Defect detection method and device, electronic equipment and computer readable storage medium

Abstract

The invention discloses a defect detection method and device, electronic equipment and a computer readable storage medium. The method comprises the steps that an ultrasonic module is controlled to transmit ultrasonic signals corresponding to multiple frequencies to a screen assembly, and actually measured ultrasonic images corresponding to the multiple frequencies are determined based on echo signals corresponding to the multiple frequencies and returned by the screen assembly; performing diffraction ring detection on the actually measured ultrasonic image corresponding to each frequency, and determining an actually measured diffraction ring detection result corresponding to each frequency; and determining a defect detection result of the screen assembly based on the actually measured diffraction ring detection results corresponding to the plurality of frequencies. Due to different interactions between the defects at different positions of the screen assembly and the ultrasonic signals at different frequencies, the shapes, positions, sizes and the like of the diffraction rings existing in the actually measured ultrasonic images formed by the ultrasonic signals at different frequencies are different, so that the detection results of the actually measured diffraction rings at a plurality of frequencies can be used for detecting the detection results of the diffraction rings at different frequencies. And the accuracy of defect detection results can be ensured.

Description

Technical Field

[0001] This application relates to the field of ultrasonic testing, and more particularly to a defect detection method, apparatus, electronic device, and computer-readable storage medium. Background Technology

[0002] Ultrasonic fingerprint systems primarily transmit ultrasonic signals to the screen assembly via an ultrasonic module. The reflected echoes from the screen assembly form an ultrasonic fingerprint image. Specifically, the image is formed based on the difference in sound wave intensity reflected from the finger's contact surface with the screen assembly, and the difference in sound wave intensity reflected from the air's contact surface with the screen assembly. Because the ultrasonic signals emitted by the ultrasonic module have a certain degree of penetrability, they can penetrate the skin to identify the unique three-dimensional features of a fingerprint. They are less affected by fingerprint surface details and can still be recognized even when the finger is slightly dirty or moist. Therefore, ultrasonic fingerprint recognition technology is gaining increasing attention. After the screen assembly is manufactured, various defect inspections are required to ensure that the screen assembly meets quality requirements. Current defect inspection processes for screen assemblies typically use electrical performance testing or charge-coupled devices (CCDs) to acquire optical images, which suffers from low accuracy. Summary of the Invention

[0003] This application provides a defect detection method, apparatus, electronic device, and computer-readable storage medium to address the problem of low accuracy in existing screen component defect detection processes.

[0004] A defect detection method, comprising: The ultrasonic module is controlled to emit ultrasonic signals corresponding to multiple frequencies to the screen component. Based on the echo signals corresponding to multiple frequencies returned by the screen component, the measured ultrasonic images corresponding to multiple frequencies are determined. Diffraction ring detection is performed on the measured ultrasonic image corresponding to each frequency to determine the measured diffraction ring detection result for each frequency. Based on the measured diffraction ring detection results corresponding to multiple frequencies, the defect detection results of the screen component are determined.

[0005] Preferably, the step of performing diffraction ring detection on the measured ultrasonic image corresponding to each frequency, and determining the measured diffraction ring detection result for each frequency, includes: For each frequency, feature recognition is performed on the measured ultrasonic image to determine all abnormal feature points. The abnormal feature points are feature points where the absolute value of the difference between the measured signal intensity and the mean signal intensity is greater than a preset intensity difference. The mean signal intensity is the mean of the measured signal intensity of all feature points in the measured ultrasonic image. Determine the fitted shape for all the aforementioned anomalous feature points; When the fitted shape is ring-shaped or ring-like, the measured diffraction ring detection result corresponding to the frequency is determined to indicate the presence of a diffraction ring. When the fitted shape is neither circular nor quasi-circular, the measured diffraction ring detection result corresponding to the frequency is determined to be that no diffraction ring exists.

[0006] Preferably, the step of performing diffraction ring detection on the measured ultrasonic image corresponding to each frequency, and determining the measured diffraction ring detection result for each frequency, includes: A bandpass filtering algorithm is used to filter the measured ultrasonic image corresponding to each frequency to determine the valid ultrasonic image corresponding to the frequency. Diffraction ring detection is performed on the effective ultrasonic image corresponding to the frequency to determine the measured diffraction ring detection result corresponding to the frequency.

[0007] Preferably, determining the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies includes: Based on the measured diffraction ring detection results corresponding to multiple frequencies, the defect location mapping table corresponding to the screen component is queried to determine the defect detection result of the screen component. The defect location mapping table is used to reflect whether diffraction rings exist at different defect locations at different frequencies.

[0008] Preferably, determining the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies includes: When the measured diffraction ring detection result corresponding to the frequency indicates the presence of a defective ring, the current defect feature corresponding to the frequency is determined. The current defect feature includes the diffraction ring radius and the diffraction ring signal difference. The diffraction ring signal difference is determined based on the measured signal intensity corresponding to all the feature points within the diffraction ring. Based on the current defect features corresponding to all the frequencies and the pre-trained defect localization model, the defect location corresponding to the screen component is determined.

[0009] Preferably, the defect detection method further includes: When the measured diffraction ring detection result corresponding to the frequency indicates the presence of a defective ring, the current defect feature corresponding to the frequency is determined. The current defect feature includes the diffraction ring radius and the diffraction ring signal difference. The diffraction ring signal difference is determined based on the measured signal intensity corresponding to all the feature points within the diffraction ring. Based on the current defect characteristics corresponding to each frequency, the defect level corresponding to the diffraction ring detected at each frequency is determined. The defect level of the screen component is determined based on the defect level corresponding to the diffraction rings detected at all the frequencies.

[0010] Preferably, the difference in the diffraction ring signal is the difference between the first characteristic mean and the second characteristic mean; The first feature mean is the mean of the M largest measured signal intensities among all the measured signal intensities corresponding to the feature points within the diffraction ring; The second characteristic mean is the mean of the M smaller measured signal intensities among all the characteristic points within the diffraction ring; M ≥ 1.

[0011] A defect detection device, comprising: The measured image acquisition module is used to control the ultrasonic module to emit ultrasonic signals corresponding to multiple frequencies to the screen component, and to determine the measured ultrasonic images corresponding to multiple frequencies based on the echo signals corresponding to multiple frequencies returned by the screen component. The diffraction ring detection module is used to perform diffraction ring detection on the measured ultrasonic image corresponding to each frequency, and to determine the measured diffraction ring detection result corresponding to each frequency. The defect detection result determination module is used to determine the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies.

[0012] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned defect detection method.

[0013] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described defect detection method.

[0014] This application provides a defect detection method, apparatus, electronic device, and computer-readable storage medium. Because the interaction between defects at different locations of a screen component and ultrasonic signals of different frequencies varies, the shape, position, and size of diffraction rings in the measured ultrasonic images formed by ultrasonic signals of different frequencies differ. By performing diffraction ring detection on the measured ultrasonic images corresponding to each frequency, the measured diffraction ring detection results for multiple frequencies are determined. Based on the measured diffraction ring detection results for multiple frequencies, it is possible to accurately determine whether a defect exists in the screen component. When a defect exists, the location and level of the defect are determined to ensure the accuracy of the defect detection results. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a defect detection method in an embodiment of this application; Figure 2 yes Figure 1 A flowchart of step S102; Figure 3 yes Figure 1 Another flowchart of step S102; Figure 4 yes Figure 1 A flowchart of step S103; Figure 5 yes Figure 1 Another flowchart following step S103; Figure 6 This is a schematic diagram of a screen component in an embodiment of this application; Figure 7 This is a schematic diagram of measured ultrasonic images formed by the OC layer and the screen at low, medium and high frequencies in an embodiment of this application; Figure 8 This is a schematic diagram of a defect location mapping table corresponding to the screen component in an embodiment of this application; Figure 9 This is a schematic diagram of a defect detection device in an embodiment of this application. Detailed Implementation

[0017] To make the technical problems, technical solutions, and beneficial effects solved by this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0018] This application provides a defect detection method applicable to detecting defects in terminal devices integrating ultrasonic modules and screen components, and determining the defect detection results. The ultrasonic module is located below the screen component and includes, but is not limited to, an ultrasonic piezoelectric sensor, an analog-to-digital converter, a storage unit, and a data processing center. The ultrasonic piezoelectric sensor emits ultrasonic signals, receives echo signals, converts the echo signals into digital signals via analog-to-digital conversion, and stores them in the storage unit. The data processing center processes the digital signals stored in the storage unit to determine the ultrasonic image.

[0019] like Figure 6As shown, the screen assembly includes a screen and multiple stacked layers sequentially attached beneath the screen. These stacked layers are structural layers positioned between the screen and the ultrasonic module. For example, the stacked layers can be, but are not limited to, optical coatings (OC layers), Si layers, and other structural layers. The Si layer is a structural layer related to Si, and other structural layers refer to structural layers formed of materials other than the OC layer and the Si layer.

[0020] The defect detection method provided in this application is applicable to electronic devices. The electronic device is connected to an ultrasonic module disposed below a screen assembly. The ultrasonic module is used to control the ultrasonic module to emit ultrasonic signals corresponding to multiple frequencies to the screen assembly and to receive the echo signals returned by the screen assembly. Based on the echo signal corresponding to each frequency, a measured ultrasonic image corresponding to that frequency is determined. Defect detection is then performed on the measured ultrasonic images of multiple frequencies to determine whether a defect exists in the screen assembly. If a defect exists, the defect location and defect level are determined, forming a defect detection result. This enables quality inspection of the screen assembly. Furthermore, different defect detection results can be statistically analyzed to analyze the related processing technology and control parameters, thereby leading to process improvements. In this example, a "defect" refers to an irregular structure or abnormal morphology appearing inside the material between the screen assembly and the ultrasonic module.

[0021] This application provides a defect detection method, which will be described using an electronic device connected to an ultrasonic module as an example. Figure 1 As shown, the defect detection methods include: S101: Control the ultrasonic module to emit ultrasonic signals corresponding to multiple frequencies to the screen component, and determine the measured ultrasonic images corresponding to multiple frequencies based on the echo signals corresponding to multiple frequencies returned by the screen component. S102: Perform diffraction ring detection on the measured ultrasonic image corresponding to each frequency, and determine the measured diffraction ring detection result for each frequency. S103: Based on the measured diffraction ring detection results corresponding to multiple frequencies, determine the defect detection results of the screen component.

[0022] Among them, the diffraction ring in ultrasound images is a ring or quasi-ring signal feature formed on the ultrasound image centered on the obstacle when ultrasound waves encounter obstacles (such as tissue interfaces, micro-lesions, bubbles, etc.) during propagation.

[0023] As an example, in step S101, the electronic device can control the ultrasonic module to emit ultrasonic signals of multiple frequencies to the screen assembly. Each frequency's ultrasonic signal passes through multiple layers to the screen, penetrates the screen, and reaches the air. After transmission and reflection through the multiple layers and the screen, the ultrasonic signal forms an echo signal corresponding to that frequency. Upon receiving the echo signal, the ultrasonic module processes it to determine the measured ultrasonic image corresponding to that frequency. Here, the measured ultrasonic image is the actually detected ultrasonic image.

[0024] As an example, in step S102, the electronic device can use a pre-set diffraction ring detection algorithm to perform diffraction ring detection on the measured ultrasonic image corresponding to each frequency, in order to determine whether a diffraction ring exists in the measured ultrasonic image corresponding to that frequency, and to determine the measured diffraction ring detection result corresponding to that frequency. Here, the measured diffraction ring detection result can be either that a diffraction ring exists or that a diffraction ring does not exist. The diffraction ring detection algorithm refers to an algorithm that can identify diffraction rings. It can be, but is not limited to, traditional algorithms based on image features and intelligent algorithms based on deep learning. For example, a traditional algorithm first separates the diffraction ring from the background through threshold segmentation (such as Otsu adaptive thresholding, fixed thresholding), eliminates low grayscale noise, and then uses edge detection operators (such as Canny, Sobel, Robert) to extract the contour edge of the diffraction ring. Finally, it determines the diffraction ring through circle fitting (such as least squares method, Hough circle transform).

[0025] As an example, in step S103, when the electronic device acquires the measured diffraction ring detection results corresponding to multiple frequencies, it performs combined analysis on the measured diffraction ring detection results corresponding to multiple frequencies to determine whether there is a defect in the screen component. If a defect exists, it determines the defect location and defect level, forming a defect detection result for the screen component, thereby achieving quality inspection of the screen component. Here, the defect location specifically indicates the layer of the screen component where the defect is located; for example, determining whether the defect is in the screen or the OC layer.

[0026] In this example, the electronic device pre-stores detailed information on the presence of diffraction rings in ultrasonic images of different frequencies when defects exist in different locations of the screen assembly. When defect detection is required on the screen assembly before shipment, the pre-stored detailed information can be queried based on the measured diffraction ring detection results corresponding to the measured ultrasonic images of multiple frequencies. The defect information recorded in the detailed information that matches the measured diffraction ring detection results of multiple frequencies is determined as the defect detection result of the screen assembly. For example, as... Figure 7As shown, when the measured ultrasonic images corresponding to low frequency, mid frequency, and high frequency are (A), (B), and (C), respectively, it can be determined that the measured diffraction ring detection results for low frequency, mid frequency, and high frequency are: no diffraction ring, presence of diffraction ring (larger radius), and presence of diffraction ring (smaller radius), respectively. Querying the detailed information, it can be determined that the defect detection result of the screen component is a defect in the OC layer. Similarly, when the measured ultrasonic images corresponding to low frequency, mid frequency, and high frequency are (D), (E), and (F), it can be determined that the measured diffraction ring detection results for low frequency, mid frequency, and high frequency are: no diffraction ring, presence of diffraction ring (smaller radius), and presence of diffraction ring (larger radius), respectively. Querying the detailed information, it can be determined that the defect detection result of the screen component is a defect in the screen.

[0027] In this embodiment, since the interaction between defects at different locations of the screen component and ultrasonic signals of different frequencies is different, the shape, position, and size of the diffraction rings in the measured ultrasonic images formed by ultrasonic signals of different frequencies are different. By performing diffraction ring detection on the measured ultrasonic images corresponding to each frequency, the measured diffraction ring detection results of multiple frequencies can be determined. Based on the measured diffraction ring detection results of multiple frequencies, it is possible to accurately determine whether there are defects in the screen component. When defects exist, the location and level of defects can be determined to ensure the accuracy of the defect detection results.

[0028] In one embodiment, such as Figure 2 As shown, step S102, which involves performing diffraction ring detection on the measured ultrasonic image corresponding to each frequency and determining the measured diffraction ring detection result for each frequency, includes: S201: Perform feature recognition on the measured ultrasonic image corresponding to each frequency to determine all abnormal feature points. Abnormal feature points are feature points where the absolute value of the difference between the measured signal intensity and the mean signal intensity is greater than a preset intensity difference. The mean signal intensity is the mean of the measured signal intensity of all feature points in the measured ultrasonic image. S202: Determine the fitted shape for all anomalous feature points; S203: When the fitted shape is ring-shaped or quasi-ring-shaped, the measured diffraction ring detection result corresponding to the determined frequency indicates the presence of a diffraction ring; S204: When the fitted shape is neither circular nor quasi-circular, the measured diffraction ring detection result corresponding to the determined frequency is that no diffraction ring exists.

[0029] Among them, anomalous feature points are those that differ significantly from most feature points in the measured ultrasound image. Diffraction ring evaluation conditions are pre-set conditions used to assess whether a diffraction ring has formed. For example, the diffraction ring evaluation condition is whether all anomalous feature points can fit into a ring or a ring-like structure.

[0030] As an example, the electronic device performs feature recognition on the measured ultrasonic image corresponding to each frequency to determine each feature point in the measured ultrasonic image. Corresponding measured signal strength Then, the mean of the measured signal intensity corresponding to all feature points in the measured ultrasonic image is calculated and determined as the mean signal intensity. ,Right now , The first step is to determine the total number of feature points in the measured ultrasound image; then, the number of each feature point is calculated. Corresponding measured signal strength With the mean signal strength absolute value of the difference Finally, the absolute value of the difference is... Difference from preset strength The comparison is performed when the absolute value of the difference is greater than the preset strength difference (i.e., When identifying feature points These are abnormal feature points; the absolute value of the difference is not greater than the preset intensity difference (i.e., ... When identifying feature points Not an anomalous feature point. In this example, in a black and white ultrasound image, the measured signal intensity corresponding to each feature point can be quantized as the pixel grayscale value of each feature point; in a color ultrasound image, the measured signal intensity corresponding to each feature point can be quantized as the color brightness of each feature point.

[0031] As an example, in step S202, after determining all abnormal feature points in each measured ultrasonic image, the electronic device needs to perform curve fitting on all abnormal feature points to determine the fitted shape formed by all abnormal feature points. In this example, the electronic device can use curve fitting algorithms such as genetic algorithms and particle swarm optimization to fit all abnormal feature points and determine the fitted shape of all abnormal feature points.

[0032] As an example, in step S203, when the fitted shape of all abnormal feature points of the electronic device is circular or quasi-circular, it is determined that the measured ultrasonic image corresponding to that frequency has a diffraction ring. At this time, it can be concluded that the ultrasonic signal based on that frequency detected a diffraction ring formed around the defect in the screen assembly.

[0033] As an example, in step S204, when the fitted shape of all abnormal feature points of the electronic device is neither circular nor quasi-circular, it is determined that the measured ultrasonic image corresponding to that frequency does not have a diffraction ring. The abnormal feature point is more likely to be noise interference rather than an abnormal situation caused by defects in the screen component itself, and no further processing is required.

[0034] In this embodiment, feature recognition is first performed on the measured ultrasonic image corresponding to each frequency to identify abnormal feature points where the measured signal intensity is significantly different from other feature points; then, based on whether the fitted shape of all abnormal feature points is circular or quasi-circular, the effect of evaluating whether there is a diffraction ring in the measured ultrasonic image corresponding to that frequency is achieved.

[0035] In one embodiment, such as Figure 3 As shown, step S102, which involves performing diffraction ring detection on the measured ultrasonic image corresponding to each frequency and determining the measured diffraction ring detection result for each frequency, includes: S301: Use a bandpass filtering algorithm to filter the measured ultrasonic image corresponding to each frequency to determine the valid ultrasonic image corresponding to the frequency. S302: Perform diffraction ring detection on the effective ultrasonic image corresponding to the frequency, and determine the measured diffraction ring detection result corresponding to the frequency.

[0036] Among them, the effective ultrasound image is the ultrasound image after filtering the measured ultrasound image.

[0037] As an example, in step S301, after acquiring measured ultrasonic images corresponding to multiple frequencies, the electronic device needs to use a bandpass filtering algorithm to filter the measured ultrasonic images corresponding to each frequency to determine the valid ultrasonic image corresponding to that frequency. Since the measured ultrasonic image is an image formed by the echo signal returned after the interaction (reflection, scattering, or transmission) between the ultrasonic signal of a specific frequency and the screen component, only the echo signal with a frequency close to or within its harmonic range will carry the structural characteristics of the screen component. Therefore, the valid signal in the ultrasonic image is limited to a specific frequency range. Using a bandpass filtering algorithm to filter the measured ultrasonic image can retain only the valid signal within that specific frequency range and filter out all interference outside that specific frequency range, so that the determined valid ultrasonic image can accurately retain the valid signal that can reflect whether there are defects in the screen component.

[0038] As an example, in step S302, the electronic device can use a pre-set diffraction ring detection algorithm to perform diffraction ring detection on the valid ultrasonic image corresponding to each frequency, in order to determine whether a diffraction ring exists in the valid ultrasonic image corresponding to that frequency, and thus determine the measured diffraction ring detection result for that frequency. In this example, since all interference outside a specific frequency range is filtered out from the valid ultrasonic image, the accuracy of the measured diffraction ring detection result for that frequency can be guaranteed. Understandably, the process of the electronic device performing diffraction ring detection on the valid ultrasonic image corresponding to each frequency is similar to steps S201-S203 above, and will not be described in detail here to avoid repetition.

[0039] In this embodiment, when there is a defect in the screen component, the measured ultrasonic image is processed by a bandpass filtering algorithm, which makes the filtered effective ultrasonic image able to highlight the diffraction ring formed by the defect. Then, the effective ultrasonic image corresponding to the frequency is subjected to diffraction ring detection to ensure the accuracy of the measured diffraction ring detection result corresponding to the determined frequency.

[0040] In one embodiment, step S103, which is to determine the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies, includes: Based on the measured diffraction ring detection results corresponding to multiple frequencies, the defect location mapping table corresponding to the screen component is queried to determine the defect detection result of the screen component. Among them, the defect location mapping table is used to reflect whether diffraction rings exist at different defect locations at different frequencies.

[0041] As an example, during the manufacturing process of any screen component, it is necessary to test screen components with defects in different locations (this can be simulation testing and / or real testing). The ultrasonic module is controlled to emit ultrasonic signals of different frequencies. Based on the echo signals of different frequencies, test ultrasonic images corresponding to different frequencies are determined. Diffraction ring detection is performed on the test ultrasonic images of different frequencies to determine the test diffraction ring detection results. These results reflect the presence of diffraction rings in the test ultrasonic images. Then, the defect locations are associated with and stored with the test diffraction ring detection results of different frequencies, forming a defect location mapping table corresponding to that screen component. In this example, because the materials and thicknesses of the screens and layers of different screen components are different, their acoustic impedances are different, resulting in different interactions between the ultrasonic signals of different frequencies and different screen components. This leads to differences in the defect detection results for different defect locations. Therefore, a corresponding defect location mapping table needs to be determined for each type of screen component.

[0042] like Figure 8 As shown, when the screen assembly includes, but is not limited to, the screen, OC layer, Si layer, and other structural layers, and when a defect exists in the OC layer or screen layer of the screen, the ultrasonic module is controlled to emit ultrasonic signals of 10.9MHz, 11.1MHz, 11.4MHz, 11.7MHz, 11.9MHz, 12.0MHz, 12.2MHz, 12.5MHz, 12.9MHz, 13.1MHz, 13.3MHz, 13.8MHz, and 13.9MHz. Diffraction ring detection is performed on the test ultrasonic images formed by the echo signals of different frequencies. When the test diffraction ring detection result indicates the presence of a diffraction ring, it can be represented by 1; when the test diffraction ring detection result indicates the absence of a diffraction ring, it can be represented by 0. The defect location is associated with and stored with the test diffraction ring detection results of different frequencies to form a defect location mapping table corresponding to the screen assembly.

[0043] As an example, an electronic device can form a query array based on the measured diffraction ring detection results corresponding to multiple frequencies. This query array is determined based on the characterization values ​​of the measured diffraction ring detection results corresponding to multiple frequencies arranged in sequence. When the measured diffraction ring detection result indicates the presence of a diffraction ring, its characterization value is represented by 1; when the measured diffraction ring detection result indicates the absence of a diffraction ring, its characterization value is represented by 0. Then, by querying the defect location mapping table corresponding to the same screen component based on the query array, the defect detection result of the screen component can be quickly determined. In this example, the defect detection result of the screen component includes whether the screen component has a defect. If the screen component has a defect, the defect location can also be determined to quickly identify which layer of the screen component has the defect.

[0044] In one embodiment, such as Figure 4 As shown, step S103, which is to determine the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies, includes: S401: When the measured diffraction ring detection result at the corresponding frequency indicates the presence of a defective ring, determine the current defect characteristics at the corresponding frequency. The current defect characteristics include the diffraction ring radius and the diffraction ring signal difference. The diffraction ring signal difference is determined based on the measured signal intensity corresponding to all feature points within the diffraction ring. S402: Based on the current defect features corresponding to all frequencies and the pre-trained defect localization model, determine the defect location corresponding to the screen component.

[0045] Among them, the defect localization model is a model that learns classification based on a neural network in advance to achieve defect localization. As an example, the defect localization model training process may include the following steps: (1) Obtain model training samples, which include training defect locations and training defect features corresponding to multiple frequencies. The training defect locations are the real defect locations determined during model training, and the training defect features are the defect features determined during model training. The training defect features include diffraction ring radius and diffraction ring signal difference. The process of determining the diffraction ring radius and diffraction ring signal difference here is the same as step S401 and will not be repeated here; (2) Input the training defect features corresponding to multiple frequencies into the neural network model to determine the corresponding predicted defect locations; (3) Calculate the loss based on the training defect locations and predicted defect locations corresponding to the same model training samples to determine the target loss corresponding to the neural network model; (4) When the target loss of the neural network model is greater than the preset loss, update the model parameters of the neural network model and repeat (2)-(3); (5) When the target loss of the neural network model is not greater than the preset loss, determine the last updated neural network model as the trained defect localization model.

[0046] As an example, in step S401, after the electronic device determines that the measured diffraction ring detection result for each frequency indicates the presence of a diffraction ring, it needs to perform feature analysis on the diffraction ring in the measured ultrasonic image to determine the diffraction ring radius and the diffraction ring signal difference. The diffraction ring radius is the radius of the identified diffraction ring. The diffraction ring signal difference is a signal characteristic reflecting the overall situation of the diffraction ring. This diffraction ring signal difference is determined based on the measured signal intensity corresponding to all feature points within the diffraction ring, specifically reflecting the difference in the measured signal intensity of all feature points within the diffraction ring.

[0047] As an example, in step S402, after determining the current defect features corresponding to all frequencies, the electronic device can input the current defect features corresponding to all frequencies into a pre-trained defect localization model for identification, so as to determine the defect location corresponding to the screen component. The defect location can be located to a coarse location in different layers (e.g., to the OC layer), or only to a detailed location within different layers (e.g., to 1 / 2 of the OC layer). The specific location depends on the localization accuracy of the defect localization model training process.

[0048] In this example, the detection of defect locations in the screen components mainly relies on the analysis of a large amount of data. For example, the diffraction ring detection results of the OC layer at various frequencies are: present at low frequencies, present at mid frequencies, and present at high frequencies; the diffraction ring radii are 3, 4, and 5; the diffraction ring signal differences are A1, A2, and A3. The diffraction ring detection results of screen defects are: absent at low frequencies, present at mid frequencies, and present at high frequencies; the diffraction ring radii are 0, 3, and 5; the diffraction ring signal differences are B1, B2, and B3. The specific processing methods are as follows: First, based on the relevant data corresponding to a large number of OC layer defects and screen defects, a defect location mapping table corresponding to the screen components can be formed. This allows for subsequent lookup of the defect location mapping table based on the measured diffraction ring detection results corresponding to multiple frequencies, which can quickly determine the defect detection results of the screen components. This defect detection module can quickly determine which layer of the screen component the defect is located on, thus determining the approximate location. Secondly, a model training sample can be formed based on a large amount of data related to OC layer defects and screen defects. Based on this model training sample, a defect localization model can be trained so that the defect localization model can locate the current defect features corresponding to multiple frequencies, thereby locating the detailed location of the defect within different layers.

[0049] In one embodiment, such as Figure 5 As shown, the method also includes: S501: When the measured diffraction ring detection result at the corresponding frequency indicates the presence of a defective ring, determine the current defect characteristics at the corresponding frequency. The current defect characteristics include the diffraction ring radius and the diffraction ring signal difference. The diffraction ring signal difference is determined based on the measured signal intensity corresponding to all feature points within the diffraction ring. S502: Based on the current defect characteristics corresponding to each frequency, determine the defect level corresponding to the diffraction ring detected at each frequency; S503: Determine the defect level of the screen component based on the defect levels corresponding to the diffraction rings detected at all frequencies.

[0050] As an example, in step S501, after the electronic device determines that the measured diffraction ring detection result for each frequency indicates the presence of a diffraction ring, it needs to perform feature analysis on the diffraction ring in the measured ultrasonic image to determine the diffraction ring radius and the diffraction ring signal difference. The diffraction ring radius is the radius of the identified diffraction ring. The diffraction ring signal difference is a signal characteristic reflecting the overall situation of the diffraction ring. This diffraction ring signal difference is determined based on the measured signal intensity corresponding to all feature points within the diffraction ring, specifically reflecting the difference in the measured signal intensity of all feature points within the diffraction ring.

[0051] As an example, in step S502, after determining the diffraction ring radius and diffraction ring signal difference corresponding to each frequency, the electronic device needs to compare the diffraction ring radius with a preset radius and the diffraction ring signal difference with a preset signal difference. When the diffraction ring radius is greater than the preset radius and the diffraction ring signal difference is greater than the preset signal difference, it indicates that the formed diffraction ring radius is large and the signal difference inside the diffraction ring is large, indicating that the defect in the screen assembly is serious. Therefore, the defect level corresponding to the diffraction ring detected at that frequency can be determined to be a serious level. When the diffraction ring radius is not greater than the preset radius or the diffraction ring signal difference is not greater than the preset signal difference, the defect level corresponding to the diffraction ring detected at that frequency is determined to be a normal level.

[0052] As an example, in step S503, after determining the defect levels corresponding to the diffraction rings detected at multiple frequencies, the electronic device can perform statistical analysis on the defect levels corresponding to the diffraction rings detected at multiple frequencies to determine the defect level corresponding to the screen component. In this example, the electronic device emits ultrasonic signals corresponding to K frequencies to the screen component, determines the measured ultrasonic images corresponding to K frequencies, and among the measured ultrasonic images corresponding to K frequencies, the defect level corresponding to the diffraction rings detected in the measured ultrasonic images corresponding to K frequencies is of the severe level. Then, the measured proportion (i.e., k / K) of the number of frequencies with the severe level defect level is calculated. Based on the measured proportion, the defect level corresponding to the screen component is determined. For example, when a preset proportion is set, if the measured proportion is greater than the preset proportion, the defect level of the screen component is determined to be of the severe level; if the measured proportion is not greater than the preset proportion, the defect level of the screen component is determined to be of the normal level. As another example, when at least two preset proportions are set, at least three proportion intervals are determined based on the at least two preset proportions, and each proportion interval corresponds to a preset level, the preset level of the proportion area where the measured proportion is located can be determined as the defect level of the screen component.

[0053] In this embodiment, the defect level corresponding to the diffraction ring detected at a single frequency is determined based on the diffraction ring radius and diffraction ring signal difference detected by the measured ultrasonic image corresponding to each frequency; then, based on the defect levels corresponding to the diffraction rings detected at all frequencies, the defect level of the screen component is comprehensively determined to ensure the accuracy of the defect level corresponding to the screen component, so that it can comprehensively reflect the overall performance of the screen component.

[0054] Furthermore, after determining the defect detection results and defect level of the screen component, the electronic device can further conduct big data analysis on the processing data of the screen component corresponding to the severe defect level, so as to subsequently analyze the process flow and / or process parameters that cause the severe defect in the screen component, so as to improve the analyzed process flow and / or process parameters, thereby improving the efficiency of defect cause analysis and process improvement.

[0055] In one embodiment, the difference in the diffraction ring signal is the difference between the first characteristic mean and the second characteristic mean; The first characteristic mean is the mean of the M largest measured signal intensities among all the measured signal intensities corresponding to all characteristic points within the diffraction ring; The second characteristic mean is the mean of the M smaller measured signal intensities among all the characteristic points within the diffraction ring; M≥1.

[0056] As an example, after determining that a diffraction ring exists at each frequency, the electronic device needs to sort the measured signal intensities of all feature points within the diffraction ring from largest to smallest. Then, it calculates the mean of the top M measured signal intensities with the largest intensities, defining it as the first characteristic mean; and calculates the mean of the next M measured signal intensities with the smallest intensities, defining it as the second characteristic mean. Finally, the difference between the first and second characteristic means is determined as the diffraction ring signal difference for that frequency. In this example, the diffraction ring signal difference reflects the overall difference between the M largest and M smallest measured signal intensities, comprehensively reflecting the signal differences within the diffraction ring and thus indicating the defect level of the diffraction ring.

[0057] In this example, the quantity M can be directly determined as a fixed value, or it can be determined based on the number of all feature points in the diffraction ring and a preset ratio, where the preset ratio is a pre-set proportion. For example, if the number of all feature points in the diffraction ring is 100, and the preset ratio is 20%, then the average of the 20 larger measured signal intensities is determined as the first feature average, and the average of the 20 smaller measured signal intensities is determined as the second feature average. The difference between the first and second feature averages is then determined as the diffraction ring signal difference. For example, Figure 7In the diffraction ring shown in (F), the greater the measured signal intensity, the whiter it appears; the smaller the measured signal intensity, the blacker it appears. The average of the first 20% of the measured signal intensities is determined as the first characteristic average, and the average of the last 20% of the measured signal intensities is determined as the second characteristic average. The difference between the first characteristic average and the second characteristic average is determined as the diffraction ring signal difference. This diffraction ring signal difference can comprehensively reflect the characteristic differences between the white area outside the diffraction ring and the black area inside the diffraction ring, and can effectively reflect the severity of the defect.

[0058] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0059] In one embodiment, a defect detection device is provided, which corresponds one-to-one with the defect detection methods described in the above embodiments. For example... Figure 9 As shown, the defect detection device includes: The measured image acquisition module 901 is used to control the ultrasonic module to transmit ultrasonic signals corresponding to multiple frequencies to the screen component, and to determine the measured ultrasonic images corresponding to multiple frequencies based on the echo signals corresponding to multiple frequencies returned by the screen component. The diffraction ring detection module 902 is used to perform diffraction ring detection on the measured ultrasonic image corresponding to each frequency and determine the measured diffraction ring detection result corresponding to each frequency. The defect detection result determination module 903 is used to determine the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies.

[0060] Specific limitations regarding the defect detection device can be found in the limitations of the defect detection method described above, and will not be repeated here. Each module in the aforementioned defect detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of the electronic device in hardware form or independently of it, or stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of each module.

[0061] In one embodiment, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the defect detection method described in the above embodiment, for example... Figure 1 S101-S103 are shown below. To avoid repetition, they will not be described again here.

[0062] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When executed by a processor, the computer program implements the defect detection method described in the above embodiment, for example... Figure 1 S101-S103 are shown below. To avoid repetition, they will not be described again here.

[0063] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A defect detection method, characterized in that, include: The ultrasonic module is controlled to emit ultrasonic signals corresponding to multiple frequencies to the screen component. Based on the echo signals corresponding to multiple frequencies returned by the screen component, the measured ultrasonic images corresponding to multiple frequencies are determined. Diffraction ring detection is performed on the measured ultrasonic image corresponding to each frequency to determine the measured diffraction ring detection result for each frequency. Based on the measured diffraction ring detection results corresponding to multiple frequencies, the defect detection results of the screen component are determined.

2. The defect detection method according to claim 1, characterized in that, The step of performing diffraction ring detection on the measured ultrasonic image corresponding to each frequency, and determining the measured diffraction ring detection result for each frequency, includes: For each frequency, feature recognition is performed on the measured ultrasonic image to determine all abnormal feature points. The abnormal feature points are feature points where the absolute value of the difference between the measured signal intensity and the mean signal intensity is greater than a preset intensity difference. The mean signal intensity is the mean of the measured signal intensity of all feature points in the measured ultrasonic image. Determine the fitted shape for all the aforementioned anomalous feature points; When the fitted shape is ring-shaped or ring-like, the measured diffraction ring detection result corresponding to the frequency is determined to indicate the presence of a diffraction ring. When the fitted shape is neither circular nor quasi-circular, the measured diffraction ring detection result corresponding to the frequency is determined to be that no diffraction ring exists.

3. The defect detection method according to claim 1, characterized in that, The step of performing diffraction ring detection on the measured ultrasonic image corresponding to each frequency, and determining the measured diffraction ring detection result for each frequency, includes: A bandpass filtering algorithm is used to filter the measured ultrasonic image corresponding to each frequency to determine the valid ultrasonic image corresponding to the frequency. Diffraction ring detection is performed on the effective ultrasonic image corresponding to the frequency to determine the measured diffraction ring detection result corresponding to the frequency.

4. The defect detection method according to claim 1, characterized in that, The determination of the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies includes: Based on the measured diffraction ring detection results corresponding to multiple frequencies, the defect location mapping table corresponding to the screen component is queried to determine the defect detection result of the screen component. The defect location mapping table is used to reflect whether diffraction rings exist at different defect locations at different frequencies.

5. The defect detection method according to claim 1, characterized in that, The determination of the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies includes: When the measured diffraction ring detection result corresponding to the frequency indicates the presence of a defective ring, the current defect feature corresponding to the frequency is determined. The current defect feature includes the diffraction ring radius and the diffraction ring signal difference. The diffraction ring signal difference is determined based on the measured signal intensity corresponding to all the feature points within the diffraction ring. Based on the current defect features corresponding to all the frequencies and the pre-trained defect localization model, the defect location corresponding to the screen component is determined.

6. The defect detection method according to claim 1, characterized in that, The defect detection method further includes: When the measured diffraction ring detection result corresponding to the frequency indicates the presence of a defective ring, the current defect feature corresponding to the frequency is determined. The current defect feature includes the diffraction ring radius and the diffraction ring signal difference. The diffraction ring signal difference is determined based on the measured signal intensity corresponding to all the feature points within the diffraction ring. Based on the current defect characteristics corresponding to each frequency, the defect level corresponding to the diffraction ring detected at each frequency is determined. The defect level of the screen component is determined based on the defect level corresponding to the diffraction rings detected at all the frequencies.

7. The defect detection method according to claim 5 or 6, characterized in that, The difference in the diffraction ring signal is the difference between the first characteristic mean and the second characteristic mean; The first feature mean is the mean of the M largest measured signal intensities among all the measured signal intensities corresponding to the feature points within the diffraction ring; The second characteristic mean is the mean of the M smaller measured signal intensities among all the characteristic points within the diffraction ring; M ≥ 1.

8. A defect detection device, characterized in that, include: The measured image acquisition module is used to control the ultrasonic module to emit ultrasonic signals corresponding to multiple frequencies to the screen component, and to determine the measured ultrasonic images corresponding to multiple frequencies based on the echo signals corresponding to multiple frequencies returned by the screen component. The diffraction ring detection module is used to perform diffraction ring detection on the measured ultrasonic image corresponding to each frequency, and to determine the measured diffraction ring detection result corresponding to each frequency. The defect detection result determination module is used to determine the defect detection result of the screen component based on the measured diffraction ring detection results corresponding to multiple frequencies.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the defect detection method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the defect detection method according to any one of claims 1 to 7.

Images