to be judged;
[0106] The basic external defects in the external defect combination corresponding to the feature external image to be judged are taken as the actual basic external defects.
[0107] The recognition of external defects is actually an existing technology. Without using the technical means in this solution, there are many other technical means to complete the recognition. The recognition of external defects is not the focus of this solution. The purpose of identifying external defects is to determine the basic external defect infrared image corresponding to the actual basic external defects through the actual basic external defects obtained. The actual infrared image contains both the actual internal infrared image and the basic external defect infrared image. Therefore, to obtain an infrared image containing only internal defects, it is necessary to remove the basic external defect infrared image. The removal method is to subtract the pixel value of the pixel point at the corresponding position of the basic external defect infrared image from the pixel value of the pixel point of the actual infrared image, and display the processed pixel point at the corresponding position of the actual infrared image to obtain the actual internal infrared image.
[0108] Reference Figure 7 As shown, preprocessing the actual infrared image to obtain the actual internal infrared image includes the following steps:
[0109] Acquire a basic external defect infrared image corresponding to an actual basic external defect as a characteristic basic external defect infrared image;
[0110] All characteristic basic external defect infrared images are removed from the actual infrared image to obtain the actual internal infrared image, wherein the result of superimposing the actual internal infrared image and all characteristic basic external defect infrared images is the actual infrared image.
[0111] Reference Figure 8 As shown, obtaining at least one actual basic internal defect based on the actual internal infrared image includes the following steps:
[0112] Acquire at least one internal defect combination, where the at least one internal defect combination includes all combinations of at least one basic internal defect;
[0113] Superimposing the basic internal defect infrared images of the basic internal defects in the internal defect combination to obtain an internal image to be judged;
[0114] Acquire an internal image to be judged that is consistent with the surface of the component to be tested as a feature internal image to be judged;
[0115] The basic internal defect in the internal defect combination corresponding to the image to be judged inside the feature is taken as the actual basic internal defect.
[0116] The process of obtaining at least one actual basic internal defect is actually carried out in an exhaustive manner. The actual internal infrared image is a combination of basic internal defects. Therefore, it is only necessary to obtain the combination of basic internal defects that is closest to the actual internal infrared image to obtain the actual basic internal defects. As a result, the production process of parts can be adjusted in a targeted manner to eliminate the actual basic internal defects and actual basic external defects of the parts.
[0117] A mobile phone motherboard defect visual detection system based on AI technology, used to implement the above-mentioned mobile phone motherboard defect visual detection method based on AI technology, comprising:
[0118] A parts division module, wherein the parts division module divides at least one mobile phone mainboard into parts to obtain at least one basic parts structure;
[0119] An image acquisition module, wherein the image acquisition module acquires at least one external defect of a basic part structure and an external defect image;
[0120] An external defect acquisition module, wherein the external defect acquisition module obtains at least one basic external defect image based on at least one external defect image, takes the external defect corresponding to the basic external defect image as the basic external defect, and obtains an image with the basic external defect as the basic external defect infrared image under the condition that there is no defect inside the basic part structure;
[0121] An internal defect acquisition module, wherein the internal defect acquisition module acquires at least one internal defect of a basic part structure and an internal defect infrared image, and based on the at least one internal defect infrared image, screens and obtains at least one basic internal defect infrared image, and uses the internal defect corresponding to the basic internal defect infrared image as the basic internal defect;
[0122] A parts segmentation module, wherein the parts segmentation module performs parts segmentation on the motherboard of the mobile phone to be tested, obtains at least one part to be tested, and pairs the part to be tested with the basic part structure;
[0123] A defect recognition module, wherein the defect recognition module performs surface recognition on the component to be tested to obtain at least one actual basic external defect, performs infrared recognition on the component to be tested to obtain an actual infrared image, and pre-processes the actual infrared image based on the actual basic external defect to obtain an actual internal infrared image, wherein the actual internal infrared image is an infrared image of the interior of the component to be tested, and obtains at least one actual basic internal defect based on the actual internal infrared image;
[0124] A parameter adjustment module is used to adjust the production process of the component to be tested based on the actual basic external defects and the actual basic internal defects.
[0125] Furthermore, the present solution also proposes a storage medium on which a computer-readable program is stored, and when the computer-readable program is called, the above-mentioned mobile phone motherboard defect visual detection method based on AI technology is executed.
[0126] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).
[0127] To sum up, the advantages of the present invention are: by setting an external defect acquisition module, an internal defect acquisition module, a part segmentation module and a defect recognition module, limited basic defects can be used to identify internal and external defects. During the recognition, ordinary images are used to identify external defects, and infrared images of external defects are obtained. Based on this, the overall actual infrared image is processed to remove the infrared image of the external defects, and then an infrared image containing only internal defects is obtained, and then the composition of the internal defects can be determined. Therefore, according to the actual basic external defects and the actual basic internal defects, the production process of the parts to be tested can be adjusted in a targeted manner to eliminate errors in the production process, thereby ensuring the production quality of mobile phone motherboards.
[0128] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.
Claims
1. A visual inspection method for mobile phone motherboard defects based on AI technology, characterized in that: include: Dividing at least one mobile phone mainboard into parts to obtain at least one basic parts structure; Acquire at least one external defect and an external defect image of a basic part structure; Based on at least one external defect image, at least one basic external defect image is obtained by screening, and an external defect corresponding to the basic external defect image is used as a basic external defect; Under the condition that there are no defects inside the basic part structure, an image with basic external defects is obtained as a basic external defect infrared image; Acquire at least one internal defect of a basic part structure and an infrared image of the internal defect; Based on at least one internal defect infrared image, at least one basic internal defect infrared image is obtained by screening, and the internal defect corresponding to the basic internal defect infrared image is used as the basic internal defect; Segment the tested mobile phone motherboard into components to obtain at least one component to be tested, and pair the component to be tested with the basic part structure; Perform surface identification on the tested component to obtain at least one actual basic external defect; Perform infrared recognition on the parts to be tested to obtain the actual infrared image; Based on the actual basic external defects, the actual infrared image is preprocessed to obtain the actual internal infrared image, wherein the actual internal infrared image is an infrared image of the interior of the component to be tested; Based on the actual internal infrared image, at least one actual basic internal defect is obtained; Based on the actual basic external defects and the actual basic internal defects, the production process of the parts to be tested is adjusted.
2. According to the method of visual detection of mobile phone motherboard defects based on AI technology in claim 1, it is characterized in that: The step of dividing at least one mobile phone mainboard into parts to obtain at least one basic parts structure comprises the following steps: Divide the components in the mobile phone mainboard into minimum units according to the functions to be implemented, and obtain at least one preliminary component structure; Repetitive preliminary part structures in at least one preliminary part structure are removed to obtain at least one basic part structure.
3. According to claim 2, a method for visually detecting defects of mobile phone motherboards based on AI technology is characterized in that: The method of screening and obtaining at least one basic external defect image based on at least one external defect image comprises the following steps: Based on the at least one external defect image, forming at least one external defect image combination, wherein the at least one external defect image combination includes all possible combinations of the at least one external defect image; Acquire a characteristic external defect image combination, wherein at least one external defect image can be obtained by superimposing at least one external defect image in the characteristic external defect image combination; At least one external defect image in the characteristic external defect image combination is used as at least one basic external defect image.
4. According to the method of visual detection of mobile phone motherboard defects based on AI technology in claim 3, it is characterized in that: The step of screening and obtaining at least one basic internal defect infrared image based on at least one internal defect infrared image comprises the following steps: Based on the at least one internal defect infrared image, forming at least one internal defect infrared image combination, wherein the at least one internal defect infrared image combination includes all possible combinations of the at least one internal defect infrared image; Acquire a characteristic internal defect infrared image combination, wherein at least one internal defect infrared image can be obtained by superimposing at least one internal defect infrared image in the characteristic internal defect infrared image combination; At least one internal defect infrared image in the characteristic internal defect infrared image combination is used as at least one basic internal defect infrared image.
5. The method for visually detecting defects of mobile phone motherboards based on AI technology according to claim 4 is characterized in that: The obtaining of at least one component to be tested and pairing the component to be tested with the basic part structure comprises the following steps: Divide the components in the tested mobile phone motherboard into at least one component area, where different component areas are not connected in the tested mobile phone motherboard; Acquire a basic part structure combination, wherein the total area of the basic part structures in the basic part structure combination is equal to the area of the component region; Forming at least one plane arrangement combination for the basic part structures in the basic part structure combination, wherein the at least one plane arrangement combination includes all arrangement situations of the basic part structures in the basic part structure combination; Obtain a plane arrangement combination that is consistent with the shape of the component region as a feature plane arrangement combination; The basic part structure in the feature plane arrangement combination is used as the part to be tested, and the basic part structure with the same shape is paired with the part to be tested.
6. The method for visually detecting defects of mobile phone motherboards based on AI technology according to claim 5 is characterized in that: The surface identification of the component to be tested to obtain at least one actual basic external defect comprises the following steps: Acquire at least one external defect combination, where the at least one external defect combination includes all combinations of at least one basic external defect; Superimposing the basic external defect images of the basic external defects in the external defect combination onto the surface of the basic part structure corresponding to the component to be tested, to obtain an external image to be judged; Acquire an external image to be judged that is consistent with the surface of the component to be tested as a feature external image to be judged; The basic external defects in the external defect combination corresponding to the feature external image to be judged are taken as the actual basic external defects.
7. The method for visually detecting defects of mobile phone motherboards based on AI technology according to claim 6 is characterized in that: The preprocessing of the actual infrared image to obtain the actual internal infrared image comprises the following steps: Acquire a basic external defect infrared image corresponding to an actual basic external defect as a characteristic basic external defect infrared image; All characteristic basic external defect infrared images are removed from the actual infrared image to obtain the actual internal infrared image, wherein the result of superimposing the actual internal infrared image and all characteristic basic external defect infrared images is the actual infrared image.
8. The method for visually detecting defects of mobile phone motherboards based on AI technology according to claim 7 is characterized in that: The obtaining of at least one actual basic internal defect based on the actual internal infrared image comprises the following steps: Acquire at least one internal defect combination, where the at least one internal defect combination includes all combinations of at least one basic internal defect; Superimposing the basic internal defect infrared images of the basic internal defects in the internal defect combination to obtain an internal image to be judged; Acquire an internal image to be judged that is consistent with the surface of the component to be tested as a feature internal image to be judged; The basic internal defect in the internal defect combination corresponding to the image to be judged inside the feature is taken as the actual basic internal defect.
9. A mobile phone motherboard defect visual detection system based on AI technology, used to implement the mobile phone motherboard defect visual detection method based on AI technology as described in any one of claims 1 to 8, characterized in that: include: A parts division module, wherein the parts division module divides at least one mobile phone mainboard into parts to obtain at least one basic parts structure; An image acquisition module, wherein the image acquisition module acquires at least one external defect of a basic part structure and an external defect image; An external defect acquisition module, wherein the external defect acquisition module obtains at least one basic external defect image based on at least one external defect image, takes the external defect corresponding to the basic external defect image as the basic external defect, and obtains an image with the basic external defect as the basic external defect infrared image under the condition that there is no defect inside the basic part structure; An internal defect acquisition module, wherein the internal defect acquisition module acquires at least one internal defect of a basic part structure and an internal defect infrared image, and based on the at least one internal defect infrared image, screens and obtains at least one basic internal defect infrared image, and uses the internal defect corresponding to the basic internal defect infrared image as the basic internal defect; A parts segmentation module, wherein the parts segmentation module performs parts segmentation on the motherboard of the mobile phone to be tested, obtains at least one part to be tested, and pairs the part to be tested with the basic part structure; A defect recognition module, wherein the defect recognition module performs surface recognition on the component to be tested to obtain at least one actual basic external defect, performs infrared recognition on the component to be tested to obtain an actual infrared image, and pre-processes the actual infrared image based on the actual basic external defect to obtain an actual internal infrared image, wherein the actual internal infrared image is an infrared image of the interior of the component to be tested, and obtains at least one actual basic internal defect based on the actual internal infrared image; A parameter adjustment module is used to adjust the production process of the component to be tested based on the actual basic external defects and the actual basic internal defects.
Citation Information
Cited By
Intelligent mobile phone defect detection method and system
CN120558973A
A method and system for detecting defects in a smartphone
CN120558973B
