Method and system for correction of industrial close-range photogrammetry images

Abstract

The application discloses a correction method and system for industrial close-range photogrammetry images, and relates to the technical field of computer vision, comprising: S100, constructing a cross-time baseline of light path observation under multi-time sampling conditions, performing joint extraction operation of brightness distribution and phase gradient on each frame of image, generating an anchor point sequence, and constructing an interference risk atlas according to the change trend of optical path difference, which is used as a dynamic constraint basis for feature recognition; S200, performing phase polarization consistency audit based on the anchor point sequence and the interference risk atlas, performing pixel-by-pixel tracking on the phase mutation area in the brightness field, extracting interference false peaks with phase inversion characteristics, and outputting a false edge candidate data set. The application constructs time anchor points and interference atlases, realizes interference perception and early warning, extracts real edges by combining phase audit and optical flow registration, reconstructs imaging geometry and optimizes distortion, suppresses interference by combining phase conjugation and time rearrangement, and realizes closed-loop control and high-precision image correction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer vision, in particular to a correction method and system for industrial close-range photogrammetry images. BACKGROUND

[0002] The image correction for industrial close-range photogrammetry refers to the whole process of implementing systematic image correction processing on images obtained by close-range shooting in the process of industrial measurement or three-dimensional reconstruction, so as to eliminate geometric and optical errors. Due to the complex lighting, space limitation, mechanical vibration, high reflective metal surface and other conditions of the camera in the industrial environment, the original image is prone to distortion, tilt, scale distortion, focal length drift, uneven lighting and other problems in the imaging process, thereby significantly reducing the measurement accuracy. The image correction process establishes a high-precision camera imaging model, parameterizes modeling and inverse calculation of lens distortion, principal point offset, pixel nonlinear deformation, and restores the true geometric relationship of the image by combining the spatial reference point or calibration board information, so that the image coordinates and the actual physical space coordinates are accurately corresponding. At the same time, the image correction process introduces a lighting balance and color consistency correction mechanism to compensate and reconstruct the shadow, reflection and uneven brightness area, so as to ensure that the corrected image can truly reflect the structure, shape and size characteristics of the measured object, and provide high-reliability visual data support for industrial detection, assembly precision control and digital modeling.

[0003] In the prior art, when the measured object has a high-reflectivity metal surface, the incident light will undergo multiple reflections and superpositions on the surface, forming a complex optical path interference effect. Due to the small changes in reflection angle and surface roughness, the phase superposition caused by the optical path difference will cause the phase inversion phenomenon in the local pixel area, thereby forming a false peak signal with abnormal mutation of brightness in the imaging result. The false peak presents similar characteristics to the true physical boundary in the gray gradient, so that the edge detection algorithm based on brightness change in the prior art is easy to misidentify it as an actual edge line, thereby introducing false feature points in the feature point extraction stage. With the accumulation of mis-matching points, the spatial registration model is geometrically unstable, which causes the grid topological structure to be broken or distorted in three-dimensional reconstruction, seriously affecting the spatial accuracy and geometric continuity of the close-range photogrammetry result.

[0004] The above information disclosed in the background section is only used to enhance the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0005] The purpose of the present application is to provide a correction method and system for industrial close-range photogrammetry images to solve the problems in the background.

[0006] In order to achieve the above object, the present application provides the following technical scheme: a correction method for industrial close-range photogrammetry images, comprising the following steps:

[0007] S100, under multi-time sampling conditions, constructing a cross-time baseline of light path observation, performing joint extraction of brightness distribution and phase gradient on each image frame, generating an anchor point sequence with time continuity, and constructing an interference risk map according to the change trend of optical path difference, which is used as a dynamic constraint basis for feature recognition;

[0008] S200, performing phase polarization consistency audit based on the anchor point sequence and the interference risk map, performing pixel-by-pixel tracking on the phase mutation area in the brightness field, extracting interference false peaks with phase inversion characteristics, and outputting a false edge candidate dataset containing spatial position and phase state;

[0009] S300, performing multi-view optical flow depth coupling registration based on the false edge candidate dataset and the phase audit result, removing edge points with time drift through cross-frame optical flow consistency detection, and converting edge regions with stable morphological characteristics into a real edge evidence dataset;

[0010] S400, reconstructing the imaging geometry based on the real edge evidence dataset, performing real-time adjustment of the camera internal and external parameters, and performing online inversion and recursive optimization of the nonlinear distortion parameters, establishing a self-consistent correction field covering the whole time sequence, and realizing the alignment of the light path system and the imaging coordinate system;

[0011] S500, performing dynamic control operation based on the self-consistent correction field, performing periodic reverse modulation on the interference false peak energy distribution through phase conjugate polarization rotation scanning combined with reversible time grid rearrangement mechanism, continuously suppressing the phase inversion area, and constructing a closed-loop control process for interference detection and interference source suppression.

[0012] Preferably, step S100 comprises:

[0013] Under multi-time sampling conditions, an imaging structure with spatial calibration reference is established, and an image frame sequence containing brightness information and phase influence is collected by configuring a multi-spectral imaging device with adjustable focal length;

[0014] Based on the collected image frame sequence, joint feature data of brightness distribution and phase gradient are extracted, and local gray level continuity and phase change trend analysis of pixels are performed, and a joint feature point dataset is constructed;

[0015] Based on the constructed joint feature point dataset, the region with consistent change characteristics in the time sequence is selected as the anchor point candidate region, and spatial back projection verification is performed combined with the light path observation baseline to determine the time anchor point sequence;

[0016] Based on the time anchor point sequence and the light path change trend, the gray level fluctuation and phase oscillation analysis is performed on the non-anchor point area, an interference risk map for dynamic constraint is constructed, and the distribution of the interference risk area existing in the image is recorded.

[0017] Preferably, the step S200 comprises:

[0018] Based on the time anchor point sequence and the interference risk map, the continuous pixel area with high interference risk score in the image pixel level range is screened as the audit area, and the spatial proximity relationship with the anchor point is established;

[0019] The contrast analysis of the brightness change curve and the anchor point brightness track is performed on each pixel point in the audit area, and the gray response difference under different polarization directions is combined to identify the phase inconsistent area;

[0020] The pixel-by-pixel time sequence tracking is performed on each pixel point in the phase inconsistent area, the brightness polarity reversal point and the polarization response synchronization feature are extracted, and it is determined whether there is a phase reversal phenomenon;

[0021] According to the determination result, the pseudo-edge candidate data set containing the spatial position and phase state information is output, the boundary fitting and the cross verification with the time anchor point are performed, and the pseudo-peak track set with clear structure is formed.

[0022] Preferably, in the pixel-by-pixel time sequence tracking process, the phase reversal phenomenon is jointly determined according to the distribution frequency of the brightness polarity reversal point, the spatial jump amplitude and the polarization response change synchronization, and only when the three conditions are met at the same time, the corresponding pixel is marked as an interference pseudo-peak.

[0023] Preferably, the step S300 comprises:

[0024] Based on the spatial position index and the phase history information of the pseudo-edge candidate data set, the corresponding area in the multi-view image sequence is called, and the observation frame sequence with time sequence relationship is constructed for subsequent analysis;

[0025] The optical flow field frame-by-frame tracking is performed on the pseudo-edge area in each observation frame sequence, the pixel motion track between the continuous image frames is established, and the spatial depth consistency of each pixel is calculated by introducing the parallax matching;

[0026] Based on the optical flow tracking result, the cross-frame consistency detection is performed on all pseudo-edge points, the phase response track and the interference score change are combined, and the non-stable edge points with time sequence drift and spatial jump are removed;

[0027] All the edge regions passing the consistency verification are constructed as the real edge evidence data set, the spatial position, the depth estimate and the contour shape information are extracted, and high credibility input is provided for subsequent geometric reconstruction.

[0028] Preferably, in the process of constructing the real edge evidence dataset, sub-pixel level boundary fitting is performed on the edge regions passing the consistency verification, and an edge morphology template is generated for use as an edge constraint input in subsequent imaging geometry reconstruction.

[0029] Preferably, step S400 comprises:

[0030] Based on the edge coordinates, optical flow trajectories, phase features and depth values in the real edge evidence dataset, a spatial back-projection operation is performed on each frame of image to establish the projection geometry relationship of the imaging light path;

[0031] Based on the projection geometry relationship, an error distribution map is constructed, and a dynamic adjustment operation of the camera internal parameters and external parameters is performed to optimize the camera imaging center position, focal length, optical axis offset and external attitude information;

[0032] After completing the camera parameter adjustment, the modeling and recursive optimization of the nonlinear distortion parameters are performed based on the theoretical trajectory and actual position of the edge points to compensate for the radial and tangential distortion errors;

[0033] The camera parameters and distortion model of each frame of image are fused to construct a self-consistent correction field covering the entire time sequence, which is used to guide the geometric alignment and light path consistency correction of all image frames in space.

[0034] Preferably, in the recursive optimization process of the nonlinear distortion parameters, based on the matching error of the edge profile in each frame of image, the distortion model is locally reconstructed, and a spatial consistency evaluation mechanism is used to dynamically adjust the parameter weight to improve the geometric correction accuracy.

[0035] Preferably, step S500 comprises:

[0036] Based on the imaging parameters, light path structure and distortion correction model recorded in the self-consistent correction field, a light path regulation structure with time continuity and spatial consistency is constructed, and a target regulation node with phase inversion characteristics is identified;

[0037] Through the phase conjugate polarization rotation scanning mechanism, the polarization state of the incident light is adjusted in the high reflection area range to form a structure complementary to the reflection path phase, which intervenes in the phase superposition process of the false peak;

[0038] Through the reversible time grid rearrangement mechanism, the interference intensity evolution trajectory in the image frame sequence is reconstructed, the inter-frame time structure is adjusted to weaken the false peak energy accumulation, and an interference regulation strategy in the time dimension is constructed;

[0039] Based on the aforementioned phase and time dual regulation methods, a closed-loop control process of interference detection and interference source suppression is constructed, and the regulation parameters are set before image acquisition and updated based on the response feedback after acquisition to update the next frame of regulation instruction.

[0040] The application also provides a correction system for industrial close-range photogrammetry images, comprising:

[0041] The time-scale light path observation module is configured to construct a cross-time-scale light path observation baseline under multi-time-scale sampling conditions, perform a joint extraction operation of brightness distribution and phase gradient on each frame of image, generate an anchor point sequence with time continuity, and construct an interference risk atlas according to a light path difference change trend, which is used as a dynamic constraint basis for feature recognition;

[0042] The phase consistency auditing module is configured to perform phase polarization consistency auditing based on the anchor point sequence and the interference risk atlas, perform pixel-by-pixel tracking on a phase mutation area in the brightness field, extract an interference false peak with a phase inversion feature, and output a false edge candidate dataset containing a spatial position and a phase state;

[0043] The optical flow coupling registration module is configured to perform multi-view optical flow depth coupling registration based on the false edge candidate dataset and the phase auditing result, remove edge points with time sequence drift through cross-frame optical flow consistency detection, and convert edge regions with stable morphological features into a real edge evidence dataset;

[0044] The geometric correction reconstruction module is configured to reconstruct an imaging geometry based on the real edge evidence dataset, perform real-time adjustment of camera internal parameters and external parameters, perform online inversion and recursive optimization on nonlinear distortion parameters, establish a self-consistent correction field covering the full time sequence, realize alignment of the light path system and the imaging coordinate system, and realize alignment of the light path system and the imaging coordinate system;

[0045] The interference suppression control module is configured to perform dynamic regulation and control operation based on the self-consistent correction field, perform periodic reverse modulation on interference false peak energy distribution through phase conjugate polarization rotation scanning combined with a reversible time grid rearrangement mechanism, continuously suppress phase inversion areas, and construct a closed-loop control process for interference detection and interference source suppression.

[0046] In the above technical solution, the application provides technical effects and advantages:

[0047] The application realizes dynamic perception and risk warning of image interference potential by constructing anchor point sequence with time continuity and interference risk map, accurately identifies and isolates the pseudo-edge area with phase inversion characteristics by combining phase polarization consistency audit and pixel-by-pixel tracking, further removes the unstable feature points disturbed by interference under the multi-view optical flow registration mechanism, and extracts highly reliable real edge evidence. The imaging geometry reconstruction and distortion optimization based on the evidence dataset can realize accurate correction of imaging parameters in the whole time sequence, and establish an integrated alignment relationship between the optical path and the image. On this basis, by means of phase conjugate polarization scanning and time grid rearrangement strategy, active regulation and periodic inhibition mechanism of interference energy is formed, and a closed-loop control process from interference detection to adaptive regulation is constructed, realizing full-process suppression and feedback adjustment of interference source. The overall scheme has the synergistic advantages of dynamic perception, accurate correction and adaptive suppression, and finally realizes accurate identification and stable elimination of interference false peaks in industrial close-range images, providing a solid data foundation and robust image guarantee for high-precision three-dimensional reconstruction and precision vision measurement. BRIEF DESCRIPTION OF DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0049] Figure 1 The method flow chart for the correction method of the industrial close-range photogrammetry image of the present application.

[0050] Figure 2 The module schematic diagram for the correction system of the industrial close-range photogrammetry image of the present application. DETAILED DESCRIPTION

[0051] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations may, however, be implemented in many different forms and should not be construed as limited to the examples set forth herein; rather, these example implementations are provided so that this disclosure will be thorough and complete, and will fully convey the scope of example implementations to those skilled in the art.

[0052] The present application provides a correction method for an industrial close-range photogrammetry image as shown in Figure 1 The present application provides a correction method for an industrial close-range photogrammetry image as shown in

[0053] S100, under the condition of multi-time sampling, constructing a cross-time optical path observation baseline, performing joint extraction of brightness distribution and phase gradient on each frame of image, generating an anchor point sequence with time continuity, and constructing an interference risk map according to the change trend of optical path difference, which is used as a dynamic constraint basis for feature recognition;

[0054] In the industrial close-range photogrammetry image processing, in order to solve the problem of interference false peak caused by high reflective metal surface, a technical path of constructing a cross-time optical path observation baseline under the condition of multi-time sampling is proposed, and through joint extraction of image brightness distribution and phase gradient, an anchor point sequence with time continuity is generated, and then an interference risk map is constructed according to the change trend of optical path difference, which provides a dynamic constraint basis for subsequent feature recognition and false feature elimination. The specific implementation steps are as follows:

[0055] In the industrial measurement environment with stable sampling frequency and fixed space reference, a multi-spectral imaging device with adjustable focal length is arranged, and a spatial calibration structure with sub-pixel level positioning capability is configured. The structure is composed of high-precision calibration points distributed at different positions in three-dimensional space, which is used for imaging reference comparison at multiple time nodes. By image acquisition of the same static target at consecutive time points, a group of time sequence image frames containing brightness information and phase interference effect is obtained. The sampling time interval is dynamically adjusted according to the external interference environment to ensure that the image sequence is still comparable under the conditions of light source stability change, small vibration disturbance or surface reflection change. By extracting the pixel distribution change of the calibration structure in each frame of image, a preliminary cross-time optical path observation baseline is constructed, which records the optical path information of light from incidence to imaging at different times, and is used as a time scale reference for subsequent feature stability.

[0056] On the basis of the constructed cross-time optical path observation baseline, the brightness distribution extraction operation is performed on the target object region in each frame of image. When extracting, the pixel-level gray scale layering method is adopted, and each pixel point is locally scanned to capture the continuity and stability of gray scale change. At the same time, the phase gradient extraction operation is performed on the image, that is, the local phase tilt characteristics of light after interference on the target surface are analyzed. This step deduces the spatial gradient of phase propagation path by calculating the gray trend change direction of each pixel point in different time frames. The brightness distribution reflects the strong and weak change of surface reflection energy density, and the phase gradient reflects the optical path change trend caused by surface topography disturbance. By superimposing the time sequence data of brightness and phase at the same pixel position, and comparing the change form in space, the joint feature point data set describing the continuous change characteristics of image is formed, which provides multi-dimensional support for constructing stable anchor points.

[0057] On the basis of completing the brightness and phase joint feature extraction, around the double standards of time consistency and spatial stability, the pixel region which exists in continuous multiple frames of images and has consistent change trend is screened out as the anchor point candidate region with time continuity. In order to ensure the stability of the anchor point, the spatial geometric constraint is further introduced, that is, according to the light path observation baseline established in the previous sequence, the anchor point candidate region is back projected to the three-dimensional calibration reference system, and the change amplitude and direction stability of the corresponding light path at different time are analyzed. Only when a certain anchor point candidate region presents a flat brightness change, consistent phase gradient and a light path back projection error lower than a set threshold in the entire time sequence, the region is formally defined as a time anchor point. The record of each anchor point not only includes its pixel coordinates and gray value on the image plane, but also includes its change curve in the time dimension, phase propagation direction and its geometric position mapped into the physical space, which provides accurate time and space double labels for subsequent comparison with the interference mode.

[0058] After the anchor point sequence is established, the time tracking ability of the anchor point sequence to the local region change of the image is used, combined with the light path change trend, to construct a graph reflecting the risk degree of light interference. Specifically, for each non-anchor point pixel region, first search its gray fluctuation curve in the time sequence, and compare and analyze with the nearest anchor point region. If the region has frequent gray polarity inversion, high amplitude gradient jump or nonlinear oscillation of phase change path, it will be marked as a high interference risk region. At the same time, the interference period matching mechanism between images is introduced to compare the diffusion mode of interference intensity in space and judge whether the interference phenomenon has propagation. Finally, all the identified interference regions are quantitatively scored according to the light path difference change amplitude, brightness disturbance frequency and phase shift direction, and a two-dimensional matrix is formed to form an interference risk graph covering the entire image region. The graph is not only used to determine whether there is an interference false peak risk in the local region, but more importantly, it provides a risk evolution trajectory over time, which provides a dynamic constraint reference for avoiding high-risk areas in subsequent image feature recognition, avoids including false features caused by interference into the edge extraction or geometric registration process, thereby improving the spatial reliability of image correction and the consistency of three-dimensional reconstruction.

[0059] S200, based on the anchor point sequence and the interference risk graph, performing phase polarization consistency audit, tracking the phase mutation region in the brightness field pixel by pixel, extracting the interference false peak with phase inversion feature, and outputting the false edge candidate dataset containing the spatial position and phase state;

[0060] After the anchor point sequence and the interference risk map are constructed, in order to identify the potential interference false peak region in the image, the region with significant phase mutation in the brightness field is audited in detail and tracked pixel by pixel, and a false edge candidate dataset containing spatial position and phase state is established, which provides stable and reliable data support for subsequent feature screening and geometric reconstruction. The specific steps are as follows:

[0061] Based on the constructed time anchor point sequence and the corresponding interference risk map, region pre-screening is performed on the pixel level distribution of the image. In the specific operation process, first, according to the interference risk score of each pixel in the interference risk map, the continuous pixel region with a risk level higher than the set threshold is selected as the preliminary audit region. These regions are usually located in the positions where the specular reflection is dense or the surface reflection direction changes rapidly, and show optical abnormal phenomena such as gray inversion, brightness abnormal amplification or phase mutation in multiple image frames. A preliminary pixel group set is established for these regions, and the spatial coordinates thereof are associated with the nearest multiple time anchors to form a spatial proximity relationship. The key of this step is to establish a time reference benchmark for the fluctuation state of the high-risk pixel group by means of the time continuity and spatial stability information provided by the anchor points, so as to facilitate the differential analysis in the subsequent phase analysis.

[0062] In the selected high-risk pixel region, the phase consistency audit operation is performed on the brightness sequence of the target pixel region in combination with the spatial stable region provided by the anchor point sequence and the time behavior curve thereof. Specifically, the brightness change curve of each pixel point in the target region is aligned and compared with the brightness change trajectory in the time anchor point, and whether the fluctuation trend of the gray value of the two in the same time period is obviously deviated is analyzed, to judge whether the target pixel has phase inconsistency phenomenon. At the same time, the data in the polarization response channel are introduced, the gray difference of the same pixel under different polarization directions is statistically analyzed, and whether there is a sign of optical path change mutation under a specific polarization angle is judged. If the pixel point shows different gray change directions in multiple time segments compared with the anchor point region, and presents a periodic enhancement and decay alternating response curve in the polarization channel, it is determined that the pixel point is in a phase unstable region, which may be affected by the surface interference effect, and belongs to the potential false peak source.

[0063] After completing the pixel-level phase polarization consistency audit, a pixel-by-pixel cascade tracking operation is performed on the pixels marked as phase inconsistent regions. This operation indexes by time and traces back and forward for each abnormal pixel in consecutive image frames to establish a complete time sequence evolution trajectory by tracking its spatial position, gray value and phase response changes. By analyzing the distribution frequency of the brightness polarity reversal point, the spatial jump amplitude and the polarization response synchronization in the trajectory, it is further determined whether there is a typical phase inversion phenomenon, i.e. the brightness pattern similar to the mirror structure appears in multiple frames and is alternately consistent with the polarization state. If the above conditions are met, it is determined that the trajectory contains an interference false peak. At the same time, a second screening is also needed to combine the score trend of the pixel in the interference risk map to avoid mislabeling the brightness anomaly caused by non-interference as a false peak. Finally, the key pixel frames are extracted from all the effective false peak trajectories determined by consistency audit and trajectory tracking, and are aggregated into a spatially continuous and temporally stable false peak structure.

[0064] Based on the above determination results, the false edge candidate dataset containing spatial position and phase state information is sorted and output. The specific output content includes the image frame number where the false peak pixel is located, the two-dimensional pixel coordinates, the historical gray scale change curve, the response feature sequence under the polarization channel, the phase mutation point position and the corresponding interference risk score. In addition, the false edge region needs to be reconstructed, the boundaries of multiple false peak trajectories are fitted on the image plane, the possible edge morphological features are extracted, and are marked as to-be-removed objects or strong constraint regions. In the output process, the false edge data is also cross-verified with the anchor point sequence in the previous stage to ensure that the spatial boundaries of the two types of data do not overlap, avoiding the misremoval of effective feature points due to anchor point misjudgment. Finally, a multi-dimensional dataset containing full-frame spatial coordinates, phase history, polarization response and interference annotation is formed, providing a clear structure and clear source data support foundation for subsequent feature point screening, mis-matching removal and geometric model constraint.

[0065] S300, based on the false edge candidate dataset and the phase audit result, performing multi-view optical flow depth coupling registration, removing edge points with time drift through cross-frame optical flow consistency detection, and converting edge regions with stable morphological features into a real edge evidence dataset;

[0066] After completing the extraction of the false edge candidate dataset and the phase polarization consistency audit, in order to improve the reliability of the edge features in spatial reconstruction, a depth registration operation is needed on the false edge region through multi-view information fusion and optical flow field consistency detection, to screen out real edge regions with temporal stability and spatial consistency, and to construct them into a real edge evidence dataset for subsequent imaging geometric structure reconstruction and error control. The specific steps are as follows:

[0067] Based on the spatial position index of the pseudo-edge candidate dataset and the phase history information carried by it, the candidate edge regions are synchronously recalled and feature region cropped in the multi-angle image sequence. In specific implementation, for each set of pseudo-edge pixels, according to the trajectory characteristics thereof in the continuous frames, the corresponding observation image regions are extracted from the adjacent frames, and these regions are arranged in a unified scale, relative angle interval and time sequence to form a multi-frame observation sequence with clear time sequence relationship and spatial angle distribution characteristics. The purpose of this step is to establish a corresponding region set of each pseudo-edge region in the multi-view image, thereby providing a continuous spatial basis and time sequence constraint for subsequent optical flow field coupling analysis. In this process, particular reference is made to the polarization response difference recorded in the phase consistency audit result of the previous stage, which is used for pre-screening those edge regions that still maintain phase feature continuity under different angles, and is preferentially used as a coupling candidate region.

[0068] For the multi-view image sequence constructed, a frame-by-frame optical flow field tracking operation is performed around the pseudo-edge region to establish the spatial mapping relationship of pixel motion between continuous frames. In specific operation, in each frame of image, the motion trend of the pixel points of the pseudo-edge region is analyzed in detail, the translation, rotation or scaling trend of the pixel points on the time axis is calculated by comparing the gray change direction and amplitude of the same position between adjacent frames, and a continuous motion trajectory is formed. In the optical flow tracking process, the local structure fidelity of the pixel points, i.e., whether the texture structure around the pixel remains relatively consistent in multiple frames of images, is analyzed. In order to enhance the robustness of the optical flow tracking, the phase mutation point information in the pseudo-edge trajectory extracted in the previous stage is also combined to specially mark the positions of sudden changes in the trajectory, so as to identify potential unstable factors in subsequent analysis. At the same time, the depth estimation operation of disparity matching is introduced, the relative position relationship of each pixel in multiple frames of view is geometrically reprojected, the actual depth consistency of the pixel in space is analyzed, and false edge structures that are highlighted only in a certain angle or time frame but cannot be confirmed by multiple angles are eliminated.

[0069] After obtaining the multi-frame optical flow field tracking results, cross-frame consistency detection operation is performed on the time trajectories of all pseudo-edge pixel points to screen out edge regions with stable morphological evolution characteristics. The specific method of consistency detection is to compare the continuity of the optical flow trajectory of each pseudo-edge point in all consecutive frames to determine whether there is a trajectory interruption, displacement mutation or direction jump phenomenon; for the trajectory with mutation or discontinuity, further search its evolution trajectory in the phase information, and compare the score change in the interference risk map. If it is found that the trajectory is discontinuous in the optical flow field, the phase response has asymmetric change, and the interference score presents sharp fluctuation in the time dimension, the edge point is marked as having time sequence drift risk and is removed. Through cross-frame optical flow consistency verification of all pseudo-edge points, those pixel points with unstable motion trajectory, inconsistent depth or lack of multi-angle coincidence support are gradually removed, thereby improving the time sequence stability and spatial consistency of the final edge data. This process relies on the spatial view set and optical flow trajectory library established in the previous two steps, and realizes the accurate pseudo-edge removal mechanism by fusing the three data supports of motion characteristics, depth structure and phase response.

[0070] After completing the optical flow consistency detection and unstable edge point removal, all verified pseudo-edge regions are classified as real edge regions, and a real edge evidence data set is constructed. The data set includes the position coordinates, motion trajectory, depth estimate, optical flow consistency score and phase stability evaluation index of each edge region in multiple frames of images, and also includes the score evolution curve of each region in the interference risk map for further analysis of its robustness under different sampling conditions. To ensure the spatial continuity of the data set, boundary splicing processing is performed on all verified edge regions, the contour curve is fitted with sub-pixel accuracy on the image plane, and the corresponding edge shape template is generated for use as edge constraint input in the subsequent geometric reconstruction and correction parameter optimization stage. The construction process of the data set not only improves the geometric reality and time continuity of the edge data, but also provides a high reliability edge structure reference for subsequent imaging structure reconstruction and interference influence elimination by introducing multi-view depth fusion and motion consistency evaluation mechanism, effectively ensuring the overall image measurement accuracy and the stability of the reconstruction model.

[0071] S400, reconstruct the imaging geometric structure based on the real edge evidence data set, perform real-time adjustment of the camera intrinsic parameters and extrinsic parameters, and perform online inversion and recursive optimization of the non-linear distortion parameters to establish a self-consistent correction field covering the whole time sequence, realize the alignment of the optical path system and the imaging coordinate system;

[0072] After obtaining the real edge evidence dataset with high temporal stability and spatial consistency, in order to accurately correct the imaging geometry of industrial close-range photogrammetry images, it is necessary to reconstruct the camera imaging model based on the edge dataset, dynamically adjust the internal and external parameters and optimize the nonlinear distortion effect, build an integrated correction reference with full-time coverage, and realize the accurate alignment of the light propagation path and the imaging coordinate system. The specific steps are as follows:

[0073] Based on the edge coordinates, optical flow trajectories, phase features and depth values contained in the real edge evidence dataset, the spatial back projection operation is performed on the corresponding edge region in each frame of image, and the actual projection geometry between the camera imaging center and the space point is estimated combined with the shooting parameters and the calibration reference at the actual acquisition time. This step restores the real space position of the edge point in the three-dimensional calibration space, and compares it with the pixel projection coordinates on the image plane, so as to obtain the projection offset of the imaging light path in the current frame of image. After completing the back projection of all edge points, the spatial deviation between them and the ideal projection model is calculated, and the error distribution map of the current imaging frame is constructed. The error map describes the projection offset characteristics of each point in pixels, and the imaging geometry reconstruction task is performed with the map as input, so as to establish the initial geometric structure model of the current image frame. This model is the core reference framework for the subsequent correction process.

[0074] On the basis of constructing the initial geometric structure model, the dynamic adjustment of camera internal and external parameters is performed around the error distribution of all edge points in the image frame. Specifically, the camera imaging center position, lens focal length, optical axis offset, principal point position and other parameters are taken as variables respectively, and the rotation matrix and translation vector of the camera are optimized simultaneously combined with the external pose calibration data. In this optimization process, the spatial re-projection path provided in the real edge evidence dataset is used as a constraint, and the projection error of all edge points in the image frame is minimized as a target, so as to adjust the camera internal structure and external attitude information in real time, so as to form a more accurate mapping relationship between it and the target object at the current sampling time. After this step, the internal and external parameter set of the current frame of image is taken as the calibration output of the time node, which is used for subsequent processing stage. At the same time, the change trend of each parameter in time dimension is recorded, which provides data support for subsequent recursion and interpolation.

[0075] After the preliminary adjustment of the camera parameters is completed, a nonlinear distortion compensation mechanism is further introduced to reverse model and recursively optimize the distortion phenomena in the image caused by factors such as lens distortion, uneven lens curvature, or assembly errors. In specific operations, for the edge points of each frame of image, the radial offset and tangential displacement trend are extracted by comparing the theoretical projection trajectory with the pixel position in the actual collected image, thereby constructing the nonlinear distortion offset model of the current frame. Then the distortion parameters of each frame of image are normalized in the time dimension, and the time-dependent distortion trend trajectory is constructed to establish the parameter evolution path. Combined with the adjustment results of the camera internal and external parameters in the previous stage, the distortion model is recursively inverted to correct the nonlinear relationship between the optical path and the image plane, and to eliminate the error accumulation caused by changes in lens structure or temperature drift during imaging. During the recursive optimization process, the distortion compensation results of each frame are checked for matching with the actual edge profile, and local reconstruction is performed in areas with low edge matching degree to improve the overall geometric consistency.

[0076] After the above three steps are completed, the corrected camera parameter set in each frame of image and the nonlinear distortion model are unified and fused to construct a self-consistent correction field covering the entire sampling time sequence. The correction field is indexed by time frames and records the complete parameter set applicable to each frame of image at the imaging time, including the internal parameter matrix, the external parameter pose information, the nonlinear distortion compensation model, and the edge projection reconstruction accuracy score. To improve the accuracy of subsequent multi-frame image fusion and geometric model construction, the correction field also introduces a spatial consistency evaluation mechanism to dynamically adjust the interpolation strategy and optimization weight between parameters by comparing the spatial alignment errors of the same edge structure at different times, ensuring that the changes in the imaging optical path and the actual image features remain highly consistent at different time nodes. In the image correction execution phase, all original image frames will perform fine geometric alignment processing according to the corresponding parameters provided by the correction field, so that each frame of image can be accurately mapped to its real physical position in space, ultimately realizing the integration of the imaging optical path and the coordinate reference system, and providing a unified and reliable geometric foundation for subsequent high-precision measurement, three-dimensional modeling, and error control.

[0077] S500, based on the self-consistent correction field, performing dynamic regulation and control operation, through phase conjugate polarization rotation scanning and reversible time grid rearrangement mechanism, periodically modulating the interference false peak energy distribution in reverse, continuously suppressing the phase inversion region, and constructing a closed-loop control process for interference detection and interference source suppression;

[0078] After the self-consistent correction field covering the whole time sequence is constructed, in order to further improve the adaptability of the imaging light path to the complex interference field and actively suppress the pseudo-peak effect generated by the high-reflectivity target surface, a dynamic control mechanism based on phase manipulation and time reconstruction needs to be introduced, and a closed-loop control process of interference detection and interference source suppression needs to be established to achieve accurate control and dynamic adjustment of the phase inversion region during the imaging stage. The specific steps are as follows:

[0079] Based on the imaging parameters, light path structure and distortion correction model of each frame of image recorded in the self-consistent correction field at the imaging time, the spatial alignment reference required by the dynamic control framework is constructed. In specific operation, all edge projection information, optical path distribution, camera internal and external parameters and nonlinear distortion compensation curves are jointly modeled in multiple dimensions according to the image frame index to form a light path control structure with time continuity and spatial consistency. This structure not only records the imaging geometric change trend of each pixel point in different time periods, but also marks the energy fluctuation law of the phase inversion region identified in the interference risk map in the entire time sequence. On this basis, the known interference pseudo-peak time sequence characteristics are retrieved, and the positions prone to peak enhancement under periodic modulation are identified as target control nodes to provide stable triggering conditions for subsequent phase conjugate modulation.

[0080] After the construction of the control reference structure is completed, the phase conjugate polarization rotation scanning mechanism with dynamic response capability is introduced to optically adjust the phase inversion characteristics in the high-risk region. In this step, a polarization rotation device with high-precision angle control capability is used to scan the incident light beam in the target region range, and by adjusting the polarization angle and polarization axis direction, the polarization state of the incident light is actively intervened to form a conjugate structure complementary to the original reflection path in the high-reflectivity region. At the same time, part of the signal in the reflection path is guided into the interference suppression channel to complete the shielding of local phase noise. In this process, the polarization rotation rate and rotation direction are dynamically adjusted according to the optical path change trend provided by the previous correction field to ensure that the incident light and the reflected light form a mutual cancellation state with opposite phase characteristics in the actual shooting process. This method takes the interference pseudo-peak frequency change as a reference standard, anchors the control action at the critical point where phase superposition is most likely to occur, so as to achieve the effect of source intervention on interference energy.

[0081] On the basis of the optical layer polarization phase adjustment, the reversible time grid rearrangement mechanism is introduced, and the interference intensity evolution trajectory in the image frame sequence is operated in time sequence reverse reconstruction. The operation takes the pseudo-peak area energy change curve recorded in the interference risk map as the input reference, and reorders the original time frame according to the pseudo-peak frequency and phase jump period. Specifically, the time frame interval of the high interference frequency area is weighted and shortened, so that it is more densely regulated and responded in the time sequence, while the stable frame segment maintains normal rhythm, so as to compress the accumulation period of interference peak value in time dimension. In the reverse rearrangement process, the frame overlap interpolation and space compensation are performed for the phase inversion area, the stable phase characteristics of the known low interference frame are projected into the adjacent unstable frame, and the time sequence reconstruction accuracy is verified through the comparison of optical flow consistency and depth reconstruction. The mechanism is based on the theory of time reversibility, and by controlling the time grid structure of the image sequence, the dynamic suppression of the interference propagation path is realized, and the time continuous superposition chain relied on by the pseudo-peak formation is broken.

[0082] The above phase conjugate polarization rotation adjustment and time grid reverse arrangement control mechanism are integrated to construct a closed-loop control process for interference suppression. In this process, based on the real-time parameters provided by the self-consistent correction field, each frame of image enters the acquisition stage before it is acquired, and the corresponding optical path control instruction is triggered according to its predicted interference risk level, including polarization angle preset, conjugate state configuration and time sequence fine adjustment rule loading. After the actual image acquisition is completed, according to the feedback of the gray scale field distribution and the phase response trend, it is automatically checked whether the pseudo-peak intensity of the current frame in the interference sensitive area meets the standard, and the control parameters of the next frame are adjusted combined with the response of the previous frame, forming a cycle of response-feedback-re-regulation. In this process, the interference energy is gradually weakened below the recognition threshold, the original pseudo-peak structure is continuously suppressed in space and time dimensions, and finally an imaging optical path adjustment mechanism with real-time adaptive ability is formed, which ensures that under the condition of high reflection and complex interference, the industrial close-range photogrammetry image with high geometric precision, stable texture boundary and continuous phase structure can still be output, providing reliable image basis for subsequent digital reconstruction and precise measurement.

[0083] The application realizes dynamic perception and risk warning of image interference potential by constructing anchor point sequence with time continuity and interference risk map, accurately identifies and isolates the pseudo-edge area with phase inversion characteristics by combining phase polarization consistency audit and pixel-by-pixel tracking, further removes the unstable feature points disturbed by interference under the multi-view optical flow registration mechanism, and extracts highly reliable real edge evidence. Based on the evidence dataset, imaging geometry reconstruction and distortion optimization can be realized to accurately correct the imaging parameters in the whole time sequence, establish the integrated alignment relationship between the optical path and the image. On this basis, by means of phase conjugate polarization scanning and time grid rearrangement strategy, the active regulation and periodic inhibition mechanism of interference energy is formed, and the closed-loop control process from interference detection to adaptive regulation is constructed, realizing the whole-process suppression and feedback adjustment of the interference source. The overall scheme has the synergistic advantages of dynamic perception, accurate correction and adaptive suppression, and finally realizes the accurate identification and stable elimination of the interference pseudo-peak in the industrial close-range image, providing a solid data foundation and robust image guarantee for high-precision three-dimensional reconstruction and precision vision measurement.

[0084] The embodiment provides a correction system for an industrial close-range photogrammetry image as shown in Figure 2 The embodiment provides a correction system for an industrial close-range photogrammetry image as shown in

[0085] The time-scale optical path observation module is configured to construct a cross-time-scale optical path observation baseline under multi-time-scale sampling conditions, perform joint extraction of brightness distribution and phase gradient on each frame of image, generate an anchor point sequence with time continuity, and construct an interference risk map according to the optical path difference change trend, which is used as a dynamic constraint basis for feature recognition.

[0086] The phase consistency audit module is configured to perform phase polarization consistency audit based on the anchor point sequence and the interference risk map, perform pixel-by-pixel tracking on the phase mutation area in the brightness field, extract the interference pseudo-peak with the phase inversion feature, and output the pseudo-edge candidate dataset containing the spatial position and the phase state.

[0087] The optical flow coupling registration module is configured to perform multi-view optical flow depth coupling registration based on the pseudo-edge candidate dataset and the phase audit result, remove the edge points with time sequence drift through cross-frame optical flow consistency detection, and convert the edge area with stable morphological characteristics into a real edge evidence dataset.

[0088] The geometric correction reconstruction module is configured to reconstruct the imaging geometry based on the real edge evidence dataset, perform real-time adjustment of the camera internal and external parameters, and perform online inversion and recursive optimization on the nonlinear distortion parameters, establish a self-consistent correction field covering the whole time sequence, and realize the alignment of the optical path system and the imaging coordinate system.

[0089] The interference suppression control module is configured to perform a dynamic regulation operation based on a self-consistent correction field, perform periodic reverse modulation on an interference false peak energy distribution through a reversible time grid rearrangement mechanism cooperating with a phase conjugate polarization rotation scan, continuously suppress a phase inversion region, and construct a closed-loop control process for interference detection and interference source suppression.

[0090] The correction method for the industrial close-range photogrammetry image is implemented by the correction system for the industrial close-range photogrammetry image, and the specific method and process of the correction system for the industrial close-range photogrammetry image are described in the above embodiments of the correction method for the industrial close-range photogrammetry image, which will not be described here again.

[0091] The above only describes certain exemplary embodiments of the present application in a descriptive manner, and it is needless to say that the described embodiments can be modified in various ways without departing from the spirit and scope of the present application for those skilled in the art. Therefore, the above drawings and descriptions are illustrative in nature and should not be understood as limiting the scope of protection of the claims of the present application.

Claims

1. A correction method for industrial close-range photogrammetric images, characterized in that, Includes the following steps: S100 constructs a cross-timescale optical path observation baseline under multi-timescale sampling conditions, performs joint extraction operations of brightness distribution and phase gradient on each frame of image to generate anchor point sequence, and constructs an interferometric risk map based on the trend of optical path difference change, which is used as the dynamic constraint basis for feature recognition. S200 performs phase polarization consistency audit based on anchor point sequence and interferometric risk map, performs pixel-by-pixel tracking of phase abrupt change regions in brightness field, extracts interferometric pseudo-peaks with phase reversal characteristics, and outputs pseudo-edge candidate dataset; S300 performs multi-view optical flow deep coupling registration based on pseudo edge candidate dataset and phase audit results. It eliminates edge points with temporal drift through cross-frame optical flow consistency detection and transforms edge regions with stable morphological features into real edge evidence datasets. The S400 reconstructs the imaging geometry based on a real edge evidence dataset, performs real-time adjustments to the camera's intrinsic and extrinsic parameters, and performs online inversion and recursive optimization of nonlinear distortion parameters to establish a self-consistent correction field covering the entire time series, which is used to guide the geometric alignment and optical path consistency correction of all image frames in space. The S500 performs dynamic control operations based on a self-consistent correction field. Through phase conjugate polarization rotation scanning combined with a reversible time grid rearrangement mechanism, it performs periodic reverse modulation on the energy distribution of interference pseudo-peaks, continuously suppresses the phase reversal region, and constructs a closed-loop control process for interference detection and interference source suppression. Step S200 includes: Based on the time anchor sequence and interference risk map, continuous pixel regions with high interference risk scores are selected as auditable regions within the pixel range of the image, and spatial proximity relationships with anchor points are established. The brightness change curve of each pixel in the audit area is compared with the brightness trajectory of the anchor point, and the grayscale response difference under different polarization directions is combined to identify areas with inconsistent phase. For each pixel in the phase inconsistency region, a pixel-by-pixel time series tracking is performed to extract the synchronization features of the brightness polarity reversal point and polarization response to determine whether a phase reversal phenomenon exists. During the pixel-by-pixel time series tracking, the phase reversal phenomenon is jointly determined based on the distribution frequency of the brightness polarity reversal point, the spatial jump amplitude, and the synchronization of polarization response changes. The corresponding pixel is marked as an interference pseudo-peak only when all three conditions are met. Based on the judgment results, a pseudo-edge candidate dataset containing spatial location and phase state information is output, and boundary fitting and cross-validation with time anchor points are performed to form a pseudo-peak trajectory set with a clear structure.

2. The correction method for industrial close-range photogrammetry images according to claim 1, characterized in that, Step S100 includes: An imaging structure with spatial calibration reference is established under multi-timescale sampling conditions, and image frame sequences are acquired by configuring a multispectral imaging device with adjustable focal length. Based on the acquired image frame sequence, the joint feature data of brightness distribution and phase gradient are extracted, and the local gray-level continuity and phase change trend of pixels are analyzed to construct a joint feature point dataset. Based on the constructed joint feature point dataset, regions with consistent change characteristics in the time series are selected as anchor point candidate regions, and spatial back projection verification is performed in conjunction with the optical path observation baseline to determine the time anchor point sequence; Based on the time anchor sequence and optical path change trend, grayscale fluctuation and phase oscillation analysis are performed on non-anchor regions to construct an interference risk map for dynamic constraints and record the distribution of interference risk regions in the image.

3. The correction method for industrial close-range photogrammetry images according to claim 1, characterized in that, Step S300 includes: Based on the spatial location index and phase history information of the pseudo edge candidate dataset, the corresponding regions in the multi-view image sequence are retrieved, and the observation frame sequence with temporal relationship is constructed for subsequent analysis. For each observation frame sequence, optical flow field is tracked frame by frame in the pseudo-edge region to establish pixel motion trajectories between consecutive image frames, and disparity matching is introduced to calculate the spatial depth consistency of each pixel. Based on the optical flow tracing results, cross-frame consistency detection is performed on all pseudo-edge points. Combining the phase response trajectory and the change in interference score, unstable edge points with temporal drift and spatial jump are eliminated. All edge regions that pass the consistency verification are constructed into a real edge evidence dataset, and spatial location, depth estimation and contour morphology information are extracted to provide highly reliable input for subsequent geometric reconstruction.

4. The correction method for industrial close-range photogrammetry images according to claim 3, characterized in that, In the process of constructing a real edge evidence dataset, subpixel-level boundary fitting is performed on the edge regions that pass the consistency verification, and edge morphology templates are generated to serve as edge constraint inputs in subsequent imaging geometry reconstruction.

5. The correction method for industrial close-range photogrammetry images according to claim 1, characterized in that, Step S400 includes: Based on the edge coordinates, optical flow trajectory, phase features and depth values ​​in the real edge evidence dataset, a spatial back projection operation is performed on each frame of the image to establish the projection geometry of the imaging optical path; An error distribution map is constructed based on the projection geometry, and dynamic adjustment operations are performed on the camera's intrinsic and extrinsic parameters to optimize the camera's imaging center position, focal length, optical axis offset, and external attitude information. After adjusting the camera parameters, nonlinear distortion parameters are modeled and recursively optimized based on the theoretical trajectory and actual position of the edge points to compensate for radial and tangential distortion errors. The camera parameters of each frame are fused with the distortion model to construct a self-consistent correction field covering the entire time series, which is used to guide the geometric alignment and optical path consistency correction of all image frames in space.

6. The correction method for industrial close-range photogrammetry images according to claim 5, characterized in that, In the recursive optimization process of nonlinear distortion parameters, the distortion model is locally reconstructed based on the matching error of edge contours in each frame image, and the parameter weights are dynamically adjusted using a spatial consistency evaluation mechanism to improve the geometric correction accuracy.

7. The correction method for industrial close-range photogrammetry images according to claim 1, characterized in that, Step S500 includes: Based on the imaging parameters, optical path structure and distortion correction model recorded in the self-consistent correction field, an optical path control structure with temporal continuity and spatial consistency is constructed, and a target control node with phase reversal characteristics is identified. By using a phase conjugate polarization rotation scanning mechanism, the polarization state of the incident light is adjusted within the high-reflection region to form a structure that is phase-complementary to the reflection path, thereby interfering with the phase superposition process of the interference pseudo-peaks. By reversible temporal raster rearrangement mechanism, the evolution trajectory of interference intensity in the image frame sequence is reconstructed, the inter-frame temporal structure is adjusted to weaken the accumulation of pseudo-peak energy, and a temporal dimension interference control strategy is constructed. Based on the aforementioned dual phase and time adjustment method, a closed-loop control process for interference detection and interference source suppression is constructed. The control parameters are set before image acquisition and the control command for the next frame is updated based on the response feedback after acquisition.

8. A correction system for industrial close-range photogrammetric images, used to implement the correction method for industrial close-range photogrammetric images according to any one of claims 1-7, characterized in that, include: The time-scaled optical path observation module is configured to construct a cross-time-scale optical path observation baseline under multi-time-scale sampling conditions, perform joint extraction operations of brightness distribution and phase gradient on each frame image, generate an anchor point sequence, and construct an interference risk map based on the trend of optical path difference change, which is used as the dynamic constraint basis for feature recognition. The phase consistency audit module is configured to perform phase polarization consistency audit based on anchor point sequence and interferometric risk map, track phase change regions in brightness field pixel by pixel, extract interferometric pseudo-peaks with phase reversal characteristics, and output pseudo-edge candidate dataset; The optical flow coupling registration module is configured to perform multi-view optical flow deep coupling registration based on the pseudo edge candidate dataset and phase audit results. It eliminates edge points with temporal drift through cross-frame optical flow consistency detection and transforms edge regions with stable morphological features into real edge evidence datasets. The geometric correction and reconstruction module is configured to reconstruct the imaging geometry based on the real edge evidence dataset, perform real-time adjustment of camera intrinsic and extrinsic parameters, and perform online inversion and recursive optimization of nonlinear distortion parameters to establish a self-consistent correction field covering the entire time series. The interference suppression control module is configured to perform dynamic control operations based on the self-consistent correction field. Through phase conjugate polarization rotation scanning combined with a reversible time grid rearrangement mechanism, it performs periodic reverse modulation on the energy distribution of interference pseudo-peaks, continuously suppresses the phase reversal region, and constructs a closed-loop control process for interference detection and interference source suppression.

Images