Extensible standardized micro optical array module multi-scene adaptation method

Abstract

The invention discloses an extensible standardized micro optical array module multi-scene adaptation method, and particularly relates to the technical field of array module multi-scene adaptation, and the method comprises the steps: carrying out the periodic structure calibration of a micro optical module, and precisely obtaining the space imaging deviation and time sampling delay information between the modules; and obtaining a microscopic parallax deviation accumulation coefficient and a signal synchronization delay out-of-control coefficient, constructing a dynamic dislocation fusion obstacle evaluation model based on the coefficients, and outputting a dynamic dislocation fusion failure obstacle index, thereby realizing quantitative evaluation and failure grade division of the collaborative stability of the multi-module system. A reference basis with extremely high operability is provided for subsequent control and scheduling; furthermore, by recording obstacle trend sequences under different operation situations, clustering and attribution modeling are carried out in combination with environment state vectors, a scene label-obstacle trend mode association library is established, and a scheduling strategy mapping mechanism is introduced, so that cross-scene rapid adaptation and stable operation are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of array module multi-scene adaptation, and more specifically, to a scalable standardized micro-optical array module multi-scene adaptation method. Background Art

[0002] With the increasing demand for multimodal fusion perception and high-precision visual imaging, scalable and standardized micro-optical array modules have become a key component of multi-scene intelligent perception systems. This type of module usually achieves flexible adaptation to different fields of view, focal length levels, and perception resolutions through an array combination of multiple micro-optical imaging units. However, the currently widely adopted modular splicing and parallel acquisition strategies, while improving the system's reconfigurability and adaptability, also expose a deep-seated coordination failure problem caused by spatial splicing errors and temporal inconsistencies.

[0003] Specifically, in multi-module collaborative motion capture tasks, the coupled technical challenges of microscopic parallax offset accumulation and uncontrolled module signal synchronization delays are common. The former primarily stems from micron-level deviations in optical axis alignment between modules or thermally induced deformation that cause imaging angle offsets, while the latter stems from clock drift within the module's internal processor or acquisition interface, leading to inter-frame acquisition inconsistencies. This "parallax spatial misalignment" and "temporal sampling misordering" can create dynamic misalignment fusion obstacles in the presence of high-speed moving targets or complex light fields, causing the system to experience severe nonlinear artifacts such as "edge splitting," "inter-frame jitter," and "ghost image misfusion" during image stitching. Furthermore, this type of distortion can easily induce misjudgments in subsequent algorithmic modules such as automatic exposure control, image enhancement, and target recognition, leading to deeper system degradation, such as "false triggering of the adaptive parameter adjustment system" or "failure to continuously track the target." Therefore, this type of multi-module fusion failure problem caused by spatial-temporal misalignment coupling distortion has become one of the core technical bottlenecks of current micro-optical array systems when facing dynamic and complex scenes (such as high-speed movement, low light interference, structural occlusion, etc.). There is an urgent need to build a more robust optical array module multi-scene adaptation method. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a scalable standardized micro-optical array module multi-scene adaptation method to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: A scalable standardized micro-optical array module multi-scene adaptation method includes the following steps: Step S1, performing periodic structural calibration on each micro-optical module in the system, and obtaining a spatial imaging deviation distribution map and a time sampling delay difference matrix between the modules; Step S2, obtaining microscopic parallax offset accumulation information between multiple modules according to the spatial imaging deviation distribution map between modules, wherein the microscopic parallax offset accumulation information includes a microscopic parallax offset accumulation coefficient; Step S3, obtaining signal synchronization delay out-of-control information among multiple modules according to the time sampling delay difference matrix among the modules, wherein the signal synchronization delay out-of-control information includes a signal synchronization delay out-of-control coefficient; Step S4: constructing a dynamic dislocation fusion obstacle assessment model based on the micro-parallax offset cumulative coefficient and the signal synchronization delay out-of-control coefficient, outputting a dynamic dislocation fusion failure obstacle index, and determining the fusion failure obstacle level of the multi-module group; Step S5: During long-term operation, record the obstacle trend sequences under different dynamic scenarios to build a scene label-obstacle trend pattern association library, map the scene label-obstacle trend pattern association library with the scheduling strategy, and quickly adapt across scenarios.

[0006] In a preferred embodiment, the spatial imaging deviation distribution map is obtained: Use multiple modules to collaboratively shoot static and synchronous images of the standard pattern; each module shoots the same target standard pattern; Extract feature point set for each module image: ,in is the feature point set of the i-th module, is the nth feature point; Randomly select one of the modules as the reference module, align the other modules with the reference module's viewing angle, and calculate the point deviation: ,in is the point deviation, To map the reference module perspective to the transformation of the i-th module, is the feature point set of the reference module; All modules will be The point deviations between them are interpolated and visualized on the spatial coordinates to obtain the spatial imaging deviation distribution map .

[0007] In a preferred embodiment, the time sampling delay difference matrix is obtained: Use a unified synchronous trigger signal source to trigger all modules to collect images. Each module records the time delay between receiving the trigger signal and the actual generation of image data. ; For each module pair Calculate the frame response delay difference: ,in is the frame response delay difference between the i-th module and the j-th module, is the time delay between the i-th module receiving the trigger signal and the actual generation of image data, is the time delay between the i-th module receiving the trigger signal and the actual generation of image data; Based on all modules Constructing the time sampling delay difference matrix : ,in , The total number of modules.

[0008] In a preferred embodiment, the logic for obtaining the micro-parallax offset accumulation coefficient is as follows: According to the spatial imaging deviation distribution map Compute the residual tensor: ,in is the residual tensor, for The global mean of ; using pixel blocks as nodes and residual change gradients as edge weights, a deviation propagation graph is constructed: ,in For nodes and nodes The residual gradient edge weight, For nodes The residual tensor of For nodes The residual tensor of ; The graph structure of the deviation propagation graph is: ,in is the set of pixel block nodes in the deviation propagation graph, is the node-edge set in the deviation propagation graph; According to the module structure state, the optical axis disturbance response function is defined as: ,in is the optical axis disturbance response value, is the angle between the optical axes of the i-th module and the j-th module, is the temperature gradient difference between the i-th module and the j-th module, is the peak value difference of vibration response between the i-th module and the j-th module, is the preset proportional coefficient of the optical axis angle, temperature gradient difference, and vibration response peak difference, and All greater than 0; The optical axis disturbance response value Normalized to the perturbation factor: ,in is the disturbance factor, is a maximum value acquisition function used to obtain the maximum value of the optical axis disturbance response value; In the deviation propagation diagram The shortest path search algorithm is executed on the MATLAB to calculate the spatial path offset potential energy: ,in is the space path offset potential energy, is the shortest path in the deviation propagation graph, is the residual gradient edge weight, is the number of path hops; Calculate the microscopic parallax offset cumulative coefficient: ,in is the microscopic parallax offset accumulation coefficient, is a very small constant to prevent division by zero, , The total number of modules.

[0009] In a preferred embodiment, the logic for obtaining the signal synchronization delay out-of-control coefficient is as follows: In the time scale set Extract the disturbance energy of the time sampling delay difference matrix at different time scales: ,in is the disturbance energy at the g-th time scale, is the module running time, t is the time unit, is the time index within the g-th time scale range, is the frame response delay difference between the i-th module and the j-th module, is the mean of the frame response delay difference; construct the multi-scale perturbation energy vector: ; Normalized perturbation energy: ,in is the normalized perturbation energy; calculate the information entropy of the perturbation distribution: ,in is the information entropy of the disturbance distribution; calculate the local out-of-control energy weighting coefficient: ,in is the local out-of-control energy weighting coefficient, is a minimum constant to prevent division by zero; taking each module as a node, if the frame response delay difference is greater than the preset frame response delay difference threshold, it is considered that there is a delay imbalance edge, and a delay edge weight is constructed between the module nodes. : ; Get the synchronous delay disturbance propagation diagram ,in is the set of module nodes of the synchronous delay disturbance propagation graph, is the module node edge set of the synchronous delay disturbance propagation graph; calculates the average out-of-control path length of the propagation chain of the synchronous delay disturbance propagation graph : ,in is the set of shortest paths between all module node pairs, is the number of shortest paths between all module node pairs, is the shortest path; calculate the signal synchronization delay out-of-control coefficient: ,in is the signal synchronization delay out-of-control coefficient, is the information entropy coefficient of the disturbance distribution, , are the information entropy coefficient of the disturbance distribution and the preset proportional coefficient of the average out-of-control path length of the propagation chain, respectively, and Both are greater than 0.

[0010] In a preferred embodiment, a dynamic dislocation fusion obstacle assessment model is constructed based on the microscopic parallax offset cumulative coefficient and the signal synchronization delay out-of-control coefficient, and a dynamic dislocation fusion failure obstacle index is output. The dynamic dislocation fusion obstacle assessment model is based on the following formula: , where is the dynamic dislocation fusion failure barrier index, is the microscopic parallax offset accumulation coefficient, is the signal synchronization delay out-of-control coefficient, They represent the preset proportional coefficients of the microscopic parallax offset accumulation coefficient and the signal synchronization delay out-of-control coefficient, respectively, and Both are greater than 0.

[0011] In a preferred embodiment, the dynamic dislocation fusion failure barrier index is compared with a preset dynamic dislocation fusion failure barrier index threshold to determine the fusion failure barrier level of the multi-module, as follows: If the dynamic dislocation fusion failure barrier index is greater than the dynamic dislocation fusion failure barrier index threshold, the fusion failure barrier level is assessed as a high risk level; If the dynamic dislocation fusion failure obstacle index is less than or equal to the dynamic dislocation fusion failure obstacle index threshold, the fusion failure obstacle level is assessed as normal.

[0012] The technical effects and advantages of the present invention are as follows: 1. The present invention accurately obtains the spatial imaging deviation and time sampling delay information between modules by periodically calibrating the structure of the micro-optical module, establishes a high-resolution, high-precision space-time state perception foundation from the source, and improves the parameter stability and data consistency of the module collaborative operation; obtains the micro-parallax offset accumulation coefficient, and comprehensively quantifies the error accumulation trend caused by factors such as optical axis offset and thermal deformation at the spatial imaging level; at the same time, obtains the signal synchronization delay out-of-control coefficient, effectively revealing the inter-frame acquisition inconsistency problem caused by factors such as clock drift and acquisition link difference in the multi-module system, and constructs a dynamic dislocation fusion obstacle assessment model based on the above coefficients to output a dynamic dislocation fusion failure obstacle index, thereby realizing the quantitative assessment of the collaborative stability of the multi-module system and the failure level classification, providing a highly operational reference for subsequent control and scheduling. Furthermore, by recording the obstacle trend sequences under different operating scenarios and combining them with the environmental state vector for clustering and attribution modeling, a scene label-obstacle trend pattern association library is established, and a scheduling strategy mapping mechanism is introduced to ensure that the system can quickly identify scene patterns and automatically match the optimal control strategy when facing typical complex working conditions such as high-speed movement, low light interference, and structural occlusion, thereby achieving rapid adaptation and stable operation across scenarios. Not only does it achieve accurate modeling and quantitative evaluation of the spatial imaging error and time sampling out-of-control of the micro-optical array module at the perception level, but it also establishes a control strategy mapping mechanism associated with the evolution of obstacle trends at the decision-making level, significantly improving the system's fusion robustness, adaptability, and multi-module collaborative stability in complex scenarios, breaking through the technical bottleneck of the existing technology that cannot effectively deal with the problem of space-time coupling failure. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings; Figure 1 Flowchart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0014] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0015] Example: Figure 1 The present invention provides a scalable standardized micro-optical array module multi-scene adaptation method, which includes the following steps: Step S1, performing periodic structural calibration on each micro-optical module in the system, and obtaining a spatial imaging deviation distribution map and a time sampling delay difference matrix between the modules; Step S2, obtaining microscopic parallax offset accumulation information between multiple modules according to the spatial imaging deviation distribution map between modules, wherein the microscopic parallax offset accumulation information includes a microscopic parallax offset accumulation coefficient; Step S3, obtaining signal synchronization delay out-of-control information among multiple modules according to the time sampling delay difference matrix among the modules, wherein the signal synchronization delay out-of-control information includes a signal synchronization delay out-of-control coefficient; Step S4: constructing a dynamic dislocation fusion obstacle assessment model based on the micro-parallax offset cumulative coefficient and the signal synchronization delay out-of-control coefficient, outputting a dynamic dislocation fusion failure obstacle index, and determining the fusion failure obstacle level of the multi-module group; Step S5: During long-term operation, record the obstacle trend sequences under different dynamic scenarios to build a scene label-obstacle trend pattern association library, map the scene label-obstacle trend pattern association library with the scheduling strategy, and quickly adapt across scenarios; In step S1, a periodic structural calibration is performed on each micro-optical module in the system, and a spatial imaging deviation distribution map and a temporal sampling delay difference matrix between the modules are obtained, as follows: Acquisition of spatial imaging deviation distribution map: Use multiple modules to collaboratively capture static and synchronous images of standard patterns (such as a checkerboard or laser dot matrix). Each module captures the same target standard pattern to ensure consistent imaging time and a stable, non-deformed pattern. Extract feature point set for each module image: ,in is the feature point set of the i-th module, is the nth feature point; Randomly select one of the modules as the reference module, align the other modules with the reference module's viewing angle, and calculate the point deviation: ,in is the point deviation, To map the reference module perspective to the transformation of the i-th module, is the feature point set of the reference module; All modules will be The point deviations between them are interpolated and visualized on the spatial coordinates to obtain the spatial imaging deviation distribution map ; Obtaining the time sampling delay difference matrix: Use a unified synchronous trigger signal source (such as GPIO pulse or synchronization protocol) to trigger all modules to acquire images. Each module records the time delay between receiving the trigger signal and the actual generation of image data. ; For each module pair Calculate the frame response delay difference: ,in is the frame response delay difference between the i-th module and the j-th module, is the time delay between the i-th module receiving the trigger signal and the actual generation of image data, is the time delay between the i-th module receiving the trigger signal and the actual generation of image data; Based on all modules Constructing the time sampling delay difference matrix : ,in , is the total number of modules; Step S2, obtaining microscopic parallax offset accumulation information between multiple modules according to the spatial imaging deviation distribution map between modules, wherein the microscopic parallax offset accumulation information includes a microscopic parallax offset accumulation coefficient; The micro-parallax offset accumulation coefficient (MCI) in this invention is an important indicator used to measure the intensity and degree of coupling between FOV consistency shifts across scalable, standardized micro-optical array modules. This coefficient quantitatively describes the cumulative offset effect of disparity vectors in the image domain caused by factors such as spatial structural perturbations, module thermal drift, and splicing errors during long-term operation and multi-scene adaptation. This coefficient provides a theoretical basis and data support for subsequent identification of module fusion obstacles. Specifically, the MCI extracts offset vectors corresponding to feature points between module images in multiple dynamically acquired frames, derives FOV reconstruction errors based on an optical axis geometry model, and combines the inter-frame variation trend of the registration residual to form a comprehensive quantitative indicator representing the cumulative drift intensity and instability of spatial parallax offset in the image domain. If systematic FOV offset exists between multiple modules (e.g., due to insufficient assembly precision, micro-vibration of module mounts, or optical axis torsion caused by thermal expansion), parallax cannot be offset during multi-frame reconstruction, resulting in a continuous expansion of boundary misalignment, manifested as a high MCI. A large value for this coefficient indicates a significant and non-negligible tendency for cumulative misalignment between modules during multi-frame image fusion. Even if the initial registration appears reasonable in a static state, the accumulated spatial deviations can lead to fusion problems such as stitching artifacts, edge tearing, and center region alignment failure during dynamic acquisition. The system is therefore highly susceptible to dynamic environmental disturbances, such as high-speed moving objects, vibrating platforms, and angular variations, significantly reducing the robustness and reconfigurability of module fusion. Conversely, a small cumulative micro-parallax offset coefficient indicates good optical axis consistency between modules, stable parallax matching, and no significant error diffusion across different frame sequences or dynamic scenes. This makes the system more likely to achieve high-quality image stitching and field of view fusion, making it particularly suitable for high-reliability vision tasks such as complex multi-target tracking, low-texture region recognition, and real-time 3D reconstruction. Dynamic monitoring of this coefficient provides early warning of module structural health. A gradual increase in the coefficient over a period of time indicates potential structural view drift, facilitating timely structural calibration or redundant switching, thereby improving system operational continuity. More importantly, a dynamic misalignment fusion barrier assessment model can be further constructed based on the microscopic parallax offset accumulation coefficient. Coupled with the signal synchronization delay loss-of-control coefficient, this model forms a diagnostic mechanism for "spatial-temporal coordination failure." This mechanism not only determines whether the current fusion state presents a high risk, but also assists multi-module systems in autonomously switching fusion strategies (such as scaling compensation, elastic parallax reconstruction, and multi-module fusion weight adjustment) in different scenarios. This improves the adaptability and robust fusion level of optical array modules in multiple scenarios and multi-task conditions, possessing significant engineering value and potential for supporting intelligent perception.

[0016] The logic for obtaining the micro-parallax offset accumulation coefficient is as follows: According to the spatial imaging deviation distribution map Compute the residual tensor: ,in is the residual tensor, which is used to characterize the offset strength of the imaging bias to the global trend. for The global mean of ; using pixel blocks as nodes and residual change gradients as edge weights, a deviation propagation graph is constructed: ,in For nodes and nodes The residual gradient edge weight, For nodes The residual tensor of For nodes The residual tensor of ; The graph structure of the deviation propagation graph is: ,in is the set of pixel block nodes in the deviation propagation graph, is the node-edge set in the deviation propagation graph; According to the module structure state (temperature, posture, vibration), the optical axis disturbance response function is defined as: ,in is the optical axis disturbance response value, is the angle between the optical axes of the i-th module and the j-th module, is the temperature gradient difference between the i-th module and the j-th module, is the peak value difference of vibration response between the i-th module and the j-th module, is the preset proportional coefficient of the optical axis angle, temperature gradient difference, and vibration response peak difference, and All greater than 0; It should be noted that Set it according to the actual situation. For example, adopt the expert empowerment method, that is, invite experts in related fields to determine the preset proportion coefficients of various indicators through professional opinion surveys and comprehensive evaluations, for example, It can be 0.3, 0.4, 0.3; The optical axis disturbance response value Normalized to the perturbation factor: ,in is the disturbance factor, is the maximum value acquisition function, used to obtain the maximum value of the optical axis disturbance response value. is a minimum constant to prevent division by zero (usually ); In the deviation propagation diagram The shortest path search algorithm (such as Dijkstra algorithm, Floyd-Warshall algorithm) is executed to calculate the spatial path offset potential energy: ,in is the space path offset potential energy, is the shortest path in the deviation propagation graph, is the residual gradient edge weight, is the number of path hops; Calculate the microscopic parallax offset cumulative coefficient: ,in is the microscopic parallax offset accumulation coefficient, is a minimum constant to prevent division by zero (usually ), , is the total number of modules; It should be noted that the above formulas are all dimensionless and numerical calculations. Common dimensionless methods include Min-Max normalization and Z-Score normalization, which will not be described here. Step S3, obtaining signal synchronization delay out-of-control information among multiple modules according to the time sampling delay difference matrix among the modules, wherein the signal synchronization delay out-of-control information includes a signal synchronization delay out-of-control coefficient; The signal synchronization delay out-of-control coefficient in the present invention is an important indicator for measuring the degree of degradation of timing consistency in a multi-module system. Its essence is to quantify the evolution trend and out-of-control probability of the time synchronization offset generated in the image sampling, data processing or information feedback processes between each optical module in response to the same event trigger or signal wake-up condition. This coefficient not only takes into account the initial time difference between modules, but also integrates the cumulative timing drift phenomenon in continuous sampling cycles, so as to effectively reflect the stability of the signal transmission coordination of multiple modules in a dynamic operating environment. A large signal synchronization delay out-of-control coefficient usually indicates that there is a significant signal sampling or response chain timing drift problem in the system, which means that some modules are unable to complete the sampling and reporting of image signals within the preset synchronization window due to internal clock offset, processing delay, temperature drift factors or communication link interference. In severe cases, it will cause abnormalities such as inconsistent target imaging frames between modules, failed fusion alignment, virtual image superposition or key feature misalignment, reducing the accuracy and real-time performance of the overall visual system. This issue is particularly critical in applications with stringent spatial coupling and extremely high alignment requirements, such as large-scale VR / AR fusion and micro-multi-array light field construction. Conversely, a small signal synchronization delay runaway coefficient indicates that the response delays of each module in the system remain within controllable limits under dynamic conditions, demonstrating good temporal consistency between the modules, a stable signal trigger processing chain, and strong timing recovery and self-correction capabilities. Image fusion performed under this condition not only achieves high accuracy and robustness against disturbances, but also significantly reduces inter-module signal interference or computing resource contention caused by time mismatches. Assessing multi-module fusion failure barriers based on the signal synchronization delay runaway coefficient has several beneficial effects: First, it provides a time-domain anomaly-sensitive indicator, addressing the limitation of traditional spatial deviation indicators that cannot reflect the risk of asynchronous fusion failure caused by timing drift. Second, during the initial deployment of a multi-module system or during long-term maintenance, real-time monitoring of this coefficient's changing trends can identify potential link bottlenecks, resource conflicts, or node aging, providing a basis for optimizing data sampling strategies and load balancing mechanisms. The signal synchronization delay runaway coefficient is not only a measure of system temporal coordination but also a crucial foundational technology driving the evolution of multi-module systems toward high reliability, low coupling, and strong fault tolerance. It provides theoretical support and quantitative tools for failure prediction, stable operation, and resource optimization and control in multi-module image fusion systems. While enhancing system timing robustness and fusion accuracy, it also lays a key indicator foundation for building a high-precision, highly consistent module coordination system.

[0017] The logic for obtaining the signal synchronization delay out-of-control coefficient is as follows: In the time scale set (in The disturbance energy of the time sampling delay difference matrix under different time scales is extracted on the g-th time scale, G is a positive integer: ,in is the disturbance energy at the g-th time scale, is the module running time, t is the time unit, is the time index within the g-th time scale range, is the frame response delay difference between the i-th module and the j-th module, is the mean of the frame response delay difference; construct the multi-scale perturbation energy vector: ; Normalized perturbation energy: ,in is the normalized perturbation energy; calculate the information entropy of the perturbation distribution: ,in is the information entropy of the disturbance distribution; calculate the local out-of-control energy weighting coefficient: ,in is the local out-of-control energy weighting coefficient, is a minimum constant to prevent division by zero (usually ); Taking each module as a node, if the frame response delay difference is greater than the preset frame response delay difference threshold, it is considered that there is a delay imbalance edge, and a delay edge weight is constructed between the module nodes. : ; Get the synchronous delay disturbance propagation diagram ,in is the set of module nodes of the synchronous delay disturbance propagation graph, is the module node edge set of the synchronous delay disturbance propagation graph; calculates the average out-of-control path length of the propagation chain of the synchronous delay disturbance propagation graph : ,in is the set of shortest paths between all module node pairs, is the number of shortest paths between all module node pairs, is the shortest path; calculate the signal synchronization delay out-of-control coefficient: ,in is the signal synchronization delay out-of-control coefficient, is the information entropy coefficient of the disturbance distribution, , are the information entropy coefficient of the disturbance distribution and the preset proportional coefficient of the average out-of-control path length of the propagation chain, respectively, and All greater than 0; It should be noted that the above formulas are all dimensionless and numerical calculations. Common dimensionless methods include Min-Max normalization and Z-Score normalization, which will not be described here. Set it according to the actual situation. For example, adopt the expert empowerment method, that is, invite experts in related fields to determine the preset proportion coefficients of various indicators through professional opinion surveys and comprehensive evaluations, for example, It can be 0.5, 0.5; Step S4: constructing a dynamic dislocation fusion obstacle assessment model based on the micro-parallax offset cumulative coefficient and the signal synchronization delay out-of-control coefficient, outputting a dynamic dislocation fusion failure obstacle index, and determining the fusion failure obstacle level of the multi-module group; A dynamic dislocation fusion obstacle assessment model is constructed based on the microscopic parallax offset cumulative coefficient and the signal synchronization delay out-of-control coefficient, and a dynamic dislocation fusion failure obstacle index is output. The dynamic dislocation fusion obstacle assessment model is based on the following formula: , where is the dynamic dislocation fusion failure barrier index, is the microscopic parallax offset accumulation coefficient, is the signal synchronization delay out-of-control coefficient, They represent the preset proportional coefficients of the microscopic parallax offset accumulation coefficient and the signal synchronization delay out-of-control coefficient, respectively, and All greater than 0; It should be noted that the above formulas are all dimensionless and numerical calculations. Common dimensionless methods include Min-Max normalization and Z-Score normalization, which will not be described here. Set it according to the actual situation. For example, adopt the expert empowerment method, that is, invite experts in related fields to determine the preset proportion coefficients of various indicators through professional opinion surveys and comprehensive evaluations, for example, It can be 0.5, 0.5; From the above calculation expression, it can be seen that the larger the micro-parallax offset cumulative coefficient and the larger the signal synchronization delay out-of-control coefficient, the larger the dynamic misalignment fusion failure obstacle index, indicating that the more serious the misalignment of the multi-module system in the two dimensions of spatial imaging and temporal coordination, the higher the risk of fusion failure, which is manifested as typical obstacle phenomena such as blurred fusion boundaries, decreased recognition accuracy, increased response delay, and local view distortion. Conversely, the smaller the micro-parallax offset cumulative coefficient and the smaller the signal synchronization delay out-of-control coefficient, the smaller the dynamic misalignment fusion failure obstacle index, indicating that the multi-module system has high matching accuracy in spatial parallax alignment and temporal synchronization coordination, and the overall fusion process is stable and reliable. The fusion effect is manifested as natural image edge transition, good field of view continuity, clear target recognition, and strong response consistency. The system fusion obstacle risk is at an acceptable or even optimal level. The dynamic dislocation fusion failure barrier index is compared with the preset dynamic dislocation fusion failure barrier index threshold to determine the fusion failure barrier level of the multi-module, as follows: If the dynamic misalignment fusion failure obstacle index is greater than the dynamic misalignment fusion failure obstacle index threshold, the system is considered to have a strong fusion obstacle risk, indicating that there are significant deviations between the multiple modules in microscopic parallax alignment and signal timing coordination, and the possibility of fusion failure is high. It is necessary to adjust the system parameters, optimize the module layout or calibrate the synchronization mechanism in time to avoid image quality degradation or data misjudgment. The fusion failure obstacle level is assessed as a high-risk level. When the multi-module fusion failure obstacle level is assessed as a high-risk level, the robust fusion control mechanism can be triggered, including: multi-module image reconstruction order adjustment, dynamic adjustment of key module image weighting, signal trigger delay prediction compensation, and dynamic weighted dedistortion processing of module registration points. If the dynamic misalignment fusion failure barrier index is less than or equal to the dynamic misalignment fusion failure barrier index threshold, the system is considered to be in an acceptable fusion state, indicating that the parallax offset and synchronization delay are controlled within a reasonable range, the collaborative fusion performance between multiple modules is stable and reliable, the imaging or perception effect is normal, and the fusion failure barrier level is assessed as normal, without excessive intervention; Step S5: During long-term operation, record the obstacle trend sequences under different dynamic scenarios to build a scene label-obstacle trend pattern association library, map the scene label-obstacle trend pattern association library with the scheduling strategy, and quickly adapt across scenarios, as follows: In any time window Record obstacle trend sequences in different dynamic scenarios , while recording the environment state vector at each time point: ,in is the light intensity, is the module movement speed, is the target density; It should be noted that the environmental state quantities in the environmental state vector include but are not limited to light intensity, module movement speed, and target density; Use feature clustering algorithm (such as K-Means clustering algorithm) to classify the long-term recorded environmental state vector Perform clustering to obtain a standardized scene label set: ,in Indicates the hth operating scenario label. Each label represents a typical operating scenario (such as "high-speed movement + strong light interference" or "stationary + low target density"), which serves as the basis for attributing the obstacle mode. Each scene is labeled with obstacle trend patterns such as short-term rapid rise followed by stabilization (sudden disturbance type), slow continuous climb (systematic degradation type), high-frequency fluctuation type (external interference type), and low-level stability type (normal operation type). Perform time series clustering (such as K-Means clustering algorithm) on the obstacle trend series below to obtain the obstacle trend pattern set: ,in For the fth obstacle trend pattern, the scene label-obstacle trend pattern association library is finally formed: ;Build the decision mapping function: ,in The c-th scheduling strategy (including maintaining the original fusion architecture, adjusting module weights, dynamic confidence redistribution, introducing backup modules for parallel processing, switching to low-precision redundant algorithms, etc.) The present invention performs periodic structural calibration on micro-optical modules to accurately obtain spatial imaging deviations and time sampling delay information between modules, establishes a high-resolution, high-precision spatial-temporal state perception foundation from the source, and improves the parameter stability and data consistency of the collaborative operation of modules; obtains the micro-parallax offset accumulation coefficient, and comprehensively quantifies the error accumulation trend caused by factors such as optical axis offset and thermal deformation at the spatial imaging level; at the same time, obtains the signal synchronization delay out-of-control coefficient, effectively revealing the inter-frame acquisition inconsistency problem caused by factors such as clock drift and acquisition link differences in the multi-module system, and constructs a dynamic dislocation fusion obstacle assessment model based on the above coefficients to output a dynamic dislocation fusion failure obstacle index, thereby realizing quantitative assessment of the collaborative stability of the multi-module system and failure level classification, and providing a highly operational reference basis for subsequent control and scheduling. Furthermore, by recording the obstacle trend sequences under different operating scenarios and combining them with the environmental state vector for clustering and attribution modeling, an association library of scene labels and obstacle trend patterns is established, and a scheduling strategy mapping mechanism is introduced to ensure that the system can quickly identify scene patterns and automatically match the optimal control strategy when facing typical complex working conditions such as high-speed movement, low light interference, and structural occlusion, thereby achieving rapid adaptation and stable operation across scenarios. Not only does it achieve accurate modeling and quantitative evaluation of the spatial imaging error and time sampling out-of-control of the micro-optical array module at the perception level, but it also establishes a control strategy mapping mechanism associated with the evolution of obstacle trends at the decision-making level, significantly improving the system's fusion robustness, adaptability, and multi-module collaborative stability in complex scenarios, breaking through the technical bottleneck of the existing technology that cannot effectively deal with the problem of space-time coupling failure.

[0018] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0019] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0020] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A scalable, standardized, micro-optical array module multi-scenario adaptation method, characterized by: The steps include: Step S1, performing periodic structural calibration on each micro-optical module in the system, and obtaining a spatial imaging deviation distribution map and a time sampling delay difference matrix between the modules; Step S2, obtaining microscopic parallax offset accumulation information between multiple modules according to the spatial imaging deviation distribution map between modules, wherein the microscopic parallax offset accumulation information includes a microscopic parallax offset accumulation coefficient; Step S3, obtaining signal synchronization delay out-of-control information among multiple modules according to the time sampling delay difference matrix among the modules, wherein the signal synchronization delay out-of-control information includes a signal synchronization delay out-of-control coefficient; Step S4: constructing a dynamic dislocation fusion obstacle assessment model based on the micro-parallax offset cumulative coefficient and the signal synchronization delay out-of-control coefficient, outputting a dynamic dislocation fusion failure obstacle index, and determining the fusion failure obstacle level of the multi-module group; Step S5: During long-term operation, record the obstacle trend sequences under different dynamic scenarios to build a scene label-obstacle trend pattern association library, map the scene label-obstacle trend pattern association library with the scheduling strategy, and quickly adapt across scenarios.

2. The scalable standardized micro-optical array module multi-scene adaptation method according to claim 1, characterized in that: Acquisition of spatial imaging deviation distribution map: Use multiple modules to collaboratively shoot static and synchronous images of the standard pattern; each module shoots the same target standard pattern; Extract feature point set for each module image: ,in is the feature point set of the i-th module, is the nth feature point; Randomly select one of the modules as the reference module, align the other modules with the reference module's viewing angle, and calculate the point deviation: ,in is the point deviation, To map the reference module perspective to the transformation of the i-th module, is the feature point set of the reference module; All modules will be The point deviations between them are interpolated and visualized on the spatial coordinates to obtain the spatial imaging deviation distribution map .

3. The scalable standardized micro-optical array module multi-scene adaptation method according to claim 1, characterized in that: Obtaining the time sampling delay difference matrix: Use a unified synchronous trigger signal source to trigger all modules to collect images. Each module records the time delay between receiving the trigger signal and the actual generation of image data. ; For each module pair Calculate the frame response delay difference: ,in is the frame response delay difference between the i-th module and the j-th module, is the time delay between the i-th module receiving the trigger signal and the actual generation of image data, is the time delay between the i-th module receiving the trigger signal and the actual generation of image data; Based on all modules Constructing the time sampling delay difference matrix : ,in , The total number of modules.

4. The scalable standardized micro-optical array module multi-scene adaptation method according to claim 2, characterized in that: The logic for obtaining the micro-parallax offset accumulation coefficient is as follows: According to the spatial imaging deviation distribution map Compute the residual tensor: ,in is the residual tensor, for The global mean of ; using pixel blocks as nodes and residual change gradients as edge weights, a deviation propagation graph is constructed: ,in For nodes and nodes The residual gradient edge weight, For nodes The residual tensor of For nodes The residual tensor of ; The graph structure of the deviation propagation graph is: ,in is the set of pixel block nodes in the deviation propagation graph, is the node-edge set in the deviation propagation graph; According to the module structure state, the optical axis disturbance response function is defined as: ,in is the optical axis disturbance response value, is the angle between the optical axes of the i-th module and the j-th module, is the temperature gradient difference between the i-th module and the j-th module, is the peak value difference of vibration response between the i-th module and the j-th module, is the preset proportional coefficient of the optical axis angle, temperature gradient difference, and vibration response peak difference, and All greater than 0; The optical axis disturbance response value Normalized to the perturbation factor: ,in is the disturbance factor, is a maximum value acquisition function used to obtain the maximum value of the optical axis disturbance response value; In the deviation propagation diagram The shortest path search algorithm is executed on the MATLAB to calculate the spatial path offset potential energy: ,in is the spatial path offset potential energy, is the shortest path in the deviation propagation graph, is the residual gradient edge weight, is the number of path hops; Calculate the microscopic parallax offset cumulative coefficient: ,in is the microscopic parallax offset accumulation coefficient, is a very small constant to prevent division by zero, , The total number of modules.

5. The scalable standardized micro-optical array module multi-scene adaptation method according to claim 3, characterized in that: The logic for obtaining the signal synchronization delay out-of-control coefficient is as follows: In the time scale set Extract the disturbance energy of the time sampling delay difference matrix at different time scales: ,in is the disturbance energy at the g-th time scale, is the module running time, t is the time unit, is the time index within the g-th time scale range, is the frame response delay difference between the i-th module and the j-th module, is the mean of the frame response delay difference; construct the multi-scale perturbation energy vector: ; Normalized perturbation energy: ,in is the normalized perturbation energy; calculate the information entropy of the perturbation distribution: ,in is the information entropy of the disturbance distribution; calculate the local out-of-control energy weighting coefficient: ,in is the local out-of-control energy weighting coefficient, is a minimum constant to prevent division by zero; taking each module as a node, if the frame response delay difference is greater than the preset frame response delay difference threshold, it is considered that there is a delay imbalance edge, and a delay edge weight is constructed between the module nodes. : ; Get the synchronous delay disturbance propagation diagram ,in is the set of module nodes of the synchronous delay disturbance propagation graph, is the set of module nodes and edges of the synchronous delay disturbance propagation graph; Calculate the average out-of-control path length of the propagation chain of the synchronous delay disturbance propagation graph : ,in is the set of shortest paths between all module node pairs, is the number of shortest paths between all module node pairs, is the shortest path; Calculate the signal synchronization delay out-of-control coefficient: ,in is the signal synchronization delay out-of-control coefficient, is the information entropy coefficient of the disturbance distribution, , are the information entropy coefficient of the disturbance distribution and the preset proportional coefficient of the average out-of-control path length of the propagation chain, respectively, and Both are greater than 0.

6. The scalable standardized micro-optical array module multi-scene adaptation method according to claim 1, characterized in that: A dynamic dislocation fusion obstacle assessment model is constructed based on the microscopic parallax offset cumulative coefficient and the signal synchronization delay out-of-control coefficient, and a dynamic dislocation fusion failure obstacle index is output. The dynamic dislocation fusion obstacle assessment model is based on the following formula: , where is the dynamic dislocation fusion failure barrier index, is the microscopic parallax offset accumulation coefficient, is the signal synchronization delay out-of-control coefficient, They represent the preset proportional coefficients of the microscopic parallax offset accumulation coefficient and the signal synchronization delay out-of-control coefficient, respectively, and Both are greater than 0.

7. The scalable standardized micro-optical array module multi-scene adaptation method according to claim 6, characterized in that: The dynamic dislocation fusion failure barrier index is compared with the preset dynamic dislocation fusion failure barrier index threshold to determine the fusion failure barrier level of the multi-module, as follows: If the dynamic dislocation fusion failure barrier index is greater than the dynamic dislocation fusion failure barrier index threshold, the fusion failure barrier level is assessed as a high risk level; If the dynamic dislocation fusion failure obstacle index is less than or equal to the dynamic dislocation fusion failure obstacle index threshold, the fusion failure obstacle level is assessed as normal.

Images