Method, device, equipment and medium for determining relative position of microparticles

Abstract

This disclosure provides a method, apparatus, device, and medium for determining the relative position of particles, comprising: acquiring images of point-like anomalous targets using different viewing parameters to obtain an initial image sequence; locating the same point-like anomalous target from each image in the initial image sequence and cropping it to obtain an image patch sequence; extracting a first expression feature describing the reflection and scattering structures of the point-like anomalous target for each image patch in the image patch sequence; and extracting a second expression feature distinguishing the state of particle existence; integrating the first and second expression features into a multi-view response sequence according to a preset viewing order; and determining the relative position of the point-like anomalous target based on the multi-view response sequence to obtain the relative position state of the point-like anomalous target. Thus, by integrating two types of features from multiple perspectives, point-like anomalous targets at different depths can be effectively distinguished, improving determination efficiency.

Description

Technical Field

[0001] The embodiments of this disclosure relate to the field of particulate detection technology, and more specifically, to a method, apparatus, device, and medium suitable for determining the relative position of particulates. Background Technology

[0002] In the production and inspection of transparent films, dot-like anomalies are among the most common and difficult-to-interpret anomaly types. They are small in size, have simple outlines, and limited texture, often appearing as bright spots, dark spots, or tiny blemishes in a single image, making them difficult to classify further using traditional texture features.

[0003] Point anomalies may originate from particles adsorbed on the membrane surface, impurities within the membrane, particles attached beneath the membrane, occasional imaging of airborne particles, and pseudo-points caused by high reflectivity or refraction. These objects often have similar appearances in two-dimensional images, but their risk levels and process countermeasures differ significantly. In related technologies, point anomalies are primarily located and classified manually, relying on the experience of the inspection personnel for judgment.

[0004] However, using existing technology, manual judgment is inefficient. Summary of the Invention

[0005] The embodiments described herein provide a method, apparatus, device, and medium for determining the relative position of particles, overcoming the aforementioned problems.

[0006] Firstly, based on the content of this disclosure, a method for determining the relative position of particles is provided, including: By acquiring images of point-like abnormal targets using different viewpoint parameters, an initial image sequence is obtained, arranged in a preset viewpoint order. From each image contained in the initial image sequence, the same point-like anomaly is located and cropped to obtain an image block sequence centered on the point-like anomaly and arranged in the order of the preset viewpoint. For each image patch contained in the image patch sequence, a first expression feature is extracted to describe the reflection and scattering structures of the point-like anomalous target; and a second expression feature is extracted to distinguish the state of existence of particles; the state of existence of particles includes: surface-attached state, deep-refractional state, and suspended state; According to the preset perspective order, the first expression feature and the second expression feature are integrated into a multi-perspective response sequence; The relative position of the point-like anomaly is determined based on the multi-view response sequence to obtain the relative position state of the point-like anomaly.

[0007] Secondly, according to the present disclosure, a particle relative position determination device is provided, comprising: The acquisition module is used to acquire images of point-like abnormal targets using different viewing parameters, and obtain an initial image sequence arranged in a preset viewing order; The positioning module is used to locate the same point-like abnormal target from each image contained in the initial image sequence and crop it to obtain a sequence of image blocks centered on the point-like abnormal target and arranged in the order of the preset viewpoint. The extraction module is used to extract, for each image patch contained in the image patch sequence, a first expression feature describing the reflection and scattering structure of the point-like anomalous target; and to extract a second expression feature distinguishing the state of existence of particles; the state of existence of particles includes: surface-attached state, deep-refractional state, and suspended state; An integration module is used to integrate the first expression feature and the second expression feature into a multi-view response sequence according to the preset view order; The determination module is used to determine the relative position of the point-like abnormal target based on the multi-view response sequence, and obtain the relative position state of the point-like abnormal target.

[0008] Thirdly, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the particle relative position determination method as described in any of the above embodiments.

[0009] Fourthly, a computer-readable storage medium is provided, on which a computer program is stored, and when executed by a processor, the computer program implements the steps of the particle relative position determination method as described in any of the above embodiments.

[0010] The particle relative position determination method provided in this application involves acquiring images of point-like anomalous targets using different viewing angle parameters to obtain an initial image sequence arranged in a preset viewing angle order. From each image in the initial image sequence, the same point-like anomalous target is located and cropped to obtain an image block sequence centered on the point-like anomalous target and arranged in a preset viewing angle order. For each image block in the image block sequence, a first expression feature describing the reflection and scattering structures of the point-like anomalous target is extracted; and a second expression feature distinguishing the particle's state of existence is extracted. The particle's state of existence includes: surface attachment state, deep refraction state, and suspension state. The first and second expression features are integrated into a multi-view response sequence according to the preset viewing angle order. The relative position of the point-like anomalous target is determined based on the multi-view response sequence to obtain the relative position state of the point-like anomalous target. Thus, by integrating two types of features from multiple perspectives, point-like anomalous targets at different depths can be effectively distinguished, solving the problem of accurately determining the relative position of particles in a medium in the prior art and improving determination efficiency.

[0011] The above description is merely an overview of the technical solutions of the embodiments of this application. In order to better understand the technical means of the embodiments of this application and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of this application more obvious and understandable, specific implementation methods of this application are described below. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments will be briefly described below. It should be understood that the drawings described below only relate to some embodiments of this disclosure and are not intended to limit this disclosure, wherein: Figure 1 This is a flowchart illustrating a method for determining the relative position of particles provided in this disclosure.

[0013] Figure 2 This is a schematic diagram of the structure of a particle relative position determination device provided in this disclosure.

[0014] Figure 3 This is a schematic diagram of the structure of a computer device provided in this disclosure.

[0015] It should be noted that the elements in the attached diagram are schematic and not drawn to scale. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure without creative effort are also within the scope of protection of this disclosure.

[0017] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this subject matter pertains. It will be further understood that terms such as those defined in commonly used dictionaries shall be interpreted as having the meaning consistent with their meaning in the context of the specification and in the relevant art, and shall not be interpreted in an idealized or overly formal form unless otherwise explicitly defined herein. As used herein, the statement of “connecting” or “coupling” two or more parts together shall mean that these parts are directly joined together or joined through one or more intermediate components.

[0018] The term "embodiment" as used herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of the phrase "embodiment" in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0019] In this document, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists, A and B exist simultaneously, or B exists. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship. Terms such as "first" and "second" are only used to distinguish one component (or part of a component) from another component (or another part of a component).

[0020] In the description of this application, unless otherwise stated, "multiple" means two or more (including two), and similarly, "multiple groups" means two or more (including two groups).

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0022] Figure 1 This is a schematic flowchart of a method for determining the relative position of particles provided in an embodiment of this disclosure. Figure 1 As shown, the specific process of the particle relative position determination method includes: S110. Images of point-like abnormal targets are acquired using different viewing angle parameters to obtain an initial image sequence arranged in a preset viewing angle order.

[0023] Point-like anomalies can be such as point particles, point dirt, point attachments, or other approximate point targets. Images of point-like anomalies can be acquired from multiple viewing angles using image acquisition equipment. These multiple viewing angles can be symmetrically set around the normal direction, such as a symmetrical angle sequence or discrete angle sequence acquired around the target's normal direction. Alternatively, they can be set as asymmetrical discrete angles based on constraints from the robotic arm or track.

[0024] In some embodiments, images of point-like abnormal targets are acquired using different viewing angle parameters to obtain an initial image sequence arranged in a preset viewing angle order, including: Based on different preset tilt angles of the image acquisition device relative to the transparent film, images containing the same point-like abnormal target are acquired sequentially; each viewpoint corresponds to one image; all images are arranged in order of increasing tilt angle to obtain the initial image sequence.

[0025] The image acquisition device can be an optical magnification imaging device, and different preset tilt angles are symmetrically distributed around the normal of the transparent film surface to ensure that the acquired initial image sequence can fully reflect the imaging characteristics of point-like abnormal targets under different viewpoints.

[0026] S120. Locate the same point-like anomalous target from each image in the initial image sequence and crop it to obtain an image block sequence centered on the point-like anomalous target and arranged in a preset viewpoint order.

[0027] Specifically, the same point-like anomalous target can be located in images from different perspectives, and a multi-view object-level small image sequence can be generated by cropping with the point-like anomalous target as the center, forming an image block sequence centered on the point-like anomalous target and arranged in a preset perspective order.

[0028] In some embodiments, before locating and cropping the same point-like anomalous target from each image in the initial image sequence to obtain an image block sequence centered on the point-like anomalous target and arranged in a preset viewpoint order, the method further includes: performing a filtering operation on each image in the initial image sequence to filter out noise interference in the image.

[0029] The filtering operation can employ Gaussian filtering or median filtering, selecting an appropriate filtering method for different types of noise. While filtering out noise, it preserves the edge information of point-like abnormal targets to the greatest extent possible, preventing target features from being smoothed out and ensuring the accuracy of subsequent positioning and cropping.

[0030] In some embodiments, before locating and cropping the same point-like anomalous target from each image in the initial image sequence to obtain an image block sequence centered on the point-like anomalous target and arranged in a preset viewpoint order, or after performing filtering operations on each image in the initial image sequence, the method further includes: amplifying the pixel difference between the point-like anomalous target and the background in each image using a contrast enhancement algorithm to enhance the feature recognition of the point-like anomalous target in the image.

[0031] Among them, contrast enhancement algorithms, such as histogram equalization or gamma correction, can adjust the grayscale distribution of the image for imaging scenes under different lighting conditions, avoiding deviations in subsequent localization due to the small grayscale difference between point-like abnormal targets and the background, thereby effectively improving the detection accuracy of point-like abnormal targets.

[0032] S130. For each image patch contained in the image patch sequence, extract a first expression feature for describing the reflection and scattering structures of point-like anomalous targets; and extract a second expression feature for distinguishing the state of existence of particles.

[0033] The states in which particles exist include: surface-attached state, deep-refractive state, and suspended state.

[0034] It can extract the highlight center position, peak brightness, halo radius, and annular energy distribution for image patches from each viewpoint to describe the reflection and scattering structure of the target. It can also extract features such as edge (left / right / top / bottom) scattering symmetry index, tail direction and length, and edge gradient direction to distinguish surface attachments, deep refraction, and suspended particles.

[0035] In some embodiments, for each image patch contained in the image patch sequence, first representational features for describing the reflection and scattering structures of point-like anomalous targets are extracted, including: Obtain the highlight center coordinates of the point-like anomalous target in each image patch within the image patch sequence; calculate the difference in highlight center coordinates of the point-like anomalous target at different viewpoints based on the highlight center coordinates of the point-like anomalous target corresponding to different image patches; determine the highlight center drift of the point-like anomalous target based on the highlight center coordinate difference; calculate the change in halo radius of the point-like anomalous target at different viewpoints based on the halo radius of the point-like anomalous target corresponding to different image patches; determine the halo spread rate of the point-like anomalous target based on the change in halo radius; and determine the first expression feature used to describe the reflection and scattering structures of the point-like anomalous target based on the highlight center drift and halo spread rate.

[0036] Among them, the highlight center drift can distinguish between surface particles and deep particles, and can be used to describe the distribution location of the target's reflective structure. The halo spread rate can characterize the degree of light diffusion scattered by particles at different depths; for deeper particles, the halo radius changes more significantly with the shooting angle, and the corresponding halo spread rate is higher.

[0037] The amount of highlight center shift under different viewing angles can be expressed as shown in the following formula (1).

[0038] Δ H ij =|| C i - C j ||(1) In formula (1), Δ H ij This indicates the amount of highlight center shift under different viewing angles; C i Indicates the first i The location of the highlight center from various angles; C j Indicates the first j The location of the highlight center from various perspectives; the trajectory of the highlight center as the angle changes can reflect the relative surface position of the point target.

[0039] The halo spread rate under different viewing angles can be expressed as shown in the following formula (2).

[0040] G ( θ )=( R θ - R 0) / R 0(2) In formula (2), G ( θ ) represents the halo expansion rate function, which reflects the relative degree of annular energy expansion at different angles; R θ Indicates angle θ The halo radius can be used to analyze the range and spread rate of annular scattering.

[0041] Therefore, by combining the high-light center drift and the halo spread rate, the first expression feature is obtained, which can accurately characterize the reflection and scattering properties of point-like anomalous particles at different relative positions.

[0042] In some embodiments, for each image patch contained in the image patch sequence, a second expression feature for distinguishing the state of particle presence is extracted, including: For each image patch contained in the image patch sequence, the scattering symmetry data of point-like anomalous targets in different directions are extracted; and the shadow tail direction of the point-like anomalous targets is extracted; based on the scattering symmetry data of point-like anomalous targets in different directions and the shadow tail direction, a second expression feature for distinguishing the existence state of particles is determined.

[0043] Among them, the scattering symmetry data are the statistical results of the scattering intensity distribution of point-like anomalous targets in multiple radial directions, which can be used to characterize whether the energy distribution around the target is shifted along a certain direction. The shadow tail direction is determined by the gradual change trend of the gray level of the point-like anomalous target along various directions, which can be used to describe the offset state of the particles relative to the film surface.

[0044] Therefore, by combining scattering symmetry data with shadow tail direction to obtain the second expression feature, it is possible to effectively distinguish the different existence states of particles, such as those located on the membrane surface, embedded in the membrane, or existing on the other side of the membrane.

[0045] S140. According to the preset perspective order, integrate the first expression feature and the second expression feature into a multi-perspective response sequence.

[0046] The multi-view response sequence may include: specular drift sequence, halo change sequence, scattering symmetry sequence, and shadow tail sequence.

[0047] S150. Determine the relative position of the point-like abnormal target based on the multi-view response sequence to obtain the relative position status of the point-like abnormal target.

[0048] The relative positional states are as follows: membrane surface state, membrane interior state, membrane subsurface state, and suspended state.

[0049] In some embodiments, the relative position of a point-like anomaly is determined based on a multi-view response sequence to obtain the relative position state of the point-like anomaly, including: By inputting the multi-view response sequence into the location determination model, the predicted probability of the point-like anomaly target relative to different location states is obtained; the location state corresponding to the maximum predicted probability is determined as the relative location state of the point-like anomaly target.

[0050] The location determination model can be a rule-based model, machine learning model, deep learning model, or agent inference model. The output is the probability distribution of whether a point-like anomalous target is located on the membrane surface, inside the membrane, below the membrane, or in a suspended state. The primary location label and confidence level can be output based on the highest probability. If the point-like anomalous target is temporally unstable or if different models output conflict, manual verification or supplementary data collection can be triggered.

[0051] For example, for surface particles, the system observed a significant shift in the highlight center with changing viewing angle, while the halo radius changed relatively little, ultimately resulting in the highest probability of outputting surface particles. For particles attached beneath the film, the system observed a smaller highlight drift but significant halo expansion, and a more stable tail direction, ultimately resulting in a higher probability of outputting particles attached beneath the film. For suspended particles or transient noise, the system found that the same target exhibited unstable positions, large brightness fluctuations, and inconsistent object outlines between adjacent frames, thus resulting in a higher probability of outputting suspended states and recommending re-examination.

[0052] In addition, this embodiment can also combine time series stability index to distinguish between real attached particles and transient suspended particles or pseudo-points, thereby avoiding misjudging transient noise as real attached particles.

[0053] In this embodiment, images of point-like anomalous targets are acquired using different viewing angle parameters to obtain an initial image sequence arranged in a preset viewing angle order. From each image in the initial image sequence, the same point-like anomalous target is located and cropped to obtain an image block sequence centered on the point-like anomalous target and arranged in a preset viewing angle order. For each image block in the image block sequence, a first expression feature describing the reflection and scattering structures of the point-like anomalous target is extracted; and a second expression feature distinguishing the particle's state of existence is extracted. The particle's state of existence includes: surface attachment state, deep refraction state, and suspension state. The first and second expression features are integrated into a multi-view response sequence according to the preset viewing angle order. The relative position of the point-like anomalous target is determined based on the multi-view response sequence to obtain the relative position state of the point-like anomalous target. Thus, by integrating two types of features from multiple perspectives, point-like anomalous targets at different depths can be effectively distinguished, solving the problem of accurately determining the relative position of particles in a medium in existing technologies and improving determination efficiency.

[0054] In summary, this embodiment specifically addresses the problem of determining the location of point anomalies in transparent membrane scenarios, demonstrating strong relevance and engineering application value. The method described in this embodiment can effectively distinguish between surface particles, subsurface deposits, suspended particles, and pseudo-points; it enhances the interpretability and targeted handling of point anomalies in transparent membrane detection; and it provides more direct evidence for cleaning strategies, re-inspection strategies, and determination of responsible interfaces.

[0055] Figure 2 This is a schematic diagram of a particle relative position determination device provided in this embodiment. The particle relative position determination device may include: The acquisition module 210 is used to acquire images of point-like abnormal targets through different viewing parameters, and obtain an initial image sequence arranged in a preset viewing order.

[0056] The positioning module 220 is used to locate the same point-like abnormal target from each image contained in the initial image sequence and crop it to obtain a sequence of image blocks centered on the point-like abnormal target and arranged in a preset viewpoint order.

[0057] Extraction module 230 is used to extract, for each image patch contained in the image patch sequence, a first expression feature describing the reflection and scattering structure of the point-like anomalous target; and to extract a second expression feature distinguishing the state of existence of particles; the state of existence of particles includes: surface-attached state, deep-refractional state, and suspended state.

[0058] The integration module 240 is used to integrate the first expression feature and the second expression feature into a multi-view response sequence according to a preset viewpoint order.

[0059] The determination module 250 is used to determine the relative position of point-like abnormal targets based on the multi-view response sequence, and obtain the relative position status of the point-like abnormal targets.

[0060] In some embodiments, the extraction module 230 is specifically used for: Obtain the highlight center coordinates of the point-like anomalous target in each image patch within the image patch sequence; calculate the difference in highlight center coordinates of the point-like anomalous target at different viewpoints based on the highlight center coordinates of the point-like anomalous target corresponding to different image patches; determine the highlight center drift of the point-like anomalous target based on the highlight center coordinate difference; calculate the change in halo radius of the point-like anomalous target at different viewpoints based on the halo radius of the point-like anomalous target corresponding to different image patches; determine the halo spread rate of the point-like anomalous target based on the change in halo radius; and determine the first expression feature used to describe the reflection and scattering structures of the point-like anomalous target based on the highlight center drift and halo spread rate.

[0061] In some embodiments, the extraction module 230 is specifically used for: For each image patch contained in the image patch sequence, the scattering symmetry data of point-like anomalous targets in different directions are extracted; and the shadow tail direction of the point-like anomalous targets is extracted; based on the scattering symmetry data of point-like anomalous targets in different directions and the shadow tail direction, a second expression feature for distinguishing the existence state of particles is determined.

[0062] In some embodiments, the determination module 250 is specifically used for: By inputting the multi-view response sequence into the location determination model, the predicted probability of the point-like anomaly target relative to different location states is obtained; the location state corresponding to the maximum predicted probability is determined as the relative location state of the point-like anomaly target; the relative location state is one of the following: membrane surface state, membrane interior state, membrane subsurface state, and suspension state.

[0063] In some embodiments, the acquisition module 210 is specifically used for: Based on different preset tilt angles of the image acquisition device relative to the transparent film, images containing the same point-like abnormal target are acquired sequentially; each viewpoint corresponds to one image; all images are arranged in order of increasing tilt angle to obtain the initial image sequence.

[0064] In some embodiments, a filtering module is also included.

[0065] The filtering module is used to perform filtering operations on each image in the initial image sequence to remove noise interference from the images.

[0066] In some embodiments, an enhancement module is also included.

[0067] The enhancement module is used to amplify the pixel differences between point-like anomalous targets and the background in each image through a contrast enhancement algorithm, so as to enhance the feature recognition of point-like anomalous targets in the image.

[0068] The particle relative position determination device provided in this disclosure can execute the above method embodiments. Its specific implementation principle and technical effects can be found in the above method embodiments, and will not be repeated here.

[0069] This application also provides a computer device. Please refer to the following for details. Figure 3 , Figure 3 This is a basic structural block diagram of the computer device in this embodiment.

[0070] The computer device includes a memory 310 and a processor 320 that are interconnected via a system bus. It should be noted that only a computer device with memory 310 and processor 320 is shown in the figure; however, it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented alternatively. Those skilled in the art will understand that the computer device described herein is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0071] Computer devices can include desktop computers, laptops, handheld computers, and cloud servers. These devices allow for human-computer interaction with users through keyboards, mice, remote controls, touchpads, or voice-activated devices.

[0072] The memory 310 includes at least one type of readable storage medium, including non-volatile memory or volatile memory, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. RAM may include static RAM or dynamic RAM. In some embodiments, the memory 310 may be an internal storage unit of a computer device, such as the hard disk or memory of the computer device. In other embodiments, the memory 310 may also be an external storage device of the computer device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, or flash card equipped on the computer device. Of course, the memory 310 may include both internal storage units and external storage devices of the computer device. In this embodiment, the memory 310 is typically used to store the operating system and various application software installed on the computer device, such as the program code of the method described above. In addition, the memory 310 can also be used to temporarily store various types of data that have been output or will be output.

[0073] Processor 320 is typically used to perform overall operations of a computer device. In this embodiment, memory 310 is used to store program code or instructions, including computer operation instructions, and processor 320 is used to execute the program code or instructions stored in memory 310 or process data, such as program code that runs the methods described above.

[0074] In this article, the bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus system 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.

[0075] Another embodiment of this application also provides a computer-readable medium, which may be a computer-readable signal medium or a computer-readable medium. A processor in a computer reads computer-readable program code stored in the computer-readable medium, enabling the processor to execute the functional actions specified in each step or combination of steps in the above method; and to generate means for implementing the functional actions specified in each block or combination of blocks in the block diagram.

[0076] Computer-readable media include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared memory or semiconductor systems, devices or apparatuses, or any suitable combination thereof, wherein the memory is used to store program code or instructions, the program code including computer operation instructions, and the processor is used to execute the program code or instructions of the above-described methods stored in the memory.

[0077] The definitions of memory and processor can be found in the description of the foregoing computer device embodiments, and will not be repeated here.

[0078] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0079] In the various embodiments of this application, the functional units or modules can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0080] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0081] In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" as described in this application does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. This application can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims listing several means, several units of these means may be embodied by the same item of hardware. The use of "first," "second," and "third," etc., does not indicate any order and these words should be interpreted as names. Unless otherwise specified, the steps in the above embodiments should not be construed as limiting the order of execution.

[0082] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for determining the relative position of particles, characterized in that, include: By acquiring images of point-like abnormal targets using different viewpoint parameters, an initial image sequence is obtained, arranged in a preset viewpoint order. From each image contained in the initial image sequence, the same point-like anomaly is located and cropped to obtain an image block sequence centered on the point-like anomaly and arranged in the order of the preset viewpoint. For each image patch contained in the image patch sequence, a first expression feature is extracted to describe the reflection and scattering structures of the point-like anomalous target; and a second expression feature is extracted to distinguish the state of existence of particles. The states in which the particles exist include: surface-attached state, deep-reflected state, and suspended state; According to the preset perspective order, the first expression feature and the second expression feature are integrated into a multi-perspective response sequence; The relative position of the point-like anomaly is determined based on the multi-view response sequence to obtain the relative position state of the point-like anomaly.

2. The method according to claim 1, characterized in that, For each image patch contained in the image patch sequence, extract first representation features for describing the reflection and scattering structures of the point-like anomalous target, including: Obtain the highlight center coordinates of the point-like anomalous target in each image block contained in the image block sequence; calculate the highlight center coordinate difference of the point-like anomalous target under different viewpoints based on the highlight center coordinates of the point-like anomalous target corresponding to different image blocks; determine the highlight center drift of the point-like anomalous target based on the highlight center coordinate difference; Based on the halo radius of the point-like anomaly target corresponding to different image patches, calculate the change in halo radius of the point-like anomaly target under different viewpoints; determine the halo spread rate of the point-like anomaly target based on the change in halo radius; Based on the specular center drift and halo spread of the point-like anomalous target, a first expression feature for describing the reflection and scattering structure of the point-like anomalous target is determined.

3. The method according to claim 1, characterized in that, For each image patch contained in the image patch sequence, a second expression feature for distinguishing the presence state of particles is extracted, including: For each image patch contained in the image patch sequence, the scattering symmetry data of the point-like anomalous target in different directions are extracted; and the shadow trailing direction of the point-like anomalous target is extracted. Based on the scattering symmetry data and shadow tail direction of the point-like anomalous targets in different directions, a second expression feature for distinguishing the state of particle presence is determined.

4. The method according to claim 1, characterized in that, The relative position of the point-like anomaly is determined based on the multi-view response sequence to obtain the relative position state of the point-like anomaly, including: The multi-view response sequence is input into the location determination model to obtain the predicted probability of the point-like anomaly target relative to different location states. The position state corresponding to the maximum predicted probability is determined as the relative position state of the point-like anomaly target; the relative position state is one of the following: membrane surface state, membrane interior state, membrane subsurface state, and suspension state.

5. The method according to claim 1, characterized in that, Images of point-like anomalies are acquired using different viewpoint parameters to obtain an initial image sequence arranged in a preset viewpoint order, including: Based on different preset tilt angles of the image acquisition device relative to the transparent film, images containing the same point-like abnormal target are acquired sequentially; each viewpoint corresponds to one image. Arrange all images in ascending order of tilt angle to obtain the initial image sequence.

6. The method according to claim 1, characterized in that, Before locating and cropping the same point-like anomaly target from each image in the initial image sequence to obtain an image block sequence centered on the point-like anomaly target and arranged according to the preset viewpoint order, the process further includes: Each image in the initial image sequence is filtered to remove noise interference.

7. The method according to claim 6, characterized in that, Also includes: The pixel differences between the point-like anomalous targets and the background in each image are amplified by a contrast enhancement algorithm to enhance the feature recognition of the point-like anomalous targets in the image.

8. A device for determining the relative position of particles, characterized in that, include: The acquisition module is used to acquire images of point-like abnormal targets using different viewing parameters, and obtain an initial image sequence arranged in a preset viewing order; The positioning module is used to locate the same point-like abnormal target from each image contained in the initial image sequence and crop it to obtain a sequence of image blocks centered on the point-like abnormal target and arranged in the order of the preset viewpoint. The extraction module is used to extract, for each image patch contained in the image patch sequence, a first expression feature describing the reflection and scattering structure of the point-like anomalous target; and to extract a second expression feature distinguishing the state of existence of particles. The states in which the particles exist include: surface-attached state, deep-reflected state, and suspended state; An integration module is used to integrate the first expression feature and the second expression feature into a multi-view response sequence according to the preset view order; The determination module is used to determine the relative position of the point-like abnormal target based on the multi-view response sequence, and obtain the relative position state of the point-like abnormal target.

9. A computer device, characterized in that, It includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the particle relative position determination method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the particle relative position determination method as described in any one of claims 1 to 7.

Images