A defect detection method and a defect detection system

Abstract

This application provides a defect detection method and system, relating to the field of machine vision technology. The method includes: acquiring multiple target projection patterns, wherein each target projection pattern includes at least a portion of preset projection patterns; controlling an illumination source to project the multiple target projection patterns onto a sample to be tested; controlling an image acquisition device to acquire projection images corresponding to the multiple target projection patterns; processing the projection images corresponding to the multiple target projection patterns using an image processing algorithm configured for the multiple target projection patterns to obtain result images corresponding to the multiple target projection patterns; and performing defect detection based on the multiple result images to obtain the target defect detection result for the sample to be tested. Applying this application can improve the efficiency and accuracy of defect detection.

Description

Technical Field

[0001] This application relates to the field of machine vision technology, and in particular to a defect detection method and a defect detection system. Background Technology

[0002] In industries such as automobile manufacturing, electronics manufacturing, and medical devices, product appearance defects directly affect product quality. Therefore, defect detection of product surfaces is necessary during the production process. In industrial production, machine vision inspection technology is used to automate defect detection. A machine vision inspection system typically consists of an illumination source, an industrial camera, an image processing unit, and a control unit. When using machine vision technology to detect surface defects, a patterned illumination source is projected onto the object being inspected. An industrial camera captures images of the object's surface, and image processing algorithms analyze and process the images to identify and locate defects. Currently, the main machine vision inspection methods include photometric stereo methods, phase deflection methods, phase profilometry, and structured light technology.

[0003] In related technologies, the patterns projected by the illumination source differ depending on the detection method used, and different detection methods are suitable for objects with different surface properties. However, in practical applications, the surface properties of objects are diverse, not limited to smooth surfaces or only diffuse reflective surfaces. For objects with complex surface properties, it is impossible to detect all surface defects using a single method. A corresponding defect detection scheme needs to be determined based on the surface properties of the object, which not only increases the cost of defect detection but also leads to lower efficiency. Furthermore, if the user misjudges the surface properties of the object, the defect detection results will be inaccurate. Summary of the Invention

[0004] The purpose of this application is to provide a defect detection method and a defect detection system to improve the efficiency and accuracy of defect detection. The specific technical solution is as follows:

[0005] A first aspect of this application provides a defect detection method, the method comprising:

[0006] Acquire multiple target projection patterns, wherein the target projection patterns include at least a portion of the preset projection patterns;

[0007] Control the illumination source to project the multiple target projection patterns onto the sample to be tested;

[0008] The image acquisition device is controlled to acquire projected images corresponding to the multiple target projection patterns;

[0009] Using the image processing algorithm configured for the multiple target projection patterns, the projected images corresponding to the multiple target projection patterns are processed to obtain the result images corresponding to the multiple target projection patterns;

[0010] Defect detection is performed based on multiple result images to obtain the target defect detection results of the sample to be tested.

[0011] In one possible implementation, the preset projection pattern includes at least one of the following: a projection pattern for photometric stereoscopic methods, a projection pattern for phase deflection methods, a projection pattern for phase profilometry, and a projection pattern for structured light projection technology.

[0012] In one possible implementation, the method further includes:

[0013] Display the result images corresponding to the multiple target projection patterns;

[0014] The defect detection based on multiple result images includes:

[0015] Defect detection is performed based on the result image corresponding to the target projection pattern indicated by the image selection command.

[0016] In one possible implementation, acquiring multiple target projection patterns includes:

[0017] Demonstrates at least one preset image processing algorithm and a preset projected image configured for each of the preset image processing algorithms;

[0018] Obtain the preset projection pattern indicated by the pattern selection command as the target projection pattern.

[0019] In one possible implementation, at least one of the preset projection patterns is imported before receiving the pattern selection instruction and after receiving the pattern import instruction, and the method further includes:

[0020] An image processing algorithm is assigned to the projection pattern indicated by the pattern import instruction.

[0021] In one possible implementation, controlling the illumination source to project the plurality of target projection patterns includes:

[0022] Each target projection pattern is divided into multiple pattern groups according to the type of image processing algorithm configured for each target projection pattern, and the image processing algorithms configured for each target projection pattern in the same pattern group are the same;

[0023] The illumination source is controlled to sequentially project each of the target projection patterns in each of the pattern groups.

[0024] In one possible implementation, the control image acquisition device acquires projected images corresponding to the plurality of target projection patterns, including:

[0025] The image acquisition device is controlled to acquire a preset number of image frames to obtain the original image. The target projection pattern of the sample to be tested is different in each image frame.

[0026] The original image is split according to the image processing algorithm configured for each of the target projection patterns to obtain the projection images corresponding to the multiple target projection patterns.

[0027] In one possible implementation, acquiring multiple target projection patterns includes:

[0028] A target projection pattern combination is determined from a preset projection pattern combination, wherein the preset projection pattern combination is either pre-set or user-defined.

[0029] The preset projection pattern indicated by the pattern selection instruction is obtained from the target projection pattern combination and used as the target projection pattern.

[0030] In one possible implementation, each of the preset projection pattern combinations is pre-configured with a corresponding detection scene, and the method further includes:

[0031] Determine the current detection scenario;

[0032] Determining the target projection pattern combination from the preset projection pattern combinations includes:

[0033] The projection pattern combination that matches the current detection scene in the preset projection pattern combination is identified as the target projection pattern combination.

[0034] A second aspect of this application provides a defect detection system, the defect detection system comprising: an illumination source, an image acquisition device, and a controller;

[0035] The illumination source is used to project a target projection pattern, the target projection pattern including at least a portion of a preset projection pattern;

[0036] The image acquisition device is used to acquire the projected image corresponding to the target projection pattern;

[0037] The controller is configured to acquire multiple target projection patterns; control an illumination source to project the multiple target projection patterns onto the sample to be tested; control an image acquisition device to acquire projection images corresponding to the multiple target projection patterns; process the projection images corresponding to the multiple target projection patterns using an image processing algorithm configured for the multiple target projection patterns to obtain result images corresponding to the multiple target projection patterns; and perform defect detection based on the multiple result images to obtain the target defect detection results of the sample to be tested.

[0038] In one possible implementation, the light emitted by the illumination source is projected onto the sample to be tested, and after being reflected by the sample to be tested, it enters the image acquisition device. There is a first preset angle between the optical axis of the image acquisition device and the normal of the sample to be tested, and there is a second preset angle between the optical axis of the illumination source and the normal of the sample to be tested.

[0039] In one possible implementation, the defect detection system further includes a beam splitter; light emitted from the illumination source is transmitted through the beam splitter and projected onto the sample to be tested, reflected by the sample to be tested and then irradiated by the beam splitter, and reflected by the beam splitter into the image acquisition device. The optical axis of the image acquisition device is perpendicular to the optical axis of the illumination source and perpendicular to the normal of the sample to be tested, and the optical axis of the illumination source is parallel to the normal of the sample to be tested.

[0040] In one possible implementation, the defect detection system further includes a beam splitter; the light emitted by the illumination source is reflected by the beam splitter and projected onto the sample to be tested, and after being reflected by the sample to be tested, it is transmitted through the beam splitter to the image acquisition device. The optical axis of the image acquisition device is perpendicular to the optical axis of the illumination source and parallel to the normal of the sample to be tested, and the optical axis of the illumination source is perpendicular to the normal of the sample to be tested.

[0041] In one possible implementation, the light emitted by the illumination source is projected onto the sample to be tested, and transmitted through the sample to the image acquisition device. The optical axis of the image acquisition device and the optical axis of the illumination source are parallel to the normal of the sample to be tested, respectively.

[0042] Beneficial effects of the embodiments in this application:

[0043] This application provides a defect detection method and system. Users can select multiple target projection patterns from a preset projection pattern library. After controlling a lighting source to project these patterns onto the sample under test and acquiring the corresponding projection images, an image processing algorithm is pre-configured for each pattern. This allows for defect detection on the projection images corresponding to each target pattern, resulting in a result image. Finally, defect detection is performed based on this result image to obtain the final defect detection result. This approach can detect more types of defects on the sample surface. Furthermore, even if the user misjudges the surface properties of the sample, the defect detection results obtained from each result image can reduce the problem of low accuracy in one image processing algorithm due to misjudgment of surface properties, thus improving the accuracy of defect detection. Simultaneously, it eliminates the need to determine the defect detection scheme based on the surface properties of the sample, improving the efficiency of defect detection.

[0044] Of course, implementing any product or method of this application does not necessarily require achieving all of the advantages described above at the same time. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0046] Figure 1 A schematic diagram of a defect detection method provided in an embodiment of this application;

[0047] Figure 2 Another schematic diagram of the defect detection method provided in the embodiments of this application;

[0048] Figure 3 Another schematic diagram of the defect detection method provided in the embodiments of this application;

[0049] Figure 4 This is an example diagram of an interface used to demonstrate preset image processing algorithms and preset projection patterns;

[0050] Figure 5 Another schematic diagram of the defect detection method provided in the embodiments of this application;

[0051] Figure 6 This is another example diagram of an interface used to demonstrate preset image processing algorithms and preset projection patterns;

[0052] Figure 7Another schematic diagram of the defect detection method provided in the embodiments of this application;

[0053] Figure 8a An example diagram of a projection pattern provided in an embodiment of this application;

[0054] Figure 8b Another example diagram of the projection pattern provided in the embodiments of this application;

[0055] Figure 9a Another example diagram of the projection pattern provided in the embodiments of this application;

[0056] Figure 9b Another example diagram of the projection pattern provided in the embodiments of this application;

[0057] Figure 10 Another example diagram of the projection pattern provided in the embodiments of this application;

[0058] Figure 11 Another schematic diagram of the defect detection method provided in the embodiments of this application;

[0059] Figure 12 Example diagram of the original image provided for embodiments of this application;

[0060] Figure 13 To Figure 12 An example image of the projected image obtained by splitting;

[0061] Figure 14 To Figure 12 Another example of the projected image obtained by splitting;

[0062] Figure 15 A flowchart of the defect detection method provided in the embodiments of this application;

[0063] Figure 16 This is a schematic diagram of the structure of the defect detection system provided in the embodiments of this application;

[0064] Figure 17 This is a schematic diagram of the defect detection system provided in the embodiments of this application when it is tilted and erected.

[0065] Figure 18 A schematic diagram of a coaxially mounted defect detection system provided in an embodiment of this application;

[0066] Figure 19 Another structural diagram of the defect detection system provided in this application embodiment when coaxially mounted;

[0067] Figure 20 This is a schematic diagram of the structure of the defect detection system provided in the embodiments of this application when it is mounted via transmission.

[0068] Figure 21 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0069] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art based on this application are within the scope of protection of this application.

[0070] First, the technical terms used in the embodiments of this application will be explained:

[0071] Phase polarization method: a method for measuring defects based on the reflection of sine / cosine stripe light spots on the surface of an object. The principle is that if there are defects on the surface of a mirror object, the reflected stripes will have different phase gradients. Based on the phase gradient collected by the camera, defects that are difficult to distinguish by the human eye can be clearly presented. At the same time, since it is only related to the reflection characteristics of the object surface, it is less affected by the color of the object surface and has strong anti-interference characteristics.

[0072] Photometric stereo method: a method for estimating surface defects of an object by obtaining images of the object's surface from different directions of illumination.

[0073] Phase profilometry is a high-precision, non-contact 3D topography measurement technique based on fringe projection and phase measurement. It reconstructs the 3D contour by analyzing the phase changes of the fringes on an object's surface. The basic principle is that phase profilometry projects a sinusoidal fringe pattern onto the object being measured. The height changes on the object's surface cause the fringes to distort, resulting in deformed fringes in the image captured by the camera. These fringe deformations contain the object's 3D information. By extracting the phase distribution of the fringes using phase-shifting techniques or phase demodulation algorithms, and combining this with geometric constraints, the height information of the object's surface can be calculated, thus achieving 3D contour reconstruction.

[0074] Structured light inspection technology: a technique that uses specific lighting patterns to acquire three-dimensional information about objects. Its basic principle is to project a pre-designed structured light pattern (such as stripes or light spots) onto the surface of an object using a projector, and then analyze the image captured by a camera to obtain the object's three-dimensional shape and surface features.

[0075] A first aspect of this application provides a defect detection method, see [link to previous section]. Figure 1The method includes the following steps: Step S10, acquiring multiple target projection patterns, wherein the target projection patterns include at least a portion of the preset projection patterns; Step S20, controlling an illumination source to project multiple target projection patterns onto the sample to be tested; Step S30, controlling an image acquisition device to acquire projection images corresponding to the multiple target projection patterns; Step S40, using an image processing algorithm configured for the multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns to obtain result images corresponding to the multiple target projection patterns; Step S50, performing defect detection based on the multiple result images to obtain the target defect detection results of the sample to be tested. In this embodiment, the user can select multiple target projection patterns from a preset projection pattern library. After controlling the illumination source to project multiple target projection patterns onto the sample under test and acquiring the corresponding projection images, since each target projection pattern is pre-configured with an image processing algorithm, defect detection can be performed on the corresponding projection image based on the configured image processing algorithm, resulting in a result image for each target projection pattern. Finally, defect detection is performed based on the result image to obtain the final defect detection result. This method can detect more types of defects on the surface of the sample under test. Furthermore, even if the user misjudges the surface properties of the sample under test, the defect detection results obtained from each result image can reduce the problem of low accuracy of one image processing algorithm due to misjudgment of surface properties, thus improving the accuracy of defect detection. At the same time, it eliminates the need to determine the defect detection scheme based on the surface properties of the sample under test, improving the efficiency of defect detection.

[0076] The following is a detailed explanation of steps S10-S50:

[0077] In step S10 of this embodiment, the multiple target projection patterns are selected by the user from the preset projection patterns. The preset projection patterns can be pre-set or imported by the user. Each preset projection pattern is configured with a corresponding image processing algorithm. The user can select the target projection pattern according to the type of image processing algorithm.

[0078] Each projection pattern is used to detect a specific type of defect. Different image processing algorithms are used to process the projected images for different types of defects. Users can select the target projection pattern according to their actual needs or import projection patterns and then select the target projection pattern. A set of projection patterns can be configured for each image processing algorithm, and each set of projection patterns includes at least one projection pattern.

[0079] For example, suppose the preset projection pattern includes projection pattern set 1 of image processing algorithm 1, projection pattern set 2 of image processing algorithm 2, and projection pattern set 3 of image processing algorithm 3. Projection pattern set 1 includes preset projection pattern a-preset projection pattern c, projection pattern set 2 includes preset projection pattern d-preset projection pattern f, and projection pattern set 3 includes preset projection pattern g-preset projection pattern i. The user can select preset projection pattern a, preset projection pattern c, and preset projection pattern i as target projection pattern 1-target projection pattern 3 respectively; or import projection pattern j and projection pattern k as target projection pattern 4 and target projection pattern 5 respectively; or simultaneously select preset projection pattern a, preset projection pattern c, and preset projection pattern i as target projection pattern 1-target projection pattern 3, and projection pattern j and projection pattern k as target projection pattern 4 and target projection pattern 5 respectively.

[0080] In this embodiment, the acquired multiple target projection patterns may include at least one projection pattern configured by each image processing algorithm, or may only include at least one projection pattern configured by one or more image processing algorithms; this application does not limit this. It is understood that if the acquired multiple target projection patterns include multiple projection patterns configured by the same image processing algorithm, then the multiple projection patterns are different. For example, assuming the multiple projection patterns are projection patterns used for phase deflection, the stripes in the multiple projection patterns have different phases and / or different thicknesses. Furthermore, the same target projection pattern can be regular or varied in the spatial domain. For example, for a projection pattern used for phase deflection, it can be formed by stripes with the same phase and thickness, or it can be formed by stripes with different phases and / or different thicknesses.

[0081] In step S20 of this embodiment, the multiple target projection images projected onto the sample to be tested can be all the acquired target projection patterns, or a portion of the acquired target projection patterns. When projecting the target patterns, each target projection pattern can be projected onto the surface of the sample to be tested using a programmable light source. The projection order can be random. For example, assuming the acquired multiple target projection patterns are target projection pattern 1 to target projection pattern 7, target projection patterns 1 to target projection pattern 7 can be projected sequentially in a random order; alternatively, they can be pre-set according to a preset sorting method, such as projecting multiple target projection patterns sequentially according to the type of image processing algorithm configured for each target projection pattern. The specific process of projecting multiple target projection patterns sequentially according to the type of image processing algorithm configured for each target projection pattern is detailed below and will not be repeated here. It is understood that the light intensity and projection angle are the same for each target projection pattern when it is projected.

[0082] In step S30 of this embodiment, controlling the image acquisition device to acquire projected images corresponding to multiple target projection patterns means that during the projection of multiple target projection patterns, the image acquisition device is controlled to simultaneously capture images of the sample to be tested, thereby obtaining projected images corresponding to multiple target projection patterns. The target projection patterns and projected images can be in a one-to-one correspondence, or some target projection patterns can correspond to one projected image. For example, when projecting target projection patterns 1 to 7, the acquired projected images can be projected images 1 to 7, each corresponding to one projection pattern; or they can be projected images 1 to 3, where projected image 1 is the projected image corresponding to projected patterns 1 to 3, projected image 2 is the projected image corresponding to projected patterns 4 to 6, and projected image 3 is the projected image corresponding to projected pattern 7. It is understood that the parameters of the image acquisition device are consistent each time a projected image is acquired.

[0083] In this embodiment, the multiple target projection patterns projected onto the sample to be tested can be projection patterns configured by the same image processing algorithm, or projection patterns configured by multiple image processing algorithms.

[0084] In step S40 of this embodiment, the image processing algorithm is used to process the projected pattern to obtain a result image of the features in the projected pattern. The result images corresponding to multiple target projected patterns can be one result image for each target projected pattern, or a result image for some target projected patterns. When processing the projected image corresponding to the target projected pattern using the image processing algorithm configured for each target projected pattern, the image can be processed directly using each image processing algorithm, or the image processing algorithms can be combined with traditional machine learning to process each projected image, or the image processing algorithms can be combined with deep learning to process each projected image. This embodiment of the application does not limit this.

[0085] In step S50 of this embodiment, the multiple result images can be all the result images obtained in step S40, or only a portion of them; this embodiment does not limit this. Defect detection based on the result images refers to performing defect detection on each result image separately to obtain defect detection results. The defect detection results may include one or more of the following: defect location, defect type, defect size, and confidence level. They may also include characteristic data of the sample to be tested, such as the size of the sample.

[0086] In one possible implementation, see Figure 2 The defect detection method provided in this application includes the following steps:

[0087] Step S10: Acquire multiple target projection patterns, wherein the target projection patterns include at least a portion of the preset projection patterns; Step S20: Control the illumination source to project the multiple target projection patterns onto the sample to be tested; Step S30: Control the image acquisition device to acquire the projection images corresponding to the multiple target projection patterns; Step S40: Use the image processing algorithm configured for the multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns to obtain the result images corresponding to the multiple target projection patterns; Step S21: Display the result images corresponding to the multiple target projection patterns; Step S501: Perform defect detection based on the result images corresponding to the target projection patterns indicated by the image selection command to obtain the target defect detection results of the sample to be tested. By displaying the result images corresponding to multiple target projection patterns to the user, the user can select the result image to be detected, improving the flexibility and adaptability of defect detection.

[0088] In this embodiment, step S501 is a detailed refinement of step S50 above. Steps S10-S40 are described above and will not be repeated here.

[0089] In step S21 of this embodiment, displaying the result images corresponding to multiple target projection patterns means showing the result images corresponding to the target projection patterns to the user on the user interface. This can be displayed in a list format, a thumbnail format, or in different display windows. The user can select a result image from the displayed result images, i.e., the result image indicated by the image selection command.

[0090] In one possible implementation, see Figure 3 The defect detection method provided in this application includes the following steps:

[0091] Step S101: Display at least one preset image processing algorithm and a preset projection image configured for each preset image processing algorithm; Step S102: Obtain the preset projection pattern indicated by the pattern selection instruction as the target projection image; Step S20: Control the illumination source to project multiple target projection patterns onto the sample to be tested; Step S30: Control the image acquisition device to acquire the projection images corresponding to the multiple target projection patterns; Step S40: Use the image processing algorithm configured for the multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns to obtain the result images corresponding to the multiple target projection patterns; Step S50: Perform defect detection based on the multiple result images to obtain the target defect detection result of the sample to be tested. By displaying preset image processing algorithms and preset projection images configured for each preset image processing algorithm to the user, the user can choose the image processing algorithm and projection pattern required for defect detection, improving the flexibility and adaptability of defect detection.

[0092] In this embodiment, steps S101 and S102 are detailed steps of step S10 above. Steps S20-S50 are described above and will not be repeated here.

[0093] In step S101 of this embodiment, displaying at least one preset image processing algorithm and a preset projection image configured for each preset image processing algorithm means displaying the preset image processing algorithm and the preset projection image to the user on the user interface. This can be done by displaying the preset image processing algorithm and the preset projection image in the form of a list, a thumbnail, or different display windows.

[0094] For example, such as Figure 4As shown, after a user clicks on a preset image processing algorithm control, a window displays preset projection patterns applicable to that algorithm. The user can select a preset projection pattern as the target projection pattern in this window. If none of the displayed preset projection images meet the user's needs, the user can also import a custom image file as the target projection pattern by clicking the "Import" control. It is understandable that when a user imports a target projection pattern, the imported pattern can also be added to the preset projection patterns simultaneously, providing more options for the user when selecting a target projection pattern next time.

[0095] In step S102 of this embodiment, the user clicks or checks one or more preset projection patterns among the displayed preset projection patterns. The preset projection pattern indicated by the pattern selection command is the preset projection pattern clicked or checked by the user. If the user clicks the "Import" control or enters an import command and imports a custom image file, the imported custom image file is the projection pattern indicated by the import command.

[0096] In this embodiment, the import command can be entered after selecting a preset image processing algorithm or before selecting a preset image processing algorithm. If the import command is entered before selecting a preset image processing algorithm, the user needs to configure an image processing algorithm for the imported projection pattern after importing it. For example, the user can specify a preset image processing algorithm as the image processing algorithm configured for the imported projection pattern. Alternatively, the user can import a new image processing algorithm as a new preset image processing algorithm and use the imported new preset image processing algorithm as the image processing algorithm configured for the imported projection pattern.

[0097] In another possible implementation, if the import command is entered after selecting a preset image processing algorithm, see [link to relevant documentation]. Figure 5 The image processing algorithm provided in this application includes the following steps:

[0098] Step S101: Display at least one preset image processing algorithm and a preset projection image configured for each of the preset image processing algorithms; Step S102: Obtain the preset projection pattern indicated by the pattern selection instruction as the target projection image; Step S51: Assign an image processing algorithm to the projection pattern indicated by the pattern import instruction; Step S20: Control the illumination source to project multiple target projection patterns onto the sample to be tested; Step S30: Control the image acquisition device to acquire the projection images corresponding to the multiple target projection patterns; Step S40: Process the projection images corresponding to the multiple target projection patterns using the image processing algorithms configured for the multiple target projection patterns to obtain the result images corresponding to the multiple target projection patterns; Step S50: Perform defect detection based on the multiple result images to obtain the target defect detection result of the sample to be tested. By using the embodiment of this application, the image processing algorithm selected when receiving the pattern import instruction is used as the image processing algorithm configured for the imported projection pattern, reducing the risk of defect detection failure due to human error in configuring the image processing algorithm and improving the reliability of defect detection.

[0099] In step S51 of this embodiment, the image processing algorithm assigned to the projection pattern indicated by the pattern import command can be the image processing algorithm selected when the pattern import command is received, or it can be the image processing algorithm selected by the user from the displayed preset image processing algorithms when the projection pattern indicated by the import command is received. It can also be other methods of assigning an image processing algorithm to the imported projection pattern; this embodiment does not limit this. For example, as shown... Figure 6 As shown, after the user selects the preset image processing algorithm 1, the window not only displays the preset projection pattern applicable to the image processing algorithm 1, but also displays the import pattern control. If the user imports the projection pattern A by entering the import command through the import pattern control, the image processing algorithm configured for the projection pattern is the preset image processing algorithm 1.

[0100] In another possible implementation, after the user imports the projection pattern, a prompt window can be displayed to the user. This prompt window includes the imported projection pattern and various preset image processing algorithms. The user selects the image processing algorithm configured for the imported projection pattern through this window. In this interface, after selecting the target projection pattern, the user clicks "Start Projection" to control the lighting source to project the target projection pattern.

[0101] The following describes the specific process of projecting multiple target patterns sequentially according to the type of image processing algorithm configured for each target pattern.

[0102] In one possible implementation, see Figure 7The defect detection method provided in this application includes the following steps: Step S10, acquiring multiple target projection patterns, wherein the target projection patterns include at least a portion of the preset projection patterns; Step S201, dividing each target projection pattern into multiple pattern groups according to the type of image processing algorithm configured for each target projection pattern, wherein the image processing algorithms configured for each target projection pattern in the same pattern group are the same; Step S202, controlling the illumination source to sequentially project each target projection pattern in each pattern group; Step S30, controlling the image acquisition device to acquire projection images corresponding to the multiple target projection patterns; Step S40, using the image processing algorithm configured for the multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns to obtain result images corresponding to the multiple target projection patterns; Step S50, performing defect detection based on the multiple result images to obtain the target defect detection result of the sample to be tested. In this embodiment, each target projection pattern is divided into multiple pattern groups by configuring the type of image processing algorithm for each target projection pattern. In this way, each target projection pattern in the same pattern group is configured with the same image processing algorithm. After obtaining the projection image, it is convenient to use the same image processing algorithm to perform defect detection on the projection image corresponding to the target projection pattern in the same pattern group, thereby improving the defect detection efficiency.

[0103] In this embodiment, steps S201 and S202 are detailed steps of step S20 above. Steps S10 and S30-S50 are described above and will not be repeated here.

[0104] For example, assuming the acquired multiple target projection patterns are target projection pattern 1 to target projection pattern 7, the image processing algorithm configured for target projection patterns 1, 5, 6 and 7 is image processing algorithm 1, and the image processing algorithm configured for target projection patterns 2 to target projection pattern 4 is image processing algorithm 2, then target projection patterns 1, 5, 6 and 7 can be divided into pattern group 1, and target projection patterns 2 to target projection pattern 4 can be divided into pattern group 2. When controlling the projection of the lighting source, all target projection patterns in pattern group 1 (or pattern group 2) are projected in sequence first, and then all target projection patterns in pattern group 2 (or pattern group 1) are projected in sequence.

[0105] In one possible implementation, the preset projection pattern in the above embodiments includes at least one of the following: a projection pattern for photometric stereo method, a projection pattern for phase deflection method, a projection pattern for phase profilometry, a projection pattern for structured light detection technology, and projection patterns required for other optical imaging technologies. By selecting this embodiment, users are provided with a variety of image processing algorithms to choose from, and these algorithms cover different types of defect detection. Through the complementarity and synergy of multiple image processing algorithms, the accuracy of defect detection is further improved.

[0106] In this application, the preset projection patterns differ under different preset image processing algorithms, and the same image processing algorithm may also include multiple projection pattern groups. For example, a projection pattern suitable for photometric stereochemistry could be... Figure 8a One or more of the four projection patterns shown can also be Figure 8b One or more of the five projection patterns shown may also include simultaneously. Figure 8a and Figure 8b The projection pattern in the image can be multiple; similarly, the projection pattern suitable for phase profilometry can be... Figure 9a One or more of the six projected patterns shown can also be Figure 9b One or more of the eight projection patterns shown may also include simultaneously. Figure 9a and Figure 9b Multiple projection patterns are included; projection patterns suitable for structured light detection technology can be... Figure 10 Multiple of the eight patterns shown. It is understandable that... Figure 8a , Figure 8b , Figure 9a , Figure 9b as well as Figure 10 This is merely an example of a projected pattern and does not impose any limitations on projected patterns.

[0107] In one possible implementation, see Figure 11 The defect detection method provided in this application includes the following steps: Step S10, acquiring multiple target projection patterns, wherein the target projection patterns include at least a portion of the preset projection patterns; Step S20, controlling an illumination source to project multiple target projection patterns onto the sample to be tested; S301, controlling an image acquisition device to acquire a preset number of image frames to obtain an original image, wherein the target projection patterns projected onto the sample to be tested are different in each image frame; Step S302, splitting the original image according to the image processing algorithm configured for each target projection pattern to obtain a projection image corresponding to each target projection pattern; Step S40, using the image processing algorithm configured for multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns to obtain a result image corresponding to the multiple target projection patterns; Step S50, performing defect detection based on the multiple result images to obtain the target defect detection result of the sample to be tested. By using this embodiment, parallel processing is achieved by splitting the original image according to the image processing algorithm, shortening the overall processing time, thereby shortening the defect detection time and improving the efficiency of defect detection.

[0108] In this embodiment, steps S301 and S302 are detailed steps of step S30 above. Steps S10-S20 and S40-S50 are described above and will not be repeated here.

[0109] In this embodiment, the target projection patterns can be projected sequentially in a random order, or they can be divided into multiple pattern groups and then projected according to different pattern groups. This embodiment does not limit this. In step S301 of this embodiment, the preset number is equal to the number of target projection patterns projected in step S20. During the image acquisition process, the image acquisition device can generate and output a projection image each time it acquires the sample to be tested. To enrich the image details of the projection image acquired by the camera, the image acquisition device can also expose to obtain an image frame each time a target projection pattern is projected. After all target projection patterns are projected, the image acquisition device generates an original image including a preset number of image frames.

[0110] For example, suppose the original image is as follows: Figure 12 As shown, there are 6 image frames, which are obtained by controlling the image acquisition device when projecting target patterns 1 to 6. If the image processing algorithm configured for target patterns 1, 5, and 6 is image processing algorithm 1, and the image processing algorithm configured for target patterns 2 to 4 is image processing algorithm 2, then the original images can be processed according to... Figure 13 The image is divided into 6 projected images, each containing one image frame; that is, image frames 1 through 6 represent the 6 projected images respectively. Alternatively, the original image can be divided into... Figure 14 The image is split into two projected images: image frames 1, 5, and 6 are split into one projected image, and image frames 2 to 4 are split into the other projected image.

[0111] In one possible implementation, the defect detection method provided in this application includes the following steps: Step S11, determining a target projection pattern combination from a preset projection pattern combination, wherein the preset projection pattern combination is pre-set or user-defined; Step S12, obtaining the preset projection pattern indicated by the pattern selection instruction from the target projection pattern combination as the target projection pattern, wherein the target projection pattern includes at least a portion of the projection patterns in the preset projection pattern; Step S20, controlling an illumination source to project multiple target projection patterns onto the sample to be tested; Step S30, controlling an image acquisition device to acquire projection images corresponding to the multiple target projection patterns; Step S40, using an image processing algorithm configured for the multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns to obtain result images corresponding to the multiple target projection patterns; Step S50, performing defect detection based on the multiple result images to obtain the target defect detection result of the sample to be tested. By using this embodiment, the target projection pattern can be quickly determined through the preset projection pattern combination, accelerating the defect detection process, shortening the defect detection time, and thus improving the efficiency of defect detection.

[0112] In this embodiment, steps S11 and S12 are another detailed refinement of step S10 described above. Steps S20-S50 are described above and will not be repeated here.

[0113] In step S11 of this embodiment, the preset projection pattern combination includes at least one projection pattern combination. The projection pattern combination in the preset projection pattern combination can be preset by the system or created by the user. This embodiment does not limit this.

[0114] In step S12 of this embodiment, the preset projection pattern indicated by the pattern selection command can be all preset projection patterns in the target projection pattern combination, or it can be a portion of the projection patterns in the target projection pattern combination. Specifically, the user can select all preset projection patterns in the target projection pattern combination as the target projection pattern using the select all control, or they can select a portion of the preset projection patterns as the target projection pattern.

[0115] In one possible implementation, one or more detection scenarios can be pre-configured for each preset projection pattern combination. The defect detection method provided in this application includes the following steps: Step S100, determining the current detection scenario; Step S111, determining the projection pattern combination in the preset projection pattern combination whose detection scenario matches the current detection scenario, as the target projection pattern combination, wherein the preset projection pattern combination is preset or user-defined; Step S12, obtaining the preset projection pattern indicated by the pattern selection instruction from the target projection pattern combination, as the target projection pattern, wherein the target projection pattern includes at least a portion of the projection patterns in the preset projection pattern; Step S20, controlling the illumination source to project multiple target projection patterns onto the sample to be tested; S30, controlling the image acquisition device to acquire the projection images corresponding to the multiple target projection patterns; Step S40, using the image processing algorithm configured for the multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns, obtaining the result images corresponding to the multiple target projection patterns; Step S50, performing defect detection based on the multiple result images, obtaining the target defect detection result of the sample to be tested. By using this embodiment, by configuring a corresponding detection scene for each preset projection pattern combination, the target projection pattern combination can be quickly matched according to the current detection scene, which speeds up the process of entering defect detection, shortens the defect detection time, and thus improves the efficiency of defect detection.

[0116] In this embodiment, the detailed steps of step S111 above, and steps S20-S50 are described above and will not be repeated here.

[0117] In step S100 of this embodiment, the current detection scene can be obtained by the defect detection system through image acquisition device recognition, or it can be input by the user; both are acceptable. For example, the detection scene can be: a glass scene, a glass + smooth metal scene, a glass + metal + wood material scene, a smooth surface scene, a rough surface scene, or a smooth surface + rough surface scene. It is understood that the above scenes are merely illustrative examples of detection scenes and do not limit the scope of the detection scene.

[0118] In step S111 of this embodiment, the projection pattern combination that matches the current detection scene can refer to a projection pattern combination where the detection scene is the same as the current detection scene, or it can refer to a projection pattern combination where the detection scene is similar to the current detection scene. This embodiment does not limit this. If it refers to a projection pattern combination where the detection scene is similar to the current detection scene, and there are multiple determined projection pattern combinations similar to the detection scene, then any projection pattern combination can be randomly selected from the determined projection pattern combinations as the target projection pattern combination, or the projection pattern combination with the highest similarity can be selected as the target projection pattern combination based on the similarity of the detection scenes.

[0119] In one possible implementation, after performing defect detection on each result image, the defect detection results can be fused to obtain the target defect detection result. In this case, the flow of the defect detection method provided in this application embodiment is as follows: Figure 15 As shown, it includes:

[0120] Step S1: The user selects a preset pattern through the client or imports a pattern to create a combination.

[0121] Step S2: The client sends a command to the controller according to the selected pattern;

[0122] Step S3: The controller controls the light source to project composite stripes;

[0123] Step S4: Control the camera to synchronously acquire images;

[0124] Step S5: The acquired images are segmented, and different categories of images are processed using their respective algorithms to obtain direct detection results;

[0125] Step S6: The direct detection results are fused to obtain the remaining defect detection results.

[0126] Wherein, steps S1 and S2 correspond to the aforementioned steps S101 and S102, step S3 corresponds to the aforementioned step S20, step S4 corresponds to the aforementioned step S30, step S5 corresponds to the aforementioned steps S302 and S40, and step S6 corresponds to the aforementioned step S50.

[0127] Corresponding to the first aspect mentioned above, a second aspect of the embodiments of this application provides a defect detection system, such as... Figure 16 As shown, the defect detection system includes: an illumination source 1601, an image acquisition device 1602, and a controller 1603;

[0128] The illumination source 1601 is used to project a target projection pattern, the target projection pattern including at least a portion of a preset projection pattern;

[0129] Image acquisition device 1602 is used to acquire the projected image corresponding to the target projection pattern;

[0130] The controller 1603 is used to acquire multiple target projection patterns; control an illumination source to project the multiple target projection patterns onto the sample to be tested; control an image acquisition device to acquire projection images corresponding to the multiple target projection patterns; use an image processing algorithm configured for the multiple target projection patterns to process the projection images corresponding to the multiple target projection patterns to obtain result images corresponding to the multiple target projection patterns; and perform defect detection based on the multiple result images to obtain the target defect detection results of the sample to be tested.

[0131] In this embodiment, the user can select multiple target projection patterns from a preset projection pattern library. After controlling the illumination source to project multiple target projection patterns onto the sample under test and acquiring the corresponding projection images, since each target projection pattern is pre-configured with an image processing algorithm, defect detection can be performed on the corresponding projection image based on the configured image processing algorithm, resulting in a result image for each target projection pattern. Finally, defect detection is performed based on the result image to obtain the final defect detection result. This method can detect more types of defects on the surface of the sample under test. Furthermore, even if the user misjudges the surface properties of the sample under test, the defect detection results obtained from each result image can reduce the problem of low accuracy of one image processing algorithm due to misjudgment of surface properties, thus improving the accuracy of defect detection. At the same time, it eliminates the need to determine the defect detection scheme based on the surface properties of the sample under test, improving the efficiency of defect detection.

[0132] The defect detection method provided in this application embodiment is applied to the controller of the defect detection system. In this embodiment, the controller 1603 can be a single controller or two independent sub-controllers. This application embodiment does not limit this.

[0133] The defect detection system provided in this embodiment also includes a client, which displays preset image processing algorithms and preset projection patterns, making it convenient for users to select target projection patterns and import target projection patterns.

[0134] In this application, the lighting source 1601 and the image acquisition device 1602 in the defect detection system can be mounted at an angle, coaxially, or through transmission. This application specifies the mounting method accordingly. Each mounting method is described below.

[0135] In one possible implementation, the structure is tilted as follows: Figure 17 As shown, the light emitted by the illumination source 1601 is projected onto the sample 1701 under test, and after reflection from the sample, enters the image acquisition device 1602. A first preset angle exists between the optical axis of the image acquisition device 1602 and the normal of the sample 1701, and a second preset angle exists between the optical axis of the illumination source 1601 and the normal of the sample 1701 under test. Both the first and second preset angles are less than 90°. Using this embodiment, when the surface of the sample under test is mostly a diffuse reflective rough surface, a tilted mounting is preferred, which can reduce costs and improve defect detection results.

[0136] In one possible implementation, the coaxial mounting is as follows: Figure 18 As shown, the defect detection system also includes a beam splitter 1604; light emitted from the illumination source 1601 is transmitted through the beam splitter 1604 and projected onto the sample 1701 under test. After being reflected by the sample 1701, the light illuminates the beam splitter 1604 and then enters the image acquisition device 1602 after being reflected by the beam splitter 1604. The optical axis of the image acquisition device 1602 is perpendicular to the optical axis of the illumination source 1601 and perpendicular to the normal of the sample 1701 under test. The optical axis of the illumination source 1601 is parallel to the normal of the sample 1701 under test. In one possible embodiment, coaxial mounting can also be as follows: Figure 19 As shown, the defect detection system also includes a beam splitter 1604. Light emitted from the illumination source 1601 is reflected by the beam splitter 1604 and projected onto the sample 1701 under test. After being reflected by the sample 1701, the light is transmitted through the beam splitter 1604 to the image acquisition device 1602. The optical axis of the image acquisition device 1602 is perpendicular to the optical axis of the illumination source 1601 and parallel to the normal of the sample 1701 under test. The optical axis of the illumination source 1701 is perpendicular to the normal of the sample 1701 under test. By using this embodiment, when the surface of the sample under test is a mirror or a smooth surface, coaxial mounting can be used to obtain clear defect images and improve the defect detection effect.

[0137] In one possible implementation, the transmission frame is as follows: Figure 20As shown, the light emitted by the illumination source 1601 is projected onto the sample 1701 to be tested, and then transmitted through the sample 1701 to the image acquisition device 1602. The optical axis of the image acquisition device 1602 and the optical axis of the illumination source 1601 are parallel to the normal of the sample 1701. Using this embodiment, when the sample to be tested is a transparent or translucent material, the transmission method can be used to detect internal defects in the sample.

[0138] In one possible implementation, the above-mentioned mounting methods can be combined to achieve all-round, multi-angle inspection of complex workpieces.

[0139] This application also provides an electronic device, such as... Figure 21 As shown, it includes:

[0140] Memory 2101 is used to store computer programs;

[0141] When processor 2102 executes the program stored in memory 2101, it performs the following steps:

[0142] Acquire multiple target projection patterns, wherein the target projection patterns include at least a portion of the preset projection patterns;

[0143] Control the illumination source to project the multiple target projection patterns onto the sample to be tested;

[0144] The image acquisition device is controlled to acquire projected images corresponding to the multiple target projection patterns;

[0145] Using the image processing algorithm configured for the multiple target projection patterns, the projected images corresponding to the multiple target projection patterns are processed to obtain the result images corresponding to the multiple target projection patterns;

[0146] Defect detection is performed based on multiple result images to obtain the target defect detection results of the sample to be tested.

[0147] Furthermore, the aforementioned electronic device may also include a communication bus and / or a communication interface, with the processor 2102, the communication interface, and the memory 2101 communicating with each other via the communication bus.

[0148] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0149] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0150] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0151] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0152] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the above-described defect detection methods.

[0153] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the defect detection methods described above.

[0154] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a solid-state drive (SSD), etc.

[0155] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0156] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0157] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A defect detection method, characterized in that, The method includes: Acquire multiple target projection patterns, wherein the target projection patterns include at least a portion of the preset projection patterns; Control the illumination source to project the multiple target projection patterns onto the sample to be tested; The image acquisition device is controlled to acquire projected images corresponding to the multiple target projection patterns; Using the image processing algorithm configured for the multiple target projection patterns, the projected images corresponding to the multiple target projection patterns are processed to obtain the result images corresponding to the multiple target projection patterns; Defect detection is performed based on multiple result images to obtain the target defect detection results of the sample to be tested.

2. The method according to claim 1, characterized in that, The preset projection pattern includes at least one of the following: a projection pattern for photometric stereoscopic method, a projection pattern for phase deflection method, a projection pattern for phase profilometry, and a projection pattern for structured light projection technology.

3. The method according to claim 1, characterized in that, The acquisition of multiple target projection patterns includes: A target projection pattern combination is determined from a preset projection pattern combination, wherein the preset projection pattern combination is either pre-set or user-defined. The preset projection pattern indicated by the pattern selection instruction is obtained from the target projection pattern combination and used as the target projection pattern.

4. The method according to claim 3, characterized in that, Each of the preset projection pattern combinations is pre-configured with a corresponding detection scene, and the method further includes: Determine the current detection scenario; Determining the target projection pattern combination from the preset projection pattern combinations includes: The projection pattern combination that matches the current detection scene in the preset projection pattern combination is identified as the target projection pattern combination.

5. The method according to claim 1, characterized in that, The acquisition of multiple target projection patterns includes: Demonstrates at least one preset image processing algorithm and a preset projected image configured for each of the preset image processing algorithms; Obtain the preset projection pattern indicated by the pattern selection command as the target projection pattern.

6. The method according to claim 5, characterized in that, At least one of the preset projection patterns is imported before receiving the pattern selection instruction and after receiving the pattern import instruction, and the method further includes: An image processing algorithm is assigned to the projection pattern indicated by the pattern import instruction.

7. The method according to claim 1, characterized in that, Controlling the illumination source to project the multiple target projection patterns includes: Each target projection pattern is divided into multiple pattern groups according to the type of image processing algorithm configured for each target projection pattern, and the image processing algorithms configured for each target projection pattern in the same pattern group are the same; The illumination source is controlled to sequentially project each of the target projection patterns in each of the pattern groups.

8. The method according to claim 1, characterized in that, The method further includes: Display the result images corresponding to the multiple target projection patterns; The defect detection based on multiple result images includes: Defect detection is performed based on the result image corresponding to the target projection pattern indicated by the image selection command.

9. The method according to claim 1, characterized in that, The controlled image acquisition device acquires projected images corresponding to the multiple target projection patterns, including: The image acquisition device is controlled to acquire a preset number of image frames to obtain the original image. The target projection pattern of the sample to be tested is different in each image frame. The original image is split according to the image processing algorithm configured for each of the target projection patterns to obtain the projection images corresponding to the multiple target projection patterns.

10. A defect detection system, characterized in that, The defect detection system includes: an illumination source, an image acquisition device, and a controller; The illumination source is used to project a target projection pattern, the target projection pattern including at least a portion of a preset projection pattern; The image acquisition device is used to acquire the projected image corresponding to the target projection pattern; The controller is configured to acquire multiple target projection patterns; control an illumination source to project the multiple target projection patterns onto the sample to be tested; control an image acquisition device to acquire projection images corresponding to the multiple target projection patterns; process the projection images corresponding to the multiple target projection patterns using an image processing algorithm configured for the multiple target projection patterns to obtain result images corresponding to the multiple target projection patterns; and perform defect detection based on the multiple result images to obtain the target defect detection results of the sample to be tested.

11. The defect detection system according to claim 10, characterized in that, The light emitted by the illumination source is projected onto the sample to be tested, and after being reflected by the sample to be tested, it enters the image acquisition device. There is a first preset angle between the optical axis of the image acquisition device and the normal of the sample to be tested, and there is a second preset angle between the optical axis of the illumination source and the normal of the sample to be tested.

12. The defect detection system according to claim 10, characterized in that, The defect detection system further includes a beam splitter; the light emitted by the illumination source is transmitted through the beam splitter and projected onto the sample to be tested, reflected by the sample to be tested and then irradiated by the beam splitter, and reflected by the beam splitter and then enters the image acquisition device. The optical axis of the image acquisition device is perpendicular to the optical axis of the illumination source and perpendicular to the normal of the sample to be tested, and the optical axis of the illumination source is parallel to the normal of the sample to be tested.

13. The defect detection system according to claim 10, characterized in that, The defect detection system further includes a beam splitter; the light emitted by the illumination source is reflected by the beam splitter and projected onto the sample to be tested, and after being reflected by the sample to be tested, it is transmitted through the beam splitter to the image acquisition device. The optical axis of the image acquisition device is perpendicular to the optical axis of the illumination source and parallel to the normal of the sample to be tested, and the optical axis of the illumination source is perpendicular to the normal of the sample to be tested.

14. The defect detection system according to claim 10, characterized in that, The light emitted by the illumination source is projected onto the sample to be tested, and transmitted through the sample to the image acquisition device. The optical axis of the image acquisition device and the optical axis of the illumination source are parallel to the normal of the sample to be tested.

Images