Construction node automatic checking method based on image recognition and progress plan

Abstract

The present application relates to the field of communication construction monitoring, and more particularly to a construction node automatic verification method based on image recognition and progress plan, comprising: determining the category of each verification analysis image to determine whether cross-validation analysis is performed for each verification analysis image; in cross-validation analysis, determining whether to perform positioning calibration analysis on the verification image for the anomaly according to the displacement matching error index and the flag reference error index; in positioning calibration analysis, determining whether to perform anomaly diagnosis analysis on the verification image for each anomaly according to the error correction index; in anomaly diagnosis analysis, determining the setting mode of the feature anomaly index according to the target environment matching index of each anomaly diagnosis execution image to determine whether to perform verification failure prompt on each anomaly diagnosis execution image; and outputting the verification analysis image for verification progress update or verification failure prompt respectively. The present application improves the supervision quality and efficiency of construction progress.

Description

Technical Field

[0001] This invention relates to the field of communication construction monitoring, and in particular to an automatic verification method for construction nodes based on image recognition and schedule planning. Background Technology

[0002] Image recognition of construction nodes can provide project managers with real-time progress tracking and quality verification support. To ensure the accuracy of actual supervision and verification, a large amount of construction image data collected on-site is required. This data often contains numerous construction images of varying quality, and the requirements for node feature extraction differ significantly between images of different quality levels. The feature extraction and verification strategies employed must meet the adaptability requirements of the corresponding image content quality. For example, high-quality images require refined feature mining to retain key node details, while low-quality images require robust feature extraction methods to avoid introducing excessive noise or misjudgments. Therefore, how to implement targeted intelligent verification methods based on actual image conditions to accurately identify and correct positioning drift anomalies, thereby comprehensively improving the accuracy and efficiency of construction progress supervision, is a problem that urgently needs to be solved by those skilled in the art.

[0003] Chinese Patent Application Publication No. CN117875872A discloses a method for intelligently controlling construction progress based on on-site monitoring images, a computer-readable storage medium, and a construction management system. This invention establishes a difference map between the construction drawings of two construction nodes based on the time correspondence between the construction plan and the construction drawings, serving as the plan drawing for the next construction node. At each stage of the construction plan, construction progress images of the construction site are captured from the perspective of the plan drawings. Image coverage analysis of preset equipment is performed on the plan drawings and construction progress images. The status of preset equipment in the construction progress images is determined; if the number and / or status of preset equipment does not reach the preset level, it is determined that the current construction progress is inconsistent with the construction plan, and a signal to control the construction progress is issued. However, the above technical solution has the following drawbacks: it fails to perform targeted analysis and evaluation of the acquired on-site monitoring images according to the actual situation, resulting in the risk of abnormal positioning of the acquired on-site monitoring images, thus leading to low quality and efficiency of supervision of the construction process. Summary of the Invention

[0004] To address this, the present invention provides an automatic verification method for construction nodes based on image recognition and schedule planning. This method overcomes the problem in the prior art that the acquired on-site monitoring images are not analyzed and evaluated in a targeted manner according to the actual situation, which leads to the risk of abnormal positioning of the acquired on-site monitoring images and thus results in low quality and efficiency of supervision of the construction process.

[0005] To achieve the above objectives, this invention provides an automatic verification method for construction nodes based on image recognition and schedule planning, comprising: Obtain the verification and analysis image; The category of the corresponding verification analysis image is determined based on the initial anomaly evaluation parameters of each verification analysis image, so as to determine whether cross-validation analysis should be performed on each verification analysis image. In cross-validation analysis, the displacement matching error index and the flag reference error index are used to determine whether to perform positioning calibration analysis on abnormal verification images. Through positioning calibration analysis, the error correction index of each positioning drift anomaly image is determined based on the reference deviation reference value or the flag reference error index, and the anomaly diagnosis analysis is performed on each anomaly image based on the error correction index. Through anomaly diagnosis analysis, the setting method of the characteristic anomaly index of the corresponding anomaly diagnosis execution image is determined according to the target environment matching index of each anomaly diagnosis execution image, and the verification failure prompt is determined according to the characteristic anomaly index. Output verification analysis images that update the verification progress or indicate verification failure.

[0006] Furthermore, the categories of the verification analysis images include abnormal execution verification images and update execution verification images; The abnormal execution verification image is a verification analysis image where the initial abnormal evaluation parameter is greater than the preset initial abnormal evaluation parameter; The update execution verification image is a verification analysis image where the initial anomaly evaluation parameter is less than or equal to the preset initial anomaly evaluation parameter; The initial anomaly assessment parameters are determined based on the coordinate validity index and the coverage validity index.

[0007] Furthermore, cross-validation analysis is performed on the abnormal execution verification image, including: The displacement matching error index and the flag reference error index of the abnormal execution verification image are detected. For abnormal execution verification images where the displacement matching error index is greater than the preset displacement matching error index or the flag reference error index is greater than the preset flag reference error index, the abnormal execution verification images are identified as positioning drift abnormal images. If abnormal positioning drift images are found, then positioning calibration analysis is required. The displacement matching error index is determined based on the acquisition and positioning coordinates of each time-series correlation verification image; The flag reference error index is determined based on the feature matching parameters of the reference point of the abnormal execution verification image.

[0008] Further, the positioning calibration analysis includes: For positioning drift anomaly images with an error correction index greater than the preset error correction index, they are identified as abnormal diagnostic execution images; If an abnormal diagnostic execution image is found, then an abnormal diagnostic analysis will be performed. The method for setting the error correction index is determined based on the reference valid index.

[0009] Furthermore, for positioning drift anomaly images where the reference validity index is greater than the preset reference validity index, the error correction index is determined based on the reference deviation reference value; For positioning drift anomaly images where the reference effective index is less than or equal to the preset reference effective index, the error correction index is determined based on the marker reference error index.

[0010] Furthermore, anomaly diagnostic analysis is performed on each anomaly diagnostic execution image, including: The target environment matching index of each anomaly diagnosis execution image is determined based on the feature matching parameters of each reference point, so as to determine the setting method of the feature anomaly index of the corresponding anomaly diagnosis execution image. For abnormal diagnostic images where the feature anomaly index is greater than the preset feature anomaly index, a verification failure prompt is displayed.

[0011] Furthermore, coupling interferometry analysis is performed on abnormal diagnostic execution images where the target environment matching index is greater than the preset target environment matching index. The characteristic abnormality index of the abnormal diagnostic execution image is determined based on the displacement matching error index and the signal coupling index. The characteristic anomaly index is positively correlated with the displacement matching error index and the signal coupling index, respectively.

[0012] Furthermore, the signal coupling index is determined based on the degree of correlation coupling of each localization interference target present in the anomaly diagnosis execution image and the coupling cross parameter; The signal coupling index is positively correlated with the degree of correlation coupling and the coupling cross parameter of each positioning interference target.

[0013] Furthermore, feature distortion analysis is performed on abnormal diagnostic execution images whose target environment matching index is less than or equal to the preset target environment matching index, and the feature abnormality index of the abnormal diagnostic execution image is determined based on the reflection curvature deviation index. The characteristic anomaly index and the reflection curvature deviation index are positively correlated.

[0014] Furthermore, for a single verification analysis image, which is either an image for updating the verification execution or a positioning drift anomaly image whose error correction index is less than or equal to a preset error correction index, the verification progress is updated for that verification analysis image.

[0015] Compared with existing technologies, the advantages of this invention are that the technical solution periodically defines the anomaly screening range of each verification analysis image by collecting data and initial anomaly assessment parameters, and determines the anomaly tendency corresponding to each anomaly execution verification image by combining cross-validation analysis, so as to characterize the correlation strength between positioning drift and actual installation errors. Based on this, a targeted diagnostic analysis method is determined for each anomaly execution verification image, and the effectiveness of the obtained verification analysis image for node verification is given according to the analysis results, so as to determine whether supplementary verification is needed. This avoids the problem of one-sided risk prediction and omission of hidden defects caused by isolated analysis of each construction node image in traditional solutions, and significantly improves the comprehensiveness and accuracy of node verification in complex indoor construction scenarios.

[0016] Furthermore, this invention categorizes data using displacement matching error index and flag reference error index to achieve a differentiated analysis strategy. The determined categories characterize the risk tendency of each abnormal execution verification image to be affected by positioning system drift or actual installation errors, thereby determining a targeted analysis method that balances analysis quality and efficiency and avoids full-scale complex analysis. This invention reduces the analysis redundancy in low-risk scenarios while ensuring the analysis depth of complex construction scenarios, achieving synergistic optimization of verification analysis quality and efficiency.

[0017] Furthermore, in this invention, the error correction index of each positioning drift anomaly image is periodically determined based on the reference validity index, and the error correction index of each positioning drift anomaly image is optimized and adjusted according to the reference deviation reference value or the flag reference error index. This ensures the effectiveness of the positioning calibration process for each positioning drift anomaly image, thereby ensuring the evaluation quality of subsequent anomaly diagnosis analysis and timely optimizing the verification analysis method for each anomaly execution verification image. This invention improves the node progress verification quality of each construction node image.

[0018] Furthermore, this invention performs coupling interference analysis on abnormal diagnostic execution images where the target environment matching index is greater than a preset threshold. Since these abnormal diagnostic execution images are subject to a high degree of environmental characteristics and signal interference during construction, and may exhibit positioning deviations or signal coupling anomalies, further analysis of the actual installation errors affecting these abnormal diagnostic execution images is necessary. The setting method for the feature anomaly index is determined based on the displacement matching error index and signal coupling index of each abnormal diagnostic execution image. The signal coupling index characterizes the degree of signal interference between the acquisition execution target and the positioning interference target during the target analysis cycle, i.e., the degree of clustering of the signal overlap area. This determines the targeted feature anomaly index analysis method, making the determination of the anomaly type for each abnormal diagnostic execution image more consistent with actual operational conditions. This improves the effectiveness of the evaluation results for the actual installation errors of each construction node image. This invention improves the node verification analysis efficiency of the indoor distribution system construction process.

[0019] Furthermore, this invention performs feature distortion analysis on abnormal diagnostic execution images that have completed coupling interference analysis or whose target environment matching index is less than or equal to a preset threshold. By quantitatively evaluating the reflection curvature deviation index of each abnormal diagnostic execution image in real time, the physical deformation anomalies existing in the corresponding abnormal diagnostic execution images are estimated. Based on the feature anomaly index, it is determined whether to issue a prompt for verification failure. This invention improves the node verification efficiency and early warning timeliness of the construction process of indoor distributed systems in complex construction scenarios. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the automatic verification method for construction nodes based on image recognition and schedule planning according to the present invention; Figure 2 This is a flowchart illustrating how the present invention determines the category of a verification analysis image based on the initial anomaly assessment parameters of each verification analysis image. Figure 3 This is a flowchart illustrating the process of determining whether to perform cross-validation analysis based on the category of the image being analyzed, according to the present invention. Figure 4 This is a flowchart illustrating the process of determining whether to perform anomaly diagnostic analysis based on the error correction index, as described in this invention. Detailed Implementation

[0021] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0022] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0023] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate the direction or positional relationship, are based on the direction or positional relationship shown in the drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0024] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0025] Please see Figures 1 to 4 As shown, this invention provides an automatic verification method for construction nodes based on image recognition and schedule planning, comprising: Obtain the verification and analysis image; The category of the corresponding verification analysis image is determined based on the initial anomaly evaluation parameters of each verification analysis image, so as to determine whether cross-validation analysis should be performed on each verification analysis image. In cross-validation analysis, the displacement matching error index and the flag reference error index are used to determine whether to perform positioning calibration analysis on abnormal verification images. Through positioning calibration analysis, the error correction index of each positioning drift anomaly image is determined based on the reference deviation reference value or the flag reference error index, and the anomaly diagnosis analysis is performed on each anomaly image based on the error correction index. Through anomaly diagnosis analysis, the setting method of the characteristic anomaly index of the corresponding anomaly diagnosis execution image is determined according to the target environment matching index of each anomaly diagnosis execution image, and the verification failure prompt is determined according to the characteristic anomaly index. Output verification analysis images that update the verification progress or indicate verification failure.

[0026] This invention is used to verify and analyze images acquired for automatic verification of construction nodes in complex communication construction scenarios. It determines verification failure prompts for updating verification progress or re-acquiring images, ensuring the quality and efficiency of the actual construction node verification process and avoiding image verification inefficiencies in complex communication construction scenarios. The construction area for verification is designated as the target verification area. During actual construction, images are periodically acquired from construction points within the target verification area, and these acquired images are recorded as verification analysis images. This periodic image acquisition can be periodic; for example, a cyclic verification acquisition period is used, the duration of which can be determined by the user. At the end of each verification acquisition period, the verification analysis images for each construction point are acquired. The higher the user's requirements for the verification analysis quality of construction nodes, the shorter the verification acquisition period. One verification acquisition period is provided, with a period of 5 days. This invention includes equipment related to the progress node verification during the construction process within the target verification area. Nodes are designated as construction points. The types of construction points in this invention include, but are not limited to, different models of ceiling-mounted antennas, wall-mounted antennas, and feeder terminals. To assist in determining the shooting location and verifying environmental consistency, thereby helping to calibrate positioning errors or identify environmental matching errors, several reference points are set for each construction point. For a single construction point, the existing reference points are all fixed devices within the reference space range corresponding to that construction point. The categories corresponding to the reference points include, but are not limited to, different models of lamps, fire sprinklers, ventilation openings, and smoke detectors. The reference space range is a spherical space with the coordinate point corresponding to the construction point as the center and the reference execution distance as the radius. The value of the reference execution distance can be set by the user according to the actual situation. One possible value for the reference execution distance is 0.7m. In this invention, drone equipment is used to automatically collect verification and analysis images. No specific limitations are imposed on the mobile devices used to acquire verification and analysis images at each construction point. However, when acquiring images, it is necessary to be able to automatically collect images and achieve spatial positioning. This invention includes several verification analysis records. Each verification analysis record contains at least one set of detection values ​​used in the technical solution of this invention during the cross-validation analysis of each verification analysis image, such as initial anomaly assessment parameters, displacement matching error index, and flag reference error index. Each verification analysis record also has a corresponding pass / fail mark. The pass / fail mark indicates whether the verification analysis quality and efficiency of the construction node meet the user's requirements. The pass / fail mark can be recorded manually. It is understood that the user can determine whether the verification analysis quality and efficiency of the construction node meet the requirements based on self-defined indicators.

[0027] Specifically, the categories of the verification analysis images include abnormal execution verification images and update execution verification images; The abnormal execution verification image is a verification analysis image where the initial abnormal evaluation parameter is greater than the preset initial abnormal evaluation parameter; The update execution verification image is a verification analysis image where the initial anomaly evaluation parameter is less than or equal to the preset initial anomaly evaluation parameter; The initial anomaly assessment parameters are determined based on the coordinate validity index and the coverage validity index.

[0028] In this invention, each verification and analysis image is associated with a corresponding acquisition metadata record. This metadata includes the acquisition timestamp and acquisition location coordinates of the corresponding verification and analysis image. For a single verification and analysis image, the initial anomaly evaluation parameter is the sum of the coordinate validity index and the coverage validity index of that verification and analysis image. The coordinate validity index... , This is the shortest distance between the acquisition and positioning coordinates corresponding to the verification and analysis image and the construction point corresponding to the verification and analysis image. To evaluate the distance parameter, the user can set its value according to actual needs. The higher the user's requirements for the quality and efficiency of the verification and analysis of construction nodes, the smaller the value of the evaluation distance parameter. One proposed value for the evaluation distance parameter is 7m. The coverage effectiveness index... , This represents the number of verification marker points corresponding to the construction points present within the verification analysis image. This represents the number of verification marker points corresponding to the construction points in the verification analysis image; The value of the preset initial anomaly assessment parameter can be understood as follows: the higher the user's requirements for the quality and efficiency of the verification analysis of the construction node, the smaller the value of the preset initial anomaly assessment parameter. The initial anomaly assessment index represents the degree of anomaly in the initial assessment of the verification analysis image. A method for determining the value of the preset initial anomaly assessment parameter is provided, which is the maximum value of the initial anomaly assessment parameter in the verification analysis record that meets the user's requirements for the quality and efficiency of the verification analysis of the construction node and updates the verification image.

[0029] Specifically, cross-validation analysis is performed on the abnormal execution verification image, including: The displacement matching error index and the flag reference error index of the abnormal execution verification image are detected. For abnormal execution verification images where the displacement matching error index is greater than the preset displacement matching error index or the flag reference error index is greater than the preset flag reference error index, the abnormal execution verification images are identified as positioning drift abnormal images. If abnormal positioning drift images are found, then positioning calibration analysis is required. The displacement matching error index is determined based on the acquisition and positioning coordinates of each time-series correlation verification image; The flag reference error index is determined based on the feature matching parameters of the reference point of the abnormal execution verification image.

[0030] Specifically, for a single abnormal execution verification image, since the initial abnormality assessment parameter of this image is large, it indicates that there is a risk to the monitoring and acquisition quality of the corresponding construction point. Therefore, cross-validation analysis is performed on this abnormal execution verification image to detect its displacement matching error index and marker reference error index. The displacement matching error index... , The average displacement distance parameter of each time-series correlated image combination used to verify the anomaly is calculated. The standard deviation between the displacement distance parameters of each time-series related image combination of the abnormal execution verification image is defined as follows: each time-series related image combination contains two time-series related images, and there are no other time-series related images between the timestamps of the acquisition of the two time-series related images. For a single time-series related image combination, the displacement distance parameter is the shortest moving distance between the acquisition and positioning coordinates of the two time-series related images. The displacement matching error index is used to evaluate whether there is a significant displacement change during the continuous acquisition of images within a certain time range. The time-series related image is a verification analysis image whose timestamp is within the time-series related range of the abnormal execution verification image. The time-series related range corresponds to a time range, and the midpoint of the time-series related range is the timestamp of the acquisition of the abnormal execution verification image. The duration of the time-series related range can be set by the user according to the actual situation. The higher the user's requirements for the verification analysis quality and efficiency of the construction node, the longer the duration of the time-series related range. One possible duration of the time-series related range is 2 minutes. For a single abnormal execution verification image, the flag reference error index is the average value of the feature matching parameters of each reference point corresponding to the construction point in the abnormal execution verification image. For a single reference point, the similarity between the reference point and the standard template of the fixed equipment of the corresponding category is detected. The obtained values ​​are normalized, and the values ​​determined by the normalization process are recorded as the feature matching parameters of the reference point. The similarity detection process includes: detecting the position of the reference point from the image, extracting the feature vector for matching from the detected area, comparing the extracted features with the pre-stored BIM template features to obtain the similarity. The similarity detection method differs for different categories of reference points, which is easily understood by those skilled in the art. For example, for a reference point of category 1 (lighting fixture), edge detection is used to extract the outline of the lighting fixture, the Hu invariant moments (7 moment features) of the outline are calculated, and the local binary pattern histogram within the lighting fixture area is extracted. For the Hu invariant moments, the Euclidean distance is calculated. For LBP histograms, calculate the chi-square distance. The similarity How to calculate the similarity between different categories of reference points and their pre-stored BIM template features is a topic already understood by those skilled in the art, and will not be elaborated upon here.

[0031] Specifically, the positioning calibration analysis includes: For positioning drift anomaly images with an error correction index greater than the preset error correction index, they are identified as abnormal diagnostic execution images; If an abnormal diagnostic execution image is found, then an abnormal diagnostic analysis will be performed. The method for setting the error correction index is determined based on the reference valid index.

[0032] Specifically, for a single abnormal execution verification image, if the displacement matching error index of the abnormal execution verification image is greater than a preset displacement matching error index or the flag reference error index is greater than a preset flag reference error index, it indicates that the abnormal execution verification image has a significant positioning offset. Therefore, this abnormal execution calibration image is recorded as a positioning drift abnormal image. The positioning drift degree of the positioning drift abnormal image is evaluated through positioning calibration analysis. The setting method of the targeted error correction index is determined based on the reference effective index. , To verify the number of reference points present in the image, perform a check on this anomaly. The number of reference points in the verification image whose feature matching parameters are greater than the preset flag reference error index is used to check for this anomaly. The values ​​of the preset displacement matching error index and the preset marker reference error index are understood to be such that the higher the user's requirements for the verification and analysis quality and efficiency of the construction node, the smaller the value of the preset displacement matching error index and the preset marker reference error index. The displacement matching error index characterizes the degree of difference in displacement change of the abnormal execution verification image within a certain time sequence, and the marker reference error index characterizes the overall abnormality of the reference objects present in the abnormal execution verification image. A method for determining the value of the preset displacement matching error index is provided, which is the average value of the displacement matching error index of the positioning drift abnormal image in the verification and analysis record that meets the user's requirements for the verification and analysis quality and efficiency of the construction node. A method for determining the value of the preset marker reference error index is also provided, which is the average value of the marker reference error index of the positioning drift abnormal image in the verification and analysis record that meets the user's requirements for the verification and analysis quality and efficiency of the construction node.

[0033] Specifically, for positioning drift anomaly images where the reference validity index is greater than the preset reference validity index, the error correction index is determined based on the reference deviation reference value; For positioning drift anomaly images where the reference effective index is less than or equal to the preset reference effective index, the error correction index is determined based on the marker reference error index.

[0034] Specifically, for a single positioning drift anomaly image, if the reference validity index of the positioning drift anomaly image is greater than the preset reference validity index, it indicates that the richness of reference points for the construction point corresponding to the positioning drift anomaly image is relatively high. The degree of error is assessed based on the actual situation of the reference points. The error correction index is a value obtained by normalizing the reference deviation reference value. The reference deviation reference value is the average of the distance deviation reference values ​​of all reference points in the positioning drift anomaly image. For a single reference point... , This refers to the shortest distance between the reference point and the corresponding construction point, determined based on the location drift anomaly image. The standard shortest interval distance between the reference point and the corresponding construction point; for a single positioning drift anomaly image, if the reference validity index of the positioning drift anomaly image is less than or equal to the preset reference validity index, it indicates that the richness of reference points for the construction point corresponding to the positioning drift anomaly image is low. The error correction index is the value obtained by normalizing the mark reference error index. The value of the preset reference validity index can be understood as follows: the higher the user's requirements for the verification and analysis quality and efficiency of the construction node, the larger the value of the preset reference validity index. The reference priority index represents the richness of reference points existing at the construction point corresponding to the positioning drift anomaly image. A method for determining the value of the preset reference validity index is provided, which is the average value of the reference validity index of the positioning drift anomaly image in the verification and analysis record that meets the user's requirements for the verification and analysis quality and efficiency of the construction node, and whose error correction index is determined based on the reference deviation reference value.

[0035] Specifically, anomaly diagnosis analysis is performed on each anomaly diagnosis execution image, including: The target environment matching index of each anomaly diagnosis execution image is determined based on the feature matching parameters of each reference point, so as to determine the setting method of the feature anomaly index of the corresponding anomaly diagnosis execution image. For abnormal diagnostic images where the feature anomaly index is greater than the preset feature anomaly index, a verification failure prompt is displayed.

[0036] Among them, the potential risk of abnormal diagnostic execution images is evaluated through abnormal diagnostic analysis. Specifically, the focus of the analysis of targeted feature abnormality indices is determined based on the target environment matching index to ensure the effectiveness of the determination results of feature abnormality indices. For a single abnormal diagnostic execution image, the target environment matching index , To diagnose this anomaly, the average value of the feature matching parameters of each reference point corresponding to the construction point in the image is calculated. The standard deviation between the feature matching parameters of each reference point corresponding to the construction point in the execution image for the anomaly diagnosis is determined. If the standard deviation between the feature matching parameters of each reference point corresponding to the construction point in the execution image for the anomaly diagnosis is 0, then it is determined that a coupling interference analysis is to be performed. For a single verification analysis image, if the verification analysis image is an abnormal diagnostic execution image with a feature anomaly index greater than the preset feature anomaly index, it indicates that the verification analysis image has a large risk of positioning drift or has obvious abnormal features. At this time, a verification failure prompt is issued for the verification analysis image, and the construction point corresponding to the acquisition positioning coordinates of the verification analysis image is marked as abnormal acquisition, and repeated acquisition and verification are required. The values ​​of the preset feature anomaly index and the preset error correction index are understood to be such that the higher the user's requirements for the verification and analysis quality and efficiency of the construction node, the smaller the value of the preset feature anomaly index. The feature anomaly index characterizes the risk of positioning drift caused by positioning signal coupling interference and the magnitude of the abnormal feature risk in the verification and analysis image. The error correction index characterizes the degree of positioning drift during image acquisition. A method for determining the value of the preset feature anomaly index is provided, where the minimum value of the feature anomaly index of the anomaly diagnosis execution image in the verification and analysis record that meets the user's requirements for the verification and analysis quality and efficiency of the construction node is recorded as the preset feature anomaly index. A feature anomaly index is defined, and a method for determining the preset error correction index is provided. If the error correction index is determined based on the reference deviation value, then the error correction index determined based on the reference deviation value in the positioning drift anomaly image that indicates verification failure in the verification analysis record, which meets the requirements of verification analysis quality and efficiency, is obtained, and the average value is calculated and recorded as the preset error correction index. If the error correction index is determined based on the marker reference error index, then the error correction index determined based on the marker reference error index in the positioning drift anomaly image that indicates verification failure in the verification analysis record, which meets the requirements of verification analysis quality and efficiency, is obtained, and the average value is calculated and recorded as the preset error correction index.

[0037] Specifically, coupling interferometry analysis is performed on abnormal diagnostic execution images where the target environment matching index is greater than the preset target environment matching index, and the characteristic abnormality index of the abnormal diagnostic execution image is determined based on the displacement matching error index and the signal coupling index. The characteristic anomaly index is positively correlated with the displacement matching error index and the signal coupling index, respectively.

[0038] Specifically, the signal coupling index is determined based on the degree of correlation coupling of each localization interference target present in the anomaly diagnosis execution image and the coupling cross parameter; The signal coupling index is positively correlated with the degree of correlation coupling and the coupling cross parameter of each positioning interference target.

[0039] Specifically, for a single abnormal diagnostic execution image, if the target environment matching index of the abnormal diagnostic execution image is less than or equal to the preset target environment matching index, it indicates that the difference between the abnormal situations of the reference points in the abnormal diagnostic execution image is low, and the overall image acquisition quality is relatively well matched. In this case, feature distortion analysis is performed on the abnormal diagnostic execution image to analyze the degree of interference encountered during the current positioning process and assess the potential positioning drift risk. The feature abnormality index of the abnormal diagnostic execution image is the value obtained after normalizing the sum of the displacement matching error index and the signal coupling index. In this invention, the maximum-minimum method is used for normalization, and the value range is determined to be [0,1]. The signal coupling index is the sum of the products of the correlation coupling degree of each positioning interference target in the abnormal diagnostic execution image and its corresponding evaluation weight coefficient. The evaluation weight coefficient corresponding to each positioning interference target is positively correlated with the coupling cross parameter, and the sum of the weight coefficients corresponding to each positioning interference target is 1. For a single positioning interference target, the correlation coupling degree... , The area of ​​the overlapping region between the positioning interference target and the acquisition execution target of the anomaly diagnosis execution image at the signal interference range corresponding to the acquisition timestamp of the anomaly diagnosis execution image. The acquisition execution target for the abnormal diagnosis execution image is the area of ​​the signal interference range corresponding to the acquisition timestamp of the abnormal diagnosis execution image. The acquisition execution target is the device that acquires the abnormal diagnosis execution image. The positioning interference target is the image acquisition device whose corresponding signal interference range overlaps with the signal interference range corresponding to the acquisition execution target. The signal interference range is a circular area with the position corresponding to the acquisition timestamp of the abnormal diagnosis execution image as the center and the signal range distance as the radius. The coupling cross parameter is the number of positioning interference targets whose signal interference range corresponding to the acquisition timestamp of the abnormal diagnosis execution image overlaps with the signal interference range corresponding to the positioning interference target. The value of the signal range distance is determined according to the signal transmission power in the actual scenario. This is a content that is easy for those skilled in the art to understand and will not be elaborated here. The value of the preset target environment matching index can be understood as follows: the higher the user's requirements for the verification and analysis quality and efficiency of the construction node, the smaller the value of the preset target environment matching index. The target environment matching index characterizes the degree of difference between the abnormal situations existing in the reference points of the abnormal diagnosis execution image. A method for determining the value of the preset target environment matching index is provided, which is the average value of the target environment matching index of the abnormal diagnosis execution image determined by the displacement matching error index and the signal coupling index in the verification and analysis record that meets the user's requirements for the verification and analysis quality and efficiency of the construction node.

[0040] Specifically, feature distortion analysis is performed on abnormal diagnostic execution images whose target environment matching index is less than or equal to the preset target environment matching index, and the feature abnormality index of the abnormal diagnostic execution image is determined based on the reflection curvature deviation index. The characteristic anomaly index and the reflection curvature deviation index are positively correlated.

[0041] Specifically, for a single abnormal diagnostic execution image, if the target environment matching index of the abnormal diagnostic execution image is less than or equal to the preset target environment matching index, it indicates that the difference between the abnormalities of the reference points in the abnormal diagnostic execution image is low, and the overall image acquisition quality is abnormal. Therefore, feature distortion analysis is performed on the abnormal diagnostic execution image to determine whether to re-acquire images for the construction points corresponding to the abnormal diagnostic execution image by analyzing its hidden defects. The feature anomaly index of the abnormal diagnostic execution image is a value obtained after normalizing the reflection curvature deviation index. , The curvature distribution vector of the construction point corresponding to the image is used to diagnose this anomaly. To diagnose this anomaly, the process of determining the curvature distribution vector of an ideal 3D model of the same type of equipment at the construction site corresponding to the image in the execution image includes: reconstructing a 3D point cloud for the construction site, calculating the local Gaussian curvature or average curvature on the surface point cloud, wherein a quadratic surface is fitted to the neighborhood of each point, and the principal curvature is calculated. Gaussian curvature mean curvature The curvature values ​​are discretized into histograms to obtain curvature distribution vectors. The process of determining the curvature distribution vector of the construction point corresponding to the anomaly diagnosis execution image is based on the anomaly diagnosis execution image. The process of determining the curvature distribution vector of the ideal three-dimensional model is based on the ideal three-dimensional model of the corresponding equipment.

[0042] Specifically, for a single verification analysis image, the verification progress is updated when the verification analysis image is an updated verification image or when the error correction index is less than or equal to the preset error correction index.

[0043] For a single verification analysis image, if the verification analysis image is an updated verification image, or if the error correction index is less than or equal to the preset error correction index, it indicates that the acquisition quality of the verification analysis image is high or the degree of correction error is low. In this case, the verification progress is updated for the verification analysis image, and the construction point corresponding to the acquisition positioning coordinates of the verification analysis image is marked as a valid acquisition, which can be directly used for verification analysis without repeated acquisition and verification.

[0044] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A method for automatic verification of construction milestones based on image recognition and schedule planning, characterized in that, include: Obtain the verification and analysis image; The category of the corresponding verification analysis image is determined based on the initial anomaly evaluation parameters of each verification analysis image, so as to determine whether cross-validation analysis should be performed on each verification analysis image. In cross-validation analysis, the displacement matching error index and the flag reference error index are used to determine whether to perform positioning calibration analysis on abnormal verification images. Through positioning calibration analysis, the error correction index of each positioning drift anomaly image is determined based on the reference deviation reference value or the flag reference error index, and the anomaly diagnosis analysis is performed on each anomaly image based on the error correction index. Through anomaly diagnosis analysis, the setting method of the characteristic anomaly index of the corresponding anomaly diagnosis execution image is determined according to the target environment matching index of each anomaly diagnosis execution image, and the verification failure prompt is determined according to the characteristic anomaly index. Output verification analysis images that update the verification progress or indicate verification failure.

2. The automatic verification method for construction nodes based on image recognition and schedule planning according to claim 1, characterized in that, The categories of the verification analysis images include abnormal execution verification images and update execution verification images; The abnormal execution verification image is a verification analysis image where the initial abnormal evaluation parameter is greater than the preset initial abnormal evaluation parameter; The update execution verification image is a verification analysis image where the initial anomaly evaluation parameter is less than or equal to the preset initial anomaly evaluation parameter; The initial anomaly assessment parameters are determined based on the coordinate validity index and the coverage validity index.

3. The automatic verification method for construction nodes based on image recognition and schedule planning according to claim 2, characterized in that, Cross-validation analysis is performed on the abnormal execution verification image, including: The displacement matching error index and the flag reference error index of the abnormal execution verification image are detected. For abnormal execution verification images where the displacement matching error index is greater than the preset displacement matching error index or the flag reference error index is greater than the preset flag reference error index, the abnormal execution verification images are identified as positioning drift abnormal images. If abnormal positioning drift images are found, then positioning calibration analysis is required. The displacement matching error index is determined based on the acquisition and positioning coordinates of each time-series correlation verification image; The flag reference error index is determined based on the feature matching parameters of the reference point of the abnormal execution verification image.

4. The automatic verification method for construction nodes based on image recognition and schedule planning according to claim 3, characterized in that, Positioning calibration analysis, including: For positioning drift anomaly images with an error correction index greater than the preset error correction index, they are identified as abnormal diagnostic execution images; If an abnormal diagnostic execution image is found, then an abnormal diagnostic analysis will be performed. The method for setting the error correction index is determined based on the reference valid index.

5. The automatic verification method for construction nodes based on image recognition and schedule planning according to claim 4, characterized in that, For positioning drift anomalies where the reference validity index is greater than the preset reference validity index, the error correction index is determined based on the reference deviation reference value. For positioning drift anomaly images where the reference effective index is less than or equal to the preset reference effective index, the error correction index is determined based on the marker reference error index.

6. The automatic verification method for construction nodes based on image recognition and schedule planning according to claim 5, characterized in that, Anomaly diagnostic analysis is performed on each image for each anomaly diagnosis, including: The target environment matching index of each anomaly diagnosis execution image is determined based on the feature matching parameters of each reference point, so as to determine the setting method of the feature anomaly index of the corresponding anomaly diagnosis execution image. For abnormal diagnostic images where the feature anomaly index is greater than the preset feature anomaly index, a verification failure prompt is displayed.

7. The method for automatic verification of construction nodes based on image recognition and schedule planning according to claim 6, characterized in that, For abnormal diagnostic execution images with a target environment matching index greater than a preset target environment matching index, coupled interferometry analysis is performed, and the characteristic abnormality index of the abnormal diagnostic execution image is determined based on the displacement matching error index and the signal coupling index. The characteristic anomaly index is positively correlated with the displacement matching error index and the signal coupling index, respectively.

8. The method for automatic verification of construction nodes based on image recognition and schedule planning according to claim 7, characterized in that, The signal coupling index is determined based on the degree of correlation coupling of each localization interference target present in the abnormal diagnosis execution image and the coupling cross parameter; The signal coupling index is positively correlated with the degree of correlation coupling and the coupling cross parameter of each positioning interference target.

9. The method for automatic verification of construction nodes based on image recognition and schedule planning according to claim 8, characterized in that, For abnormal diagnostic execution images with a target environment matching index less than or equal to a preset target environment matching index, feature distortion analysis is performed, and the feature abnormality index of the abnormal diagnostic execution image is determined based on the reflection curvature deviation index. The characteristic anomaly index and the reflection curvature deviation index are positively correlated.

10. The method for automatic verification of construction nodes based on image recognition and schedule planning according to claim 9, characterized in that, For a single verification analysis image, if the verification analysis image is an image to be updated or a positioning drift anomaly image with an error correction index less than or equal to a preset error correction index, the verification progress is updated for that verification analysis image.

Images