An image processing method, device, storage medium, and program product

Abstract

The image processing method, device, storage medium and program product provided in the embodiments of the present application, the method comprises: acquiring acquisition images of a detection target under multiple imaging channels, and determining a frequency spectrum shift image corresponding to each acquisition image. The multiple frequency spectrum shift images are subjected to frequency spectrum fusion, and the post-fusion frequency spectrum image is subjected to defect enhancement iteration to obtain a target acquisition image. In the technical solution provided in the present application, each acquisition image is converted into frequency domain data, the frequency spectrum position is adjusted based on the acquisition angle, which lays a foundation for subsequent frequency spectrum fusion and high-precision acquisition image determination. Moreover, through frequency spectrum fusion and optimization iteration, the determination precision of the target acquisition image is significantly improved, and the defect detection efficiency of the dark field detection process is improved.

Description

Technical Field

[0001] This application belongs to the field of optical inspection technology, and in particular relates to an image processing method, device, storage medium and program product. Background Technology

[0002] Currently, dark field inspection technology has become an indispensable inspection method in the field of optical inspection technology due to its high-contrast imaging advantage for detecting tiny defects on the surface. It plays a key role in many fields that require nanoscale surface analysis, such as semiconductor advanced process yield control, materials science, and biomedicine.

[0003] Traditional dark-field inspection technology mainly acquires target images and analyzes defects through multi-channel direct observation. However, this method does not achieve effective comprehensive analysis of multi-channel acquired data. Ultimately, the imaging resolution is still limited by the numerical aperture of a single objective lens, making it difficult to further improve the acquisition accuracy and seriously affecting the detection accuracy of dark-field inspection tasks.

[0004] Therefore, how to perform high-precision image acquisition in dark field detection is an important problem that urgently needs to be solved. Summary of the Invention

[0005] This application provides an image processing method, device, storage medium, and program product that can significantly improve the image acquisition accuracy during dark field detection.

[0006] In a first aspect, embodiments of this application provide an image processing method, including: Acquire multiple images of the target object across multiple imaging channels; For each acquired image, based on the acquisition angle of the acquired image, a frequency domain transformation is performed on the acquired image to obtain the spectrum image of the acquired image, and the spectrum position of the spectrum image of the acquired image is adjusted to obtain the spectrum shift image of the acquired image; Multiple spectral shift images are fused to obtain a fused spectral image; Based on preset defect enhancement parameters, the fused spectral image is subjected to at least one round of defect enhancement processing to determine the target acquisition image of the detection target.

[0007] Secondly, embodiments of this application provide an image processing apparatus, including: The image acquisition unit is used to acquire multiple images of the target being detected in multiple imaging channels; The spectrum determination unit is used to perform frequency domain transformation on each acquired image based on the acquisition angle of the acquired image to obtain the spectrum image of the acquired image, and to adjust the spectrum position of the spectrum image of the acquired image to obtain the spectrum shift image of the acquired image. The spectrum fusion unit is used to perform spectrum fusion on multiple spectrum shift images to obtain a fused spectrum image. The image generation unit is used to perform at least one round of defect enhancement processing on the fused spectral image based on preset defect enhancement parameters to determine the target acquisition image of the detection target.

[0008] Thirdly, embodiments of this application provide an electronic device, which includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the steps of any image processing method of the embodiments of this application.

[0009] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, and when the program or instructions are executed by a processor, the steps of any image processing method of the embodiments of this application are implemented.

[0010] Fifthly, embodiments of this application provide a computer program product in which instructions, when executed by a processor of an electronic device, enable the electronic device to perform the steps of any image processing method of embodiments of this application.

[0011] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects: This application provides an image processing method, comprising: first, acquiring images of the target at multiple acquisition angles corresponding to multiple imaging channels; then, determining the corresponding spectral image for each acquired image, and adjusting the spectral position based on the acquisition angle to determine the corresponding spectral shift image. This application converts each acquired image into frequency domain data and adjusts the spectral position based on the acquisition angle, laying the foundation for subsequent spectral fusion and the determination of high-precision acquired images. Based on this, acquired images at different acquisition angles can be accurately fused into a high-precision image of the target, significantly improving the image acquisition accuracy in the dark field detection process.

[0012] Next, spectral fusion can be performed on multiple spectral shift images. Defect enhancement iterations can then be applied to the fused spectral image, resulting in a target acquisition image with defect enhancement based on at least one iteration. Through spectral fusion and defect enhancement iterations, the acquisition accuracy of the target image can be significantly improved, thereby enhancing the detection accuracy of subsequent defect detection processes.

[0013] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A schematic diagram of an image acquisition structure using a multi-path objective lens in the prior art; Figure 2 This is a schematic diagram of an image acquisition structure using a single-path objective lens in the prior art. Figure 3 A schematic flowchart of an image processing method provided in one embodiment of this application; Figure 4(a) is one of the structural schematic diagrams of a dark field detection device provided in an embodiment of this application; Figure 4(b) is a second schematic diagram of the structure of a dark field detection device provided in one embodiment of this application; Figure 5 A schematic diagram illustrating a spectrum fusion process according to an embodiment of this application; Figure 6 This is a schematic diagram of the structure of an image processing apparatus provided in another embodiment of this application; Figure 7 This is a schematic diagram of the hardware structure of an image processing device provided in another embodiment of this application. Detailed Implementation

[0016] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0017] It should be noted that, in this document, relational terms such as "first" and "second" are used merely 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..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0018] Dark-field inspection technology, with its high-contrast imaging advantage for minute surface defects, has become one of the core technologies in the field of optical inspection. It is widely used in fields with strict requirements for nanoscale structure analysis, such as yield control of advanced semiconductor processes, material surface morphology analysis, and biomedical microscopy.

[0019] However, existing dark-field detection methods have serious shortcomings in improving imaging resolution. Traditional dark-field detection schemes mainly acquire the surface morphology of the object under inspection through multiple acquisitions. Although images can be acquired simultaneously through multiple objectives, effective comprehensive analysis of the acquired data is not possible. The final image resolution of this method is limited by the numerical aperture of a single objective, resulting in low acquisition accuracy.

[0020] Furthermore, this method requires careful consideration of the interference effects between multiple optical paths. The larger the numerical aperture of each path, the more severe the interference effect, potentially necessitating high-precision hardware fabrication of the objective lens, significantly increasing inspection costs. For details on this type of inspection method, please refer to [link / reference needed]. Figure 1 As shown in the example.

[0021] Figure 1 This is a schematic diagram of an existing multi-path objective lens for image acquisition.

[0022] like Figure 1 As shown in the example, the detection surface 101 of the target object can acquire multiple images through three imaging channels 102, 103, and 104. Each imaging channel consists of a camera, a telescope, and an objective lens. To ensure image acquisition accuracy, the numerical aperture (NA) of each objective lens in this example is relatively high, for example, NA=0.5.

[0023] Since the numerical apertures of the multiple objectives are the same, in this example, to avoid optical path interference problems caused by high numerical apertures and direct proximity, the overlapping interference regions of the three objectives were cut off at adjacent locations. Figure 1 The cutting areas 105 and 106 are shown in the diagram. This type of hardware processing for the objective lens requires extremely high precision, and the success rate cannot be guaranteed, which significantly increases the cost of image acquisition and inspection.

[0024] Furthermore, in this example, only redundant verification is performed on the acquired images from each channel, such as verifying the authenticity of defects in the defect detection process based on images from three channels. There is no effective integration of images from multiple acquisition angles for comprehensive image analysis of the detection target, and the final image resolution is based solely on the numerical aperture of each channel, resulting in low acquisition accuracy.

[0025] In addition, there are dark-field inspection methods that employ high numerical aperture (NA=0.9) single-channel objectives to improve image acquisition accuracy. While this method can improve the resolution of the acquired image to some extent, the optical manufacturing cost of high numerical aperture objectives is extremely high, drastically increasing the cost of inspection equipment in practical use. Furthermore, single-channel image acquisition has blind spots, resulting in inaccurate data acquisition of certain areas of the target, leading to an increased false negative rate in dark-field inspection and severely impacting the accuracy and reliability of the inspection results. For details on this type of inspection method, please refer to [link / reference needed]. Figure 2 As shown in the example.

[0026] Figure 2 This is a schematic diagram of an image acquisition structure using a single-path objective lens in the prior art.

[0027] like Figure 2 As shown, a single-channel image of the inspection surface 204 of the target can be acquired using a camera 201, a telescope 202, and a high numerical aperture objective lens 203. This acquisition method requires a high numerical aperture objective lens with extremely high manufacturing costs to ensure image acquisition accuracy. However, because the dark-field scattered light at the edge of the high numerical aperture objective lens cannot be received due to optical path obstruction (e.g., in semiconductor defect detection scenarios such as the edge region of a wafer), data from some inspection areas cannot be accurately acquired, leading to a risk of missed detections during defect inspection.

[0028] Furthermore, high numerical aperture (NA) lenses have limitations in depth of field; for example, the depth of field of an objective lens with NA=0.9 is only one-third that of an objective lens with NA=0.5. Therefore, objectives with higher numerical apertures place higher demands on the performance of the autofocus system in the image acquisition equipment during dark-field detection, further increasing costs.

[0029] To address the aforementioned technical problems, this application provides an image processing method, device, storage medium, and program product. The method includes: acquiring images of a detected target at multiple acquisition angles corresponding to multiple imaging channels. For each acquired image, a corresponding spectral image can be determined, and based on the acquisition angle, the spectral position can be adjusted to determine a corresponding spectral shift image.

[0030] Next, multiple spectral shift images can be fused, and defect enhancement iterations can be performed on the fused spectral images. Based on at least one iteration, a target acquisition image of the detected target after defect enhancement can be obtained.

[0031] The technical solution provided in this application converts each acquired image into frequency domain data and adjusts the spectral position based on the acquisition angle, laying the foundation for subsequent spectral fusion and the determination of high-precision acquired images. Based on this, acquired images from different acquisition angles can be accurately fused into a high-precision image of the detection target, significantly improving the image acquisition accuracy in the dark field detection process.

[0032] This application significantly improves the acquisition accuracy of target images and subsequent defect detection accuracy through spectral fusion and defect enhancement iteration. Furthermore, the technical solution provided in this application can achieve high-precision image acquisition based on multiple imaging channels with lower numerical apertures. The objective lenses of the lower numerical aperture imaging channels have smaller mechanical diameters, and the combination of objective lenses from multiple imaging channels avoids issues similar to those encountered in other applications. Figure 1 The problem of optical path interference in the dark field is eliminated, so there is no need to cut and process the objective lens, which greatly reduces the cost of the objective lens and processing, saves hardware processing time, and further improves the efficiency of dark field detection.

[0033] The execution entity used in the embodiments of this application can be the dark field detection device itself, or a terminal device capable of controlling the dark field detection device, such as a desktop computer or laptop computer, or a remote device, such as a server. In addition, the execution entity used in the embodiments of this application can also be a software entity, such as a client or software program installed on the dark field detection device or terminal device. This application does not strictly limit the execution entity used in applying the technical solutions provided in the embodiments of this application; it can be flexibly selected and set according to actual application scenarios and needs.

[0034] The specific application scenarios of the image processing methods, devices, storage media, and program products provided in the embodiments of this application are not strictly limited in this application, and can be determined according to actual needs.

[0035] For example, in the semiconductor field, in practical scenarios where dark-field inspection equipment is used to detect defects on the surface of a wafer, the dark-field inspection equipment using the technical solution provided in this application can acquire images of the target wafer surface at multiple acquisition angles through multiple imaging channels. Then, based on the acquisition angle of each acquired image, the position of the spectral image of each acquired image can be adjusted to obtain spectral shift images corresponding to multiple acquisition angles.

[0036] Furthermore, spectral fusion can be performed based on multiple spectral shift images to obtain a fused spectral image of the target wafer. For the fused spectral image, the technical solution provided in this application can also perform at least one round of defect enhancement processing to obtain a target acquisition image of the target wafer after defect enhancement. Based on the technical solution of this application, high-precision acquisition images can be obtained through multiple low numerical aperture imaging channels, reducing inspection costs while improving the accuracy of the acquisition images through multi-channel image fusion and defect enhancement. The acquisition images obtained based on the technical solution of this application can significantly improve the defect detection accuracy of subsequent defect detection processes on the target wafer, ensuring wafer production quality and yield.

[0037] It should be noted that the application scenarios described in the above embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a specific limitation on the application scenarios of the technical solutions provided in the embodiments of this application. Those skilled in the art will understand that with the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems. The image processing methods, devices, storage media, and program products provided in the embodiments of this application can be applied to various practical application scenarios that require dark-field detection of detection targets.

[0038] Figure 3 This is a schematic flowchart illustrating an image processing method according to one embodiment of this application. Figure 3 As shown in the figure, the image processing method provided in this application embodiment includes steps S301 to S304.

[0039] S301: Acquire multiple images of the target in multiple imaging channels.

[0040] In step S301, the technical solution provided in this application embodiment can acquire images of the target at multiple acquisition angles using a dark-field detection device that includes multiple imaging channels with the same numerical aperture. The specific type of the target is not strictly limited; in some embodiments, the target may be a semiconductor wafer, graphene film, fiber optic end face, etc., and can be flexibly selected according to actual needs and application scenarios. The acquisition angle represents the azimuth angle of each imaging channel relative to the horizontal plane where the target is located.

[0041] This application does not strictly limit the number of imaging channels in the dark field detection device or the specific structural composition of each imaging channel; these can be flexibly set according to actual needs and application scenarios. In some embodiments, the specific structure of the part of the dark field detection device used to acquire and detect the target image can be referred to the examples shown in Figures 4(a) and 4(b).

[0042] Figures 4(a) and 4(b) are schematic diagrams of a dark field detection device provided in one embodiment of this application.

[0043] Figure 4(a) is a schematic diagram of a dark field detection device with three imaging channels provided in this application. As shown in Figure 4(a), in this example, the detection surface 404 of the detection target can be acquired from different acquisition angles through the three imaging channels 401, 402 and 403 respectively.

[0044] In the embodiments provided in this application, the three imaging channels in Figure 4(a) can select objectives with a numerical aperture not greater than a preset low aperture threshold, such as 0.3, 0.35, or 0.4. Based on this, compared to the above... Figure 1 The example shown greatly reduces the mechanical diameter of the objective lens in each imaging channel, solving the optical path interference problem from a hardware perspective, thus eliminating the need for further cutting and processing of the objective lens and significantly saving processing costs.

[0045] The accuracy issues caused by the reduced numerical aperture can be resolved through subsequent frequency domain processing and comprehensive frequency domain analysis of multiple acquired images. This allows multiple low-numerical-aperture objectives to achieve image acquisition equivalent to a high-numerical-aperture objective. For example, acquiring images through three objective imaging channels with a numerical aperture of 0.3 is equivalent to acquiring an image of a target with a numerical aperture of 0.9. The embodiments provided in this application effectively reduce processing costs while significantly improving image acquisition accuracy. Furthermore, acquiring images from multiple angles can expand the image acquisition range, thus addressing the aforementioned issues. Figure 2 The problem of data loss in some areas due to single-channel acquisition.

[0046] In subsequent steps, the technical solution provided in this application can fuse low numerical aperture images acquired from different angles by multiple imaging channels after spectral shifting to obtain a high-precision image equivalent to a high numerical aperture image. Furthermore, since a lower numerical aperture objective lens can be selected, the influence of optical path interference is successfully avoided, improving image acquisition accuracy while significantly saving hardware processing costs.

[0047] Figure 4(b) is a schematic diagram of a dark field detection device with five imaging channels provided in this application. As shown in Figure 4(b), in this example, two more imaging channels with the same numerical aperture of the objective lens can be added to the three channels shown in Figure 4(a), namely imaging channels 405 and 406. Figure 4(b) can further receive the scattered light from the detection surface 404. Similarly, by selecting an objective lens with a lower numerical aperture, optical path interference is avoided, thus eliminating the need for processing the objective lens and saving costs.

[0048] The five imaging channels can acquire images of the detection surface 404 of the target at five different acquisition angles, thus providing sufficient practical data for subsequent frequency domain analysis and target image determination, further improving the accuracy of the acquired images and the accuracy of defect detection of the target.

[0049] In addition to the examples above, the dark field detection device in other embodiments can also consist of other numbers of imaging channels, such as four imaging channels, seven imaging channels, etc., which can be flexibly selected and set according to actual needs and application scenarios.

[0050] S302: For each acquired image, based on the acquisition angle of the acquired image, perform frequency domain transformation on the acquired image to obtain the spectrum image of the acquired image, and adjust the spectrum position of the spectrum image of the acquired image to obtain the spectrum shift image of the acquired image.

[0051] In step S302, the technical solution provided in this application embodiment can perform frequency domain conversion on the acquired image at each acquisition angle, and adjust the spectral position of the acquired image's spectral image based on the acquisition angle to obtain the spectral displacement image of each acquired image, thereby providing practical data for the subsequent spectral fusion process through spectral displacement.

[0052] In this application, we take into account that the detection target may correspond to different material types, such as silicon wafer substrate in wafer defect detection scenarios, metal particles on the wafer (copper, aluminum, etc.), etc. Different material types correspond to different scattering types. Introducing the scattering effect caused by the material type into the acquisition process can effectively improve the accuracy of the target acquisition image and make it closer to the real situation.

[0053] Based on this, in the embodiments provided in this application, the scattering parameters of the acquired images at each acquisition angle can be determined based on the material type of the target being detected. Simultaneously, the acquired images can be signal converted to obtain the acquisition signal data corresponding to each acquired image for frequency domain conversion. The specific signal conversion process can be referred to as shown in the following formula (1): Formula (1) Among them, for the first One captured image, Indicates the first The acquired image itself, and The first The horizontal and vertical coordinates of each pixel in the acquired image are not limited to a Cartesian coordinate system in this application. They can be established based on a preset reference point in the acquired image (such as the center point of the image, the lower left corner of the image, etc.). In order to target the Each acquired image corresponds to a preset point spread function for the imaging channel. This is a noise signal. For the first The acquired signal data corresponding to each acquired image.

[0054] Formula (1) can be used to determine the point spread function corresponding to each acquired image. While considering noise signals The effect is to convert each acquired image from a two-dimensional image into signal data, thus obtaining the corresponding acquired signal data.

[0055] For the acquired signal data corresponding to each acquired image, frequency domain transformation processing such as Fourier Transform (FT) can be performed to determine the spectrum image to be corrected for each acquired image. The specific transformation process can be referred to as shown in the following formula (2): Formula (2) Among them, for the first One captured image, Similar to formula (1) above, it represents the first... The acquired signal data corresponding to each acquired image Indicates the acquisition of signal data Perform Fourier transform The resulting spectrum image to be corrected and They represent The coordinates of each acquired image are The frequency components of the pixels in the horizontal axis and the frequency components in the vertical axis.

[0056] Next, based on the scattering parameters of each acquired image, scattering correction can be performed on the spectrum image to be corrected for each acquired image, so as to incorporate the different real scattering effects caused by different material types into the spectrum image, thus obtaining the spectrum image of each acquired image after scattering correction. The specific scattering correction process can be referred to as shown in the following formula (3): Formula (3) Among them, for the first One captured image, For the first The spectral image of a acquired image after scattering correction. The first one obtained by formula (2) above The spectrum image to be corrected for each acquired image. Indicates the first The scattering parameters of each acquired image can be based on the illumination wavelength of the dark field detection device. and the The acquisition angle of each acquired image is determined.

[0057] In dark field detection scenarios, scattered light from the surface of the target is mainly collected. However, different material types, such as the inherent properties of the material's microstructure, refractive index, particle size and shape, will have different effects on the intensity of scattered light at different image acquisition angles. This will cause the acquired images of the target on different imaging channels to deviate due to differences in scattering intensity. If this deviation is not addressed, it will seriously affect the accuracy of the fused image in the subsequent spectral fusion process.

[0058] Based on the above embodiments, the scattering effects of different material types can be introduced during the frequency domain conversion of acquired images, and spectral correction can be performed based on the corresponding scattering parameters. This embodiment can effectively compensate for the differences in scattering intensity caused by different receiving angles and material types in each imaging channel during image acquisition, thereby ensuring the consistency between multiple spectral images while improving the realism and accuracy of the spectral images. This lays an accurate data foundation for the subsequent spectral fusion process, and ultimately improves the detection accuracy of the dark field detection process.

[0059] Regarding the specific process of determining the scattering parameters of each acquired image, in one embodiment provided in this application, the scattering model parameters corresponding to the detected target can be determined based on the material type of the detected target and the pre-built correspondence between the material type and the scattering model parameters.

[0060] Then, based on the scattering model parameters, the preset illumination wavelength of the dark field detection device, and the acquisition angle of the acquired image, the scattering parameters corresponding to the acquired image at different acquisition angles can be determined, such as those shown in formula (3) above. .

[0061] The scattering model parameters are used to simulate the scattering effects of images acquired under different material types and acquisition angles through a scattering simulation model. The specific model type and structure are not strictly limited in this embodiment and can be flexibly set according to actual needs and application scenarios. In some embodiments, the scattering simulation model can be a Mie scattering model, a Rayleigh scattering model, a geometrical optics approximation model (GO), etc., which can be flexibly selected and used.

[0062] In the above embodiments, the corresponding scattering model parameters and scattering simulation model can be accurately determined based on the material type of the target being detected. This effectively improves the accuracy and adaptability of scattering correction for spectral data of different material types and at different acquisition angles, fully ensuring the accuracy and reliability of spectral correction results from multiple acquisition angles, making the spectral image more realistic and accurate, and further improving the determination accuracy of the target acquisition image and the detection accuracy of the subsequent dark field detection process.

[0063] Furthermore, after determining the spectral image at each acquisition angle, spectral shifting is required to eliminate spectral offsets caused by different acquisition angles in order to ensure the accuracy of the subsequent spectral fusion process. Spectral shifting allows for precise positional calibration of the spectral images at different acquisition angles, eliminating image deviations caused by different acquisition angles and fully guaranteeing the accuracy of the spectral fusion process.

[0064] Based on this, for each acquired image's spectral image, the corresponding spectral displacement distance can be determined according to the acquisition angle corresponding to that image. The spectral displacement distance corresponding to the center angle of multiple acquisition angles can be zero, meaning no spectral displacement processing is required. For acquired images at other acquisition angles, the spectral displacement distance can be determined based on the spectral image at the center angle. The specific method for determining the spectral displacement distance can be flexibly selected according to actual needs. In some embodiments, it can be to calculate the spectral translation vector caused by the change in acquisition angle in the frequency domain space, and further derive the spectral displacement distance corresponding to the spectral image at different acquisition angles based on the spectral translation vector.

[0065] The aforementioned center angle refers to, for example, the acquisition angle of imaging channel 402 in Figure 4(a), or the acquisition angle of imaging channel 402 in Figure 4(b), that is, the acquisition angle perpendicular to the surface of the target being detected. The non-center angle refers to other acquisition angles besides the center angle.

[0066] Furthermore, the position of each frequency component in the spectrum image of the acquired image is shifted according to the spectral displacement distance to obtain the spectral displacement image of each acquired image.

[0067] For each acquisition angle, the spectral displacement distance represents the direction and magnitude of displacement of the spectral image at that acquisition angle. In the embodiments provided in this application, the spectral displacement distance may include the horizontal spectral displacement distance and the vertical spectral displacement distance of the acquired image. The horizontal spectral displacement distance and the vertical spectral displacement distance respectively represent the displacement distance of each frequency component in the spectral image of the acquired image along the horizontal axis and the vertical axis in a two-dimensional Cartesian coordinate system established with a preset reference point (e.g., the center point, the lower left corner, etc.) in the spectral image as the origin.

[0068] Specifically, the process of determining the lateral spectral displacement distance and the longitudinal spectral displacement distance can be referred to the following formula (4): Formula (4) Among them, for the first One captured image, Indicates that it is for the first The lateral spectral displacement distance of each acquired image. Indicates the first In the embodiments provided in this application, the lateral spectral displacement distance at each acquisition angle is the same as the longitudinal spectral displacement distance for each acquired image. Indicates the first The acquisition angle of each acquired image. The illumination wavelength for dark field detection equipment.

[0069] The spectral displacement distance of the spectrum image at each acquisition angle can be accurately calculated using the above formula (4). Based on the spectral displacement distance, the position of the spectrum image can be further adjusted to determine the spectral displacement image at each acquisition angle.

[0070] Specifically, in the embodiments provided in this application, for each non-center angle acquired image, the lateral displacement of each frequency component on the spectrum image can be performed based on the aforementioned lateral spectrum displacement distance.

[0071] Simultaneously, based on the vertical spectral displacement distance of the acquired image, each frequency component on the spectral image can be vertically shifted. The spectral image after both horizontal and vertical displacements is the spectral displacement image of the acquired image. The specific spectral displacement process described above can be referred to in the following formula (5): Formula (5) Among them, for the first One captured image, Indicates the first Spectral shift image of each acquired image, Indicates the first Spectral image of each acquired image Lateral spectral shift distance for each frequency component Lateral translation and longitudinal spectral displacement distance The longitudinal translation.

[0072] Through the displacement adjustment process, such as that in formula (5), the spectral displacement images of the dark field detection device at different acquisition angles can be accurately determined. The spectral displacement images after spectral displacement can be used in the subsequent spectral fusion process to effectively eliminate the spectral shift caused by different acquisition angles, improve the determination accuracy of the fused spectral images, and thus improve the determination accuracy of the target acquisition images.

[0073] S303: Perform spectral fusion on multiple spectral shift images to obtain a fused spectral image.

[0074] In step S303, the technical solution provided by this application can perform spectral stitching on the spectral shift image that has undergone scattering correction and spectral shifting in the above steps to determine a fused spectral image of multiple spectral shift images.

[0075] In the embodiments provided in this application, spectral displacement images corresponding to multiple acquisition angles can be superimposed and stitched together using the same reference point as a reference, thereby fusing the spectral displacement images from multiple acquisition angles into a complete and continuous global spectral image, which serves as the aforementioned fused spectral image. The reference point can be any point at the same location in the multiple spectral displacement images, such as the image center point. The above spectral stitching process can be referred to... Figure 5 As shown.

[0076] Figure 5 This is a schematic diagram of a spectrum fusion process provided in one embodiment of this application.

[0077] like Figure 5 As shown in the example, the dark field detection device includes five imaging channels 501-505, each of which can acquire a corresponding image 506. The processing procedure of step S302 described above is performed on each acquired image, that is, the spectral image 507 of each acquired image can be obtained by frequency domain transformation and scattering correction.

[0078] Then, based on the acquisition angle corresponding to each acquired image, the spectral image 507 can be corrected for spectral position offset, resulting in a spectral displacement image 508 after calibration to correct the spectral offset caused by different acquisition angles. Since the distribution of the scattered signal in the spatial frequency domain is shifted due to incident light at different angles, each spectral image needs to be mapped and aligned according to its corresponding angle information.

[0079] In some embodiments, the acquisition angle can be divided into a horizontal axis acquisition angle and a vertical axis acquisition angle. Specifically, the acquisition angle corresponding to imaging channel 501 can be X=30°, Y=30°; the acquisition angle corresponding to imaging channel 502 can be X=30°, Y=-30°; the acquisition angle corresponding to imaging channel 503 can be X=0°, Y=0°, which represents the center angle of the acquisition angle corresponding to imaging channel 503; the acquisition angle corresponding to imaging channel 504 can be X=-30°, Y=30°; and the acquisition angle corresponding to imaging channel 505 can be X=-30°, Y=-30°. In this example, when the horizontal axis angle (X) is positive, the spectral components of the corresponding spectrum image shift to the left, and vice versa. When the vertical axis acquisition angle (Y) is positive, the spectral components of the corresponding spectrum image shift upward, and vice versa. The above angle parameters are only for illustrative purposes and can be flexibly set according to actual needs and application scenarios.

[0080] Furthermore, multiple spectral shift images can be stitched together based on the same image center point, such as... Figure 5 As shown, a fused spectral image 509 is obtained from multiple spectral shift images. The final fused spectral image 509 contains complete spatial frequency information acquired from multiple angles, covering a wider spatial range, which is equivalent to achieving a high numerical aperture imaging effect. Subsequently, by performing inverse Fourier transform and subsequent defect enhancement processing on the fused spectral image, a high-resolution target acquisition image can be reconstructed, significantly improving the defect detection capability.

[0081] The process of determining the above-mentioned fused spectral image can be referred to the following formula (6): Formula (6) in, Representing multiple spectral shift images The fused spectrum image obtained after spectrum stitching contains frequency domain features of images acquired from multiple different angles, which can significantly improve the image accuracy of the corresponding spatial domain image. This enables the determination of a high-resolution acquired image based on multiple low numerical aperture acquired images in the subsequent process of determining the target acquired image.

[0082] S304: Based on preset defect enhancement parameters, perform at least one round of defect enhancement processing on the fused spectral image to determine the target acquisition image of the detection target.

[0083] In step S304, this application considers that the target image acquired during the dark field detection process can serve as a practical basis for defect detection. To further improve the detection accuracy and efficiency of the defect detection process, the technical solution provided in this application embodiment can perform at least one round of defect enhancement processing on the fused spectrum image based on preset defect enhancement parameters during spatial domain transformation, thereby determining a target acquisition image that contains high-precision detection targets and has a defect enhancement effect.

[0084] In the embodiments provided in this application, the fused spectral image determined in the above steps can first be spatially transformed from frequency domain data to a spatial domain image to determine the image to be enhanced for the target being detected. The specific processing procedure can be referred to as shown in the following formula (7): Formula (7) in, To fuse the spectral images, the inverse Fourier transform shown in formula (7) is used. The fused spectral image can be converted from the frequency domain to the spatial domain to obtain the image to be enhanced. .

[0085] Then, based on the defect enhancement parameters, at least one round of defect enhancement processing can be performed on the fused spectral image and the image to be enhanced to determine the target acquisition image after defect enhancement.

[0086] Specifically, in one embodiment provided in this application, in each round of defect enhancement processing, the image to be enhanced in that round can be frequency domain transformed to obtain the spectral image of the image to be enhanced in that round.

[0087] Then, the Euclidean norm (L2 norm) of the spectral difference image between the image to be enhanced and the fused spectral image in this round can be calculated. Simultaneously, the Manhattan norm (L1 norm) of the image to be enhanced in this round can be determined.

[0088] Based on the Euclidean norm and Manhattan norm calculated in this round, as well as the defect enhancement parameters, the defect enhancement loss for this round can be determined. Further, based on the defect enhancement loss, defect enhancement is performed on the image to be enhanced in this round, resulting in a defect-enhanced image. This defect-enhanced image is then used as the image to be enhanced in the next round.

[0089] After a maximum number of defect enhancement rounds, the image to be enhanced corresponding to the round with the minimum defect enhancement loss can be used as the target acquisition image of the detection target. Alternatively, in other embodiments, the iteration process can be stopped when the defect enhancement loss is less than a preset loss threshold, and the image to be enhanced in the last round can be used as the target acquisition image. The above defect enhancement process can be specifically referred to as the following formula (8): Formula (8) Among them, the defect enhancement processing for each round, For the images to be enhanced in this round, To fuse spectral images, This represents a pre-defined regularization parameter based on the sparse distribution characteristics of defects during defect detection. It can be flexibly set according to actual needs and application scenarios, serving as the aforementioned pre-defined defect enhancement parameter.

[0090] This represents the Euclidean norm of the spectral difference image between the image to be enhanced and the fused spectral image in this round. This represents the Manhattan norm of the image to be enhanced in this round. This indicates the defect enhancement loss in this round, through... The image to be enhanced with the lowest defect enhancement loss can be selected from multiple rounds of defect enhancement processing. The target image is acquired as the detection target.

[0091] The above embodiments allow for the construction of an iteratively optimized loss function based on defect enhancement parameters, the image to be enhanced, and the fused spectral image. This iterative optimization mechanism separates and enhances weak defect signals from the complex background during the iteration process. This enhances the accuracy of the acquired target image while making defect information and features increasingly prominent in the image, significantly improving the detection accuracy and efficiency of defect detection in dark-field detection, thereby improving the production quality and yield of the detected target.

[0092] The above describes the specific implementation of the image processing method provided in this application. The technical solution provided in this application converts each acquired image into frequency domain data and adjusts the spectral position based on the acquisition angle, laying the foundation for subsequent spectral fusion and the determination of high-precision acquired images. This process fully considers the scattering influence of different material types on the acquired images, introduces scattering models and scattering parameters, and performs scattering correction on the frequency domain data, improving the accuracy and realism of the spectral image determination. Based on this, acquired images from different acquisition angles can be accurately fused into a high-precision image of the detection target, significantly improving the image acquisition accuracy in the dark field detection process.

[0093] Furthermore, this application significantly improves the acquisition accuracy of target images and subsequent defect detection accuracy through spectral fusion and missing term enhancement iteration. Moreover, the technical solution provided in this application can achieve high-precision image acquisition based on multiple imaging channels with lower numerical apertures. The objective lenses of the lower numerical aperture imaging channels have smaller mechanical diameters, and the combination of objective lenses from multiple imaging channels avoids issues similar to those encountered in other applications. Figure 1 The problem of optical path interference in the dark field is eliminated, so there is no need to cut and process the objective lens, which greatly reduces the cost of the objective lens and processing, saves hardware processing time, and further improves the efficiency of dark field detection.

[0094] Based on the image processing method provided in the above embodiments, this application also provides specific implementations of the image processing apparatus, as described in the following embodiments.

[0095] Figure 6 This is a schematic diagram of the structure of an image processing apparatus provided in another embodiment of this application. (See attached diagram.) Figure 6 As shown, the image processing apparatus 600 includes: The image acquisition unit 601 is used to acquire multiple images of the target under multiple imaging channels; The spectrum determination unit 602 is used to perform frequency domain transformation on each acquired image based on the acquisition angle of the acquired image to obtain the spectrum image of the acquired image, and to adjust the spectrum position of the spectrum image of the acquired image to obtain the spectrum shift image of the acquired image. The spectrum fusion unit 630 is used to perform spectrum fusion on multiple spectrum shift images to obtain a fused spectrum image; The image generation unit 604 is used to perform at least one round of defect enhancement processing on the fused spectrum image based on preset defect enhancement parameters to determine the target acquisition image of the detection target.

[0096] In one embodiment, the spectrum determination unit 602 is specifically used for: For each acquired image, the scattering parameters of that image are determined based on the acquisition angle. The acquired image is converted into a signal to determine the corresponding acquired signal data. Perform a Fourier transform on the acquired signal data corresponding to the acquired image to determine the spectrum image to be corrected. Based on the scattering parameters of the acquired image, scattering correction is performed on the spectrum image to be corrected of the acquired image to obtain the spectrum image of the acquired image.

[0097] In one embodiment, the spectrum determination unit 602 is specifically used for: Determine the scattering model parameters of the target based on its material type; For each acquired image, the scattering parameters of the acquired image are determined based on the scattering model parameters, the preset illumination wavelength, and the acquisition angle of the acquired image.

[0098] In one embodiment, the spectrum determination unit 602 is specifically used for: For each non-center angle acquired image, the spectral displacement distance of the acquired image is determined based on the acquisition angle of the acquired image. The non-center angle refers to the acquisition angle other than the acquisition angle perpendicular to the plane where the detection target is located. Based on the spectral displacement distance of the acquired image, the position of each frequency component on the spectral image of the acquired image is translated to obtain the spectral displacement image of the acquired image.

[0099] In one embodiment, the spectrum determination unit 602 is specifically used for: For each non-central angle acquired image, the horizontal and vertical spectral displacement distances of the acquired image are determined based on the acquisition angle. The horizontal spectral displacement distance represents the displacement distance of each frequency component in the spectrum image of the acquired image along the horizontal axis of the two-dimensional Cartesian coordinate system. The vertical spectral displacement distance represents the displacement distance of each frequency component along the vertical axis of the two-dimensional Cartesian coordinate system. The two-dimensional Cartesian coordinate system is established with a reference point in the plane where the spectrum image of the acquired image is located as the origin, a first preset direction as the horizontal axis, and a second preset direction as the vertical axis. The reference point is any point in the spectrum image, and the first preset direction and the second preset direction are perpendicular to each other.

[0100] In one embodiment, the spectrum determination unit 602 is specifically used for: Based on the lateral spectral displacement distance of the acquired image, each frequency component on the spectrum image of the acquired image is horizontally shifted, and based on the longitudinal spectral displacement distance of the acquired image, each frequency component on the spectrum image of the acquired image is vertically shifted, to obtain the spectral displacement image of the acquired image.

[0101] In one embodiment, the image generation unit 604 is specifically used for: Spatial transformation is performed on the fused spectral image to obtain the image to be enhanced of the detected target; Based on the defect enhancement parameters, at least one round of defect enhancement processing is performed on the fused spectral image and the image to be enhanced to determine the target acquisition image.

[0102] In one embodiment, the image generation unit 604 is specifically used for: For each round of defect enhancement processing, the image to be enhanced in that round is transformed in the frequency domain to obtain the spectral image of the image to be enhanced in that round; Determine the Euclidean norm of the spectral difference image between the spectral image of the image to be enhanced and the fused spectral image in this round, as well as the Manhattan norm of the image to be enhanced in this round; Based on the Euclidean norm, Manhattan norm, and defect enhancement parameters of this round, determine the defect enhancement loss for this round; Based on the defect enhancement loss of this round, defect enhancement processing is performed on the image to be enhanced in this round to obtain the defect-enhanced image of this round, and the defect-enhanced image of this round is used as the image to be enhanced in the next round; The image to be enhanced in the round with the minimum defect enhancement loss in the preset maximum round is used as the target image.

[0103] Figure 7 This is a schematic diagram of the hardware structure of an image processing device provided in another embodiment of this application.

[0104] The image processing device may include a processor 701 and a memory 702 storing computer program instructions.

[0105] Specifically, the processor 701 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0106] Memory 702 may include mass storage for data or instructions. For example, and not limitingly, memory 702 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 702 may include removable or non-removable (or fixed) media. Where appropriate, memory 702 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 702 is non-volatile solid-state memory.

[0107] In a particular embodiment, memory 702 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0108] The processor 701 implements any of the image processing methods described in the above embodiments by reading and executing computer program instructions stored in the memory 702.

[0109] In one example, the image processing device may further include a communication interface 703 and a bus 710. Wherein, as... Figure 7 As shown, the processor 701, memory 702, and communication interface 703 are connected through bus 710 and complete communication with each other.

[0110] The communication interface 703 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0111] Bus 710 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 710 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0112] Furthermore, in conjunction with the image processing methods described in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the image processing methods described in the above embodiments.

[0113] This application also provides a computer program product, including a computer program, which, when executed, implements any of the image processing methods described in the above embodiments.

[0114] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0115] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0116] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0117] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0118] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. An image processing method, characterized in that, The method includes: Acquire multiple images of the target object across multiple imaging channels; For each acquired image, based on the acquisition angle of the acquired image, a frequency domain transformation is performed on the acquired image to obtain the spectrum image of the acquired image, and the spectrum position of the spectrum image of the acquired image is adjusted to obtain the spectrum shift image of the acquired image; The multiple spectral shift images are fused to obtain a fused spectral image; Based on preset defect enhancement parameters, the fused spectral image is subjected to at least one round of defect enhancement processing to determine the target acquisition image of the detection target.

2. The method according to claim 1, characterized in that, Based on the acquisition angle of the acquired image, a frequency domain transformation is performed on the acquired image to obtain its spectral image, including: For each acquired image, the scattering parameters of the acquired image are determined based on the acquisition angle of the acquired image; The acquired image is converted into a signal to determine the corresponding acquired signal data. Perform a Fourier transform on the acquired signal data corresponding to the acquired image to determine the spectrum image to be corrected. Based on the scattering parameters of the acquired image, scattering correction is performed on the spectrum image to be corrected of the acquired image to obtain the spectrum image of the acquired image.

3. The method according to claim 2, characterized in that, Based on the acquisition angle of the acquired image, the scattering parameters of the acquired image are determined, including: Based on the material type of the target being detected, determine the scattering model parameters of the target being detected; For each acquired image, the scattering parameters of the acquired image are determined based on the scattering model parameters, the preset illumination wavelength, and the acquisition angle of the acquired image.

4. The method according to claim 1, characterized in that, The spectral position of the acquired image is adjusted to obtain a spectral shift image of the acquired image, including: For each non-center angle of the acquired image, the spectral displacement distance of the acquired image is determined according to the acquisition angle of the acquired image, wherein the non-center angle refers to the acquisition angle other than the acquisition angle perpendicular to the plane where the detection target is located; Based on the spectral displacement distance of the acquired image, the position of each frequency component on the spectral image of the acquired image is translated to obtain the spectral displacement image of the acquired image.

5. The method according to claim 4, characterized in that, For each acquired image at a non-central angle, the spectral shift distance of the acquired image is determined based on the acquisition angle, including: For each non-center angle of the acquired image, the lateral spectral displacement distance and the longitudinal spectral displacement distance of the acquired image are determined according to the acquisition angle. The lateral spectral displacement distance represents the displacement distance of each frequency component in the spectrum image of the acquired image along the horizontal axis of the two-dimensional Cartesian coordinate system. The longitudinal spectral displacement distance represents the displacement distance of each frequency component along the vertical axis of the two-dimensional Cartesian coordinate system. The two-dimensional Cartesian coordinate system takes a reference point in the plane where the spectrum image of the acquired image is located as its origin. The reference point is any point in the spectrum image.

6. The method according to claim 5, characterized in that, Based on the spectral displacement distance of the acquired image, the position of each frequency component on the spectral image of the acquired image is translated to obtain the spectral displacement image of the acquired image, including: Based on the lateral spectral displacement distance of the acquired image, each frequency component on the spectrum image of the acquired image is horizontally shifted, and based on the longitudinal spectral displacement distance of the acquired image, each frequency component on the spectrum image of the acquired image is vertically shifted, to obtain the spectral displacement image of the acquired image.

7. The method according to claim 1, characterized in that, Based on preset defect enhancement parameters, the fused spectral image undergoes at least one round of defect enhancement processing to determine the target acquisition image of the detected target, including: The fused spectral image is spatially transformed to obtain the image to be enhanced of the detected target; Based on the defect enhancement parameters, at least one round of defect enhancement processing is performed on the fused spectral image and the image to be enhanced to determine the target acquired image.

8. The method according to claim 7, characterized in that, Based on the defect enhancement parameters, at least one round of defect enhancement processing is performed on the fused spectral image and the image to be enhanced to determine the target acquired image, including: For each round of defect enhancement processing, the image to be enhanced in that round is transformed in the frequency domain to obtain the spectral image of the image to be enhanced in that round; Determine the Euclidean norm of the spectral difference image between the spectral image of the image to be enhanced in this round and the fused spectral image, as well as the Manhattan norm of the image to be enhanced in this round. Based on the Euclidean norm, Manhattan norm, and the defect enhancement parameters of this round, determine the defect enhancement loss of this round; Based on the defect enhancement loss of this round, defect enhancement processing is performed on the image to be enhanced in this round to obtain the defect-enhanced image of this round, and the defect-enhanced image of this round is used as the image to be enhanced in the next round; The image to be enhanced in the round corresponding to the round with the minimum defect enhancement loss in the preset maximum round is taken as the target acquisition image.

9. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the image processing method as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, implement the image processing method as described in any one of claims 1-8.

Images