Method and system for realizing abnormality recognition in building construction based on machine vision

Through the machine vision-based abnormal identification method, the problem of low recognition efficiency caused by complex background interference and lighting changes in the construction site is solved, real-time monitoring and abnormal identification are realized, and construction quality and safety are improved.

CN117893779BActive Publication Date: 2024-11-26广东筠诚建筑科技有限公司 +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410082553.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-19
Publication Date
2024-11-26
Estimated Expiration
2044-01-19

AI Technical Summary

Technical Problem

Due to complex background interference and lighting changes in construction sites, it is difficult for the prior art to effectively identify abnormal situations, resulting in a decrease in identification efficiency.

Method used

Anomaly recognition model is constructed by acquiring site image sets, data pre-notation, denoising, enhancement, segmentation, feature extraction and model construction, and deploying it on the construction site to achieve real-time abnormal recognition.

Benefits of technology

It improves the abnormal identification efficiency at the construction site, can promptly discover and solve potential problems, and improves construction quality and safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117893779B_ABST
    Figure CN117893779B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of building construction, and proposes a method and system for realizing abnormality recognition under building construction based on machine vision. The method comprises: obtaining a set of on-site images of a building construction site, obtaining image labeling data by recognizing image data and pre-labeling, then performing denoising processing on the labeling data to obtain denoised image data, performing image enhancement processing to obtain enhanced image data, then performing image segmentation on the enhanced image data to obtain segmented image data, dividing the segmented image data into a data training set, extracting image features, identifying feature vectors and calculating feature similarity, thereby constructing an abnormality recognition model, performing environmental parameter abnormality recognition by using the trained model, extracting abnormal factors and calculating abnormal indicators, and generating early warning signals to realize abnormality recognition of the building construction site. The present invention can improve the efficiency of abnormality recognition in building construction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of building construction, and in particular to a method and system for realizing abnormality recognition in building construction based on machine vision. Background Art

[0002] Machine vision refers to the ability of a computer system to collect image or video data through a camera or other sensor, and use image processing, pattern recognition and other technologies to parse and understand this data, thereby achieving automatic analysis, identification and understanding of objects, scenes, actions, etc. in the image or video.

[0003] At present, the commonly used method for realizing abnormality recognition in construction is to use image recognition technology combined with the monitoring system of the construction site to monitor various activities and objects in the construction process in real time. The monitoring system will automatically analyze the image data and use deep learning algorithms to identify abnormalities. However, there are many types of abnormal situations in the construction site, such as dangerous behaviors of workers wearing safety equipment on site, equipment failures, etc., which are easy to reduce the recognition efficiency in the construction site due to complex background interference or changes in lighting. Therefore, a method and system for realizing abnormality recognition in construction based on machine vision is needed to improve the efficiency of abnormality recognition in construction. Summary of the invention

[0004] The present invention provides a method and system for realizing abnormality recognition in building construction based on machine vision, the main purpose of which is to improve the efficiency of abnormality recognition in building construction.

[0005] To achieve the above-mentioned purpose, the present invention provides a method for realizing abnormality recognition in building construction based on machine vision, comprising:

[0006] Acquire a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-label the image data to obtain image label data;

[0007] Performing image denoising on the image label data to obtain denoised image data, performing image enhancement on the denoised image data to obtain enhanced image data, and performing image segmentation on the enhanced image data to obtain segmented image data;

[0008] Dividing the segmented image data into a data training set, and performing feature extraction on the segmented image data to obtain image features, identifying feature vectors corresponding to the image features, calculating feature similarities corresponding to the segmented image data based on the feature vectors, and constructing an abnormality recognition model corresponding to the scene image set based on the feature similarities;

[0009] Performing data training on the anomaly recognition model using the data training set to obtain a trained anomaly recognition model, and deploying the tested anomaly recognition model to an environment corresponding to the construction site to obtain environmental parameters;

[0010] Anomalies are identified on the environmental parameters to obtain environmental anomaly parameters, environmental anomaly factors in the environmental anomaly parameters are extracted, environmental anomaly indicators corresponding to the environmental anomaly factors are calculated, and based on the environmental anomaly indicators, an early warning signal corresponding to the construction site is generated to realize abnormality identification in the construction site.

[0011] Optionally, acquiring a set of on-site images in the construction site and identifying image data corresponding to the set of on-site images includes:

[0012] Determining the corresponding photographing equipment for the construction site;

[0013] Acquiring a set of on-site images of a construction site using the photographing device;

[0014] Storing the on-site image set in a preset cloud storage space;

[0015] Performing data preprocessing on the data stored in the cloud storage space to obtain preprocessed data;

[0016] Corresponding image data in the preprocessed data is identified.

[0017] Optionally, the pre-labeling the image data to obtain image labeled data includes:

[0018] Performing data adjustment on the image data to obtain adjusted image data;

[0019] Defining a labeling category corresponding to the adjusted image data;

[0020] Based on the annotation category, pre-annotate the adjusted image data to obtain initially-annotated image data;

[0021] The initially marked image data is annotated and corrected to obtain image marked data.

[0022] Optionally, performing image denoising on the image marker data to obtain denoised image data includes:

[0023] Importing the image marking data, and identifying the noise type corresponding to the image marking data;

[0024] Based on the noise type, performing noise statistics on the image marking data to obtain noise statistical data;

[0025] The noise statistical data is subjected to frequency domain denoising to obtain denoised image data.

[0026] Optionally, dividing the segmented image data into a data training set and performing feature extraction on the segmented image data to obtain image features includes:

[0027] Proportionally dividing the segmented image data to obtain a data training set corresponding to the segmented image data;

[0028] Normalizing the data training set to obtain normalized data;

[0029] and determining a feature type corresponding to the segmented image data;

[0030] Based on the feature type, feature extraction is performed on the segmented image data to obtain image features.

[0031] Optionally, the calculating the feature similarity corresponding to the segmented image data based on the feature vector includes: segmenting the feature vector into a first feature vector and a second feature vector;

[0032] The feature similarity corresponding to the image feature is calculated using the following formula:

[0033]

[0034] Among them, CS represents the feature similarity corresponding to the image feature, and its value range is between [-1,1]. The closer the value is to 1, the more similar it is, and the closer the value is to -1, the less similar it is. Ai represents the corresponding i-th element in the first feature vector, and Bi represents the corresponding i-th element in the second feature vector.

[0035] Optionally, the step of deploying the tested anomaly recognition model to an environment corresponding to the construction site to obtain environmental parameters includes:

[0036] Determining a development environment corresponding to the building construction site;

[0037] Based on the development environment, deploy the model service corresponding to the tested anomaly recognition model;

[0038] Based on the model service, the environment of the construction site is identified to obtain environmental parameters.

[0039] Optionally, the performing abnormality identification on the environmental parameter to obtain the abnormal environmental parameter includes:

[0040] Identify an environmental parameter list corresponding to the environmental parameter; determine a real-time value in the environmental parameter list;

[0041] Setting a threshold range corresponding to the real-time value;

[0042] Based on the threshold range, the environmental parameter is identified as abnormal to obtain an environmental abnormality parameter.

[0043] Optionally, the calculating the environmental anomaly index corresponding to the environmental anomaly factor includes:

[0044] The environmental anomaly index corresponding to the environmental anomaly factor is calculated using the following formula:

[0045]

[0046] Among them, Z represents the environmental abnormality index corresponding to the environmental abnormality factor, Fi represents the i-th environmental abnormality factor, Fj represents the j-th environmental abnormality factor, n represents the number of environmental abnormality factors, and m represents the total number of environmental abnormality factors.

[0047] In order to solve the above problems, the present invention also provides a system for realizing abnormality recognition in building construction based on machine vision, the system comprising:

[0048] An image marking module is used to obtain a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-mark the image data to obtain image marking data;

[0049] An image segmentation module, used for performing image denoising on the image tag data to obtain denoised image data, performing image enhancement on the denoised image data to obtain enhanced image data, and performing image segmentation on the enhanced image data to obtain segmented image data;

[0050] A model building module, used to divide the segmented image data into a data training set, and perform feature extraction on the segmented image data to obtain image features, identify feature vectors corresponding to the image features, calculate feature similarities corresponding to the segmented image data based on the feature vectors, and build an abnormality recognition model corresponding to the scene image set based on the feature similarities;

[0051] An environmental parameter module is used to perform data training on the anomaly recognition model using the data training set to obtain a trained anomaly recognition model, and deploy the tested anomaly recognition model to the environment corresponding to the construction site to obtain environmental parameters;

[0052] The indicator calculation module is used to identify abnormalities in the environmental parameters, obtain environmental abnormality parameters, extract environmental abnormality factors in the environmental abnormality parameters, calculate environmental abnormality indicators corresponding to the environmental abnormality factors, and generate early warning signals corresponding to the construction site based on the environmental abnormality indicators to realize abnormality identification in the construction site.

[0053] The present invention can monitor the progress of the construction site in real time by acquiring a set of on-site images in the construction site and identifying the image data corresponding to the set of on-site images, which is helpful for real-time monitoring, improving quality, ensuring safety, improving efficiency and accumulating experience, and can improve construction quality and efficiency. The present invention can improve image quality, improve image analysis and processing results, protect image information privacy, and reduce data storage and transmission costs by performing image denoising on the image labeling data, thereby improving the effect and efficiency of image application. The present invention can improve the performance and generalization ability of the model and effectively evaluate the model by dividing the segmented image data into a data training set and performing feature extraction on the segmented image data to obtain image features. The present invention can obtain useful image features for subsequent image analysis and application. The present invention can obtain a trained anomaly recognition model by using the data training set to perform data training on the anomaly recognition model, which can help better understand how the model makes anomaly detection decisions, and can analyze the output of the model, feature importance, etc., so as to explain the operation process of the model and the basis for abnormal judgment. The present invention generates a warning signal corresponding to the construction site based on the environmental anomaly index to realize anomaly recognition in the construction site, and can identify abnormal situations in the construction site, and can timely discover and solve potential environmental anomaly problems to prevent the problem from further deteriorating, thereby improving the safety and reliability of the construction site. Therefore, the present invention proposes a method and system for realizing anomaly recognition under construction based on machine vision to improve the efficiency of anomaly recognition in construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A schematic diagram of a flow chart of a method for realizing abnormality recognition in building construction based on machine vision provided by an embodiment of the present invention;

[0055] Figure 2 A schematic diagram of a module of a machine vision-based abnormality recognition system for building construction provided by an embodiment of the present invention;

[0056] Figure 3 A schematic diagram of the internal structure of an electronic device for implementing an abnormality identification method under construction based on machine vision provided by one embodiment of the present invention.

[0057] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0058] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0059] The embodiment of the present application provides a method for identifying anomalies in construction based on machine vision. The execution subject of the method for identifying anomalies in construction based on machine vision includes but is not limited to at least one of the electronic devices that can be configured to execute the method provided in the embodiment of the present application, such as a server, a terminal, etc. In other words, the method for identifying anomalies in construction based on machine vision can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0060] Reference Figure 1 FIG. 1 is a flow chart of a method for realizing abnormality identification under construction based on machine vision according to an embodiment of the present invention. In this embodiment, the method for realizing abnormality identification under construction based on machine vision includes:

[0061] S1. Acquire a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-label the image data to obtain image label data.

[0062] The present invention can monitor the progress of the construction site in real time by acquiring a field image set in the construction site and identifying image data corresponding to the field image set, which is helpful for real-time monitoring, improving quality, ensuring safety, improving efficiency and accumulating experience, and can improve construction quality and efficiency.

[0063] The on-site image set refers to a set of images actually captured on-site; the image data refers to the digital information contained in the image, such as pixel value, color value, brightness value, contrast and other information.

[0064] As an embodiment of the present invention, the method of acquiring a field image set in a construction site and identifying image data corresponding to the field image set includes: determining a shooting device corresponding to the construction site; using the shooting device to acquire a field image set in the construction site; storing the field image set in a preset cloud storage space; performing data preprocessing on the data stored in the cloud storage space to obtain preprocessed data; and identifying the corresponding image data in the preprocessed data.

[0065] Among them, the shooting equipment refers to the shooting equipment used to shoot images of construction sites, such as: drones, smart cameras, smart surveillance cameras and other equipment; the preset cloud storage space refers to the cloud storage space set up in advance for storing images of construction sites, such as: cloud disk, cloud server, cloud storage service, etc.; the pre-processed data refers to the processing of image data stored in the cloud storage space to improve the accuracy and availability of subsequent image data.

[0066] Furthermore, the shooting device can be obtained through a machine learning algorithm, such as a decision tree algorithm, a random forest algorithm, a neural network algorithm and the like; the preset cloud storage space can be obtained through a distributed storage system, such as Hadoop, Ceph and other systems; the pre-processed data can be obtained through a data cleaning tool, such as OpenRefine, Trifacta Wrangler, Pandas and other tools.

[0067] The present invention provides data pre-labeling for the image data to obtain image labeled data, which can greatly simplify and accelerate the labeling process, improve the consistency and accuracy of the labeling, and provide better labeled samples for subsequent model training, thereby improving the efficiency and quality of the entire data processing process.

[0068] The image marking data refers to the data obtained by marking or annotating the image data.

[0069] As an embodiment of the present invention, the method of pre-annotating the image data to obtain image labeled data includes: performing data adjustment on the image data to obtain adjusted image data; defining a labeling category corresponding to the adjusted image data; based on the labeling category, pre-annotating the adjusted image data to obtain preliminary labeled image data; and correcting the labeling of the preliminary labeled image data to obtain image labeled data.

[0070] Among them, the adjusted image data refers to a series of processing and operations on the original image to make it more suitable for annotation and analysis; the labeled category refers to dividing the adjusted image data into different categories or labels; the initially labeled image data refers to the preliminary annotation of the adjusted image data using an automated method or algorithm.

[0071] Furthermore, the adjusted image data can be obtained through image processing tools, such as OpenCV, PIL, ImageJ and other tools; the labeled category can be obtained through labeling tools, such as LabelImg, Labelbox, RectLabel and other tools; the initial labeled image data can be obtained through a target detection model, such as Faster R-CNN.

[0072] S2. Perform image denoising on the image labeling data to obtain denoised image data, perform image enhancement on the denoised image data to obtain enhanced image data, and perform image segmentation on the enhanced image data to obtain segmented image data.

[0073] The present invention performs image denoising on the image marking data to obtain denoised image data, which can improve image quality, improve image analysis and processing results, protect image information privacy, and reduce data storage and transmission costs, thereby improving the effect and efficiency of image applications.

[0074] The denoised image data refers to clear image data obtained by removing noise from an image.

[0075] As an embodiment of the present invention, the image denoising is performed on the image marked data to obtain denoised image data, including: importing the image marked data and identifying the noise type corresponding to the image marked data; based on the noise type, performing noise statistics on the image marked data to obtain noise statistical data; performing frequency domain denoising on the noise statistical data to obtain denoised image data.

[0076] Among them, the noise type refers to the noise type corresponding to the image marking data, such as: Gaussian noise, salt and pepper noise, brightness noise, etc.; the noise statistical data refers to the data obtained by performing noise statistical analysis on the image marking data, such as: noise standard deviation, noise distribution pattern and other information.

[0077] Furthermore, the noise type can be obtained through digital image analysis tools, such as ImageJ, MATLAB and other tools; the noise statistics can be obtained through statistical distribution models, such as Gaussian distribution model, Poisson distribution model and other models.

[0078] The present invention performs image enhancement on the denoised image data to obtain enhanced image data, which can improve image quality, enhance visual effects, enhance features, provide a better data basis for subsequent processing, and facilitate information extraction and analysis.

[0079] The enhanced image data refers to image data obtained by enhancing image quality and specific features. Optionally, the enhanced image data can be obtained through an image restoration algorithm, such as least squares method, total variation regularization, non-local mean and other algorithms.

[0080] The present invention performs image segmentation on the enhanced image data to obtain segmented image data, which can separate different objects or areas in the image, thereby realizing object recognition and positioning, helping to understand the content in the image, and further perform scene understanding, thereby improving the effects of virtual reality experience and augmented reality applications.

[0081] The segmented image data refers to result data obtained after performing an image segmentation operation on the enhanced image data. Optionally, the segmented image data can be obtained through a segmentation algorithm, such as threshold segmentation, edge detection, region growing and other algorithms.

[0082] S3. Divide the segmented image data into a data training set, and perform feature extraction on the segmented image data to obtain image features, identify feature vectors corresponding to the image features, calculate feature similarities corresponding to the segmented image data based on the feature vectors, and construct an abnormality recognition model corresponding to the on-site image set based on the feature similarities.

[0083] The present invention divides the segmented image data into a data training set and performs feature extraction on the segmented image data to obtain image features, thereby improving the performance and generalization ability of the model, effectively evaluating the performance of the model, and obtaining useful image features for subsequent image analysis and application.

[0084] The data training set refers to a subset divided from the entire segmented image data for training the model; the image feature refers to important information extracted from the segmented image data that describes the image content and structure.

[0085] As an embodiment of the present invention, the segmented image data is divided into a data training set, and feature extraction is performed on the segmented image data to obtain image features, including: proportionally dividing the segmented image data to obtain a data training set corresponding to the segmented image data; normalizing the data training set to obtain normalized data; and determining the feature type corresponding to the segmented image data; and based on the feature type, extracting features from the segmented image data to obtain image features.

[0086] Among them, the normalized data refers to the process of scaling the data and converting it into a uniform numerical range; the feature type refers to the feature type that needs to be paid attention to and extracted in the image according to task requirements.

[0087] Furthermore, the normalized data can be obtained through a normalization algorithm, such as Min-Max Scaling, Z-score standardization and other algorithms; the feature type can be obtained through a feature extraction method, such as GLCM, LBP and other methods.

[0088] By identifying the feature vectors corresponding to the image features, the present invention can be optimized according to specific problems, improve computing efficiency, remove unnecessary computing steps, and reduce waste of time and resources, thereby better understanding the relationship between image features and targets.

[0089] The feature vector refers to a set of numerical representations extracted from an image by a feature extraction algorithm. Optionally, the feature vector can be obtained by a vector recognition tool, such as OpenCV, TensorFlow, and other tools.

[0090] As an embodiment of the present invention, the calculating the feature similarity corresponding to the segmented image data based on the feature vector includes: segmenting the feature vector into a first feature vector and a second feature vector;

[0091] The feature similarity corresponding to the image feature is calculated using the following formula:

[0092]

[0093] Among them, CS represents the feature similarity corresponding to the image feature, and its value range is between [-1,1]. The closer the value is to 1, the more similar it is, and the closer the value is to -1, the less similar it is. Ai represents the corresponding i-th element in the first feature vector, and Bi represents the corresponding i-th element in the second feature vector.

[0094] Based on the feature similarity, the present invention constructs an abnormality recognition model corresponding to the on-site image set, and can understand the causes and characteristics of the abnormality through visual analysis. This helps to further analyze the abnormal samples and take corresponding measures to deal with them, thereby improving the abnormality detection and monitoring tasks of the on-site image set.

[0095] Among them, the abnormality recognition model refers to a recognition model used to automatically identify abnormal samples in the scene image set. Optionally, the abnormality recognition model can be obtained by implementing the Python programming language, such as: scikit-learn, TensorFlow, PyTorch, etc.

[0096] In detail, the above content, combined with image data analysis, provides a new method for construction progress monitoring. Through feature extraction and similarity calculation, the construction progress can be evaluated more objectively, and an abnormality recognition model can be established to provide early warning of potential problems. This will effectively reduce the subjectivity of human judgment and improve the accuracy and efficiency of monitoring, thereby helping to achieve better construction management and progress control.

[0097] S4. Perform data training on the anomaly recognition model using the data training set to obtain a trained anomaly recognition model, and deploy the tested anomaly recognition model to an environment corresponding to the construction site to obtain environmental parameters.

[0098] The present invention uses the data training set to perform data training on the anomaly recognition model to obtain a trained anomaly recognition model, which can help to better understand how the model makes anomaly detection decisions, and can analyze the model's output, feature importance, etc., thereby explaining the model's operation process and the basis for anomaly judgment.

[0099] The trained anomaly recognition model refers to a model obtained through optimization and adjustment with higher accuracy and generalization ability. Optionally, the trained anomaly recognition model can be obtained by performing data training on the anomaly recognition model using the data training set.

[0100] The present invention deploys the tested anomaly recognition model to the environment corresponding to the construction site to obtain environmental parameters, thereby achieving real-time monitoring, early warning and protection, improving work efficiency, performing data analysis and optimization, and realizing automated management, thereby bringing many benefits to the construction site.

[0101] The environmental parameters refer to various physical quantities or characteristics related to the environment in the construction site.

[0102] As an embodiment of the present invention, the tested abnormality recognition model is deployed to the environment corresponding to the construction site to obtain environmental parameters, including: determining the development environment corresponding to the construction site; based on the development environment, deploying the model service corresponding to the tested abnormality recognition model; based on the model service, performing environmental recognition on the construction site to obtain environmental parameters.

[0103] The development environment refers to the software and hardware environment required for model deployment and testing at a construction site; the model service refers to deploying a tested anomaly recognition model as an accessible service.

[0104] Furthermore, the development environment can be obtained through deep learning model implementation, such as convolutional neural network, recurrent neural network, long short-term memory network; the model service can be obtained through construction tools, such as Flask, Django and other tools.

[0105] S5. Perform abnormal identification on the environmental parameters to obtain environmental abnormality parameters, extract environmental abnormality factors from the environmental abnormality parameters, calculate environmental abnormality indicators corresponding to the environmental abnormality factors, and generate an early warning signal corresponding to the construction site based on the environmental abnormality indicators to realize abnormality identification in the construction site.

[0106] The present invention obtains environmental anomaly parameters by performing anomaly identification on the environmental parameters, which can be adjusted and optimized according to specific business needs to adapt to anomaly identification tasks in different environments and parameters, and the model can be modified, updated and improved according to actual conditions.

[0107] Among them, the environmental abnormality parameters refer to factors that may affect the environmental state or performance, such as temperature, humidity, pressure, light, electromagnetic radiation and other parameters. Optionally, the environmental abnormality parameters can be obtained through operation and maintenance tools, such as monitoring systems, log analysis tools and other tools.

[0108] As an embodiment of the present invention, the abnormality identification of the environmental parameters to obtain the environmental abnormality parameters includes: identifying an environmental parameter list corresponding to the environmental parameters; determining a real-time value in the environmental parameter list; setting a threshold range corresponding to the real-time value; and based on the threshold range, identifying the abnormality of the environmental parameters to obtain the environmental abnormality parameters.

[0109] Among them, the environmental parameter list refers to the list corresponding to each environmental parameter that needs to be monitored and identified, such as: temperature, humidity, pressure, light, etc.; the real-time value refers to the current value of the environmental parameter obtained by a sensor or other device; the threshold range refers to the effective range set for each environmental parameter.

[0110] Furthermore, the environmental parameter list can be obtained through an environmental monitoring system, such as sensors, data acquisition equipment and data processing tools; the real-time value can be obtained through sensors, such as temperature, humidity, light and other sensors; the threshold range can be obtained through a clustering algorithm.

[0111] By extracting the environmental anomaly factors from the environmental anomaly parameters, the present invention can be customized and adjusted according to specific needs to adapt to the extraction of environmental anomaly factors in specific scenarios, and can flexibly select appropriate feature extraction methods and models according to different environmental parameters and abnormal situations.

[0112] Among them, the environmental anomaly factor refers to a key feature or indicator that can cause anomalies in environmental parameters. Optionally, the environmental anomaly factor can be obtained through a supervised learning model, such as a decision tree, random forest, support vector machine and other models.

[0113] As an embodiment of the present invention, the calculating the environmental anomaly index corresponding to the environmental anomaly factor includes:

[0114] The environmental anomaly index corresponding to the environmental anomaly factor is calculated using the following formula:

[0115]

[0116] Among them, Z represents the environmental abnormality index corresponding to the environmental abnormality factor, Fi represents the i-th environmental abnormality factor, Fj represents the j-th environmental abnormality factor, n represents the number of environmental abnormality factors, and m represents the total number of environmental abnormality factors.

[0117] The present invention generates a warning signal corresponding to the construction site based on the environmental abnormality index to realize the identification of abnormalities in the construction site. It can identify abnormal situations in the construction site, and can promptly discover and solve potential environmental abnormality problems to prevent the problems from further deteriorating, thereby improving the safety and reliability of the construction site.

[0118] The warning signal refers to warning information about abnormal conditions in the construction site. Optionally, the warning signal can be obtained through monitoring equipment, such as on-site monitors and other tools.

[0119] The present invention can monitor the progress of the construction site in real time by acquiring a set of on-site images in the construction site and identifying the image data corresponding to the set of on-site images, which is helpful for real-time monitoring, improving quality, ensuring safety, improving efficiency and accumulating experience, and can improve construction quality and efficiency. The present invention can improve image quality, improve image analysis and processing results, protect image information privacy, and reduce data storage and transmission costs by performing image denoising on the image labeling data, thereby improving the effect and efficiency of image application. The present invention can improve the performance and generalization ability of the model and effectively evaluate the model by dividing the segmented image data into a data training set and performing feature extraction on the segmented image data to obtain image features. The present invention can obtain useful image features for subsequent image analysis and application. The present invention can obtain a trained anomaly recognition model by using the data training set to perform data training on the anomaly recognition model, which can help better understand how the model makes anomaly detection decisions, and can analyze the output of the model, feature importance, etc., so as to explain the operation process of the model and the basis for abnormal judgment. The present invention generates a warning signal corresponding to the construction site based on the environmental anomaly index to realize anomaly recognition in the construction site, and can identify abnormal situations in the construction site, and can timely discover and solve potential environmental anomaly problems to prevent the problem from further deteriorating, thereby improving the safety and reliability of the construction site. Therefore, the present invention proposes a method and system for realizing anomaly recognition under construction based on machine vision to improve the efficiency of anomaly recognition in construction.

[0120] like Figure 2 1 is a functional module diagram of a method and system for realizing abnormality recognition in construction based on machine vision provided by one embodiment of the present invention.

[0121] The abnormality identification system 200 for building construction based on machine vision of the present invention can be installed in an electronic device. According to the functions implemented, the abnormality identification system 200 for building construction based on machine vision can include an image marking module 201, an image segmentation module 202, a model building module 203, an environmental parameter module 204 and an index calculation module 205. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0122] In this embodiment, the functions of each module / unit are as follows:

[0123] The image marking module 201 is used to obtain a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-mark the image data to obtain image marking data;

[0124] The image segmentation module 202 is used to perform image denoising on the image tag data to obtain denoised image data, perform image enhancement on the denoised image data to obtain enhanced image data, and perform image segmentation on the enhanced image data to obtain segmented image data;

[0125] The model building module 203 is used to divide the segmented image data into a data training set, and perform feature extraction on the segmented image data to obtain image features, identify feature vectors corresponding to the image features, calculate feature similarities corresponding to the segmented image data based on the feature vectors, and build an abnormality recognition model corresponding to the scene image set based on the feature similarities;

[0126] The environmental parameter module 204 is used to perform data training on the abnormality recognition model using the data training set to obtain a trained abnormality recognition model, and deploy the tested abnormality recognition model to the environment corresponding to the construction site to obtain environmental parameters;

[0127] The index calculation module 205 is used to identify abnormalities in the environmental parameters, obtain environmental abnormality parameters, extract environmental abnormality factors from the environmental abnormality parameters, calculate environmental abnormality indicators corresponding to the environmental abnormality factors, and generate early warning signals corresponding to the construction site based on the environmental abnormality indicators to realize abnormality identification in the construction site.

[0128] In detail, the modules described in the system 200 for realizing abnormality identification in construction based on machine vision described in the embodiment of the present invention adopt the same technical means as the method for realizing abnormality identification in construction based on machine vision described in the accompanying drawings when in use, and can produce the same technical effects, which will not be repeated here.

[0129] like Figure 3 The figure is a schematic diagram of the structure of an electronic device for realizing an abnormality recognition method in building construction based on machine vision according to the present invention.

[0130] The electronic device 1 may include a processor 30, a memory 31, a communication bus 32 and a communication interface 33, and may also include a computer program stored in the memory 31 and executable on the processor 30, such as an engineering safety supervision program based on artificial intelligence.

[0131] In some embodiments, the processor 30 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 30 is the control core (ControlUnit) of the electronic device 1, and uses various interfaces and lines to connect the various components of the entire electronic device, and executes or executes programs or modules stored in the memory 31 (for example, executing engineering safety supervision programs based on artificial intelligence, etc.), and calls data stored in the memory 31 to execute various functions of the electronic device and process data.

[0132] The memory 31 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 31 may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 31 may also be an external storage device of an electronic device, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Further, the memory 31 may also include both an internal storage unit of the electronic device and an external storage device. The memory 31 may not only be used to store application software and various types of data installed in the electronic device, such as the code of the database configuration connection program, but also be used to temporarily store data that has been output or is to be output.

[0133] The communication bus 32 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 31 and at least one processor 30, etc.

[0134] The communication interface 33 is used for communication between the above-mentioned electronic device 1 and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display (Display), an input unit (such as a keyboard (Keyboard)), and optionally, the user interface may also be a standard wired interface, a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device and to display a visual user interface.

[0135] Figure 3 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0136] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 30 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, and power status indicators. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0137] It should be understood that the described embodiments are for illustration purposes only.

[0138] The database configuration connection program stored in the memory 31 in the electronic device 1 is a combination of multiple computer programs. When running in the processor 30, it can achieve:

[0139] Acquire a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-label the image data to obtain image label data;

[0140] Performing image denoising on the image label data to obtain denoised image data, performing image enhancement on the denoised image data to obtain enhanced image data, and performing image segmentation on the enhanced image data to obtain segmented image data;

[0141] Dividing the segmented image data into a data training set, and performing feature extraction on the segmented image data to obtain image features, identifying feature vectors corresponding to the image features, calculating feature similarities corresponding to the segmented image data based on the feature vectors, and constructing an abnormality recognition model corresponding to the scene image set based on the feature similarities;

[0142] Performing data training on the anomaly recognition model using the data training set to obtain a trained anomaly recognition model, and deploying the tested anomaly recognition model to an environment corresponding to the construction site to obtain environmental parameters;

[0143] Anomalies are identified on the environmental parameters to obtain environmental anomaly parameters, environmental anomaly factors in the environmental anomaly parameters are extracted, environmental anomaly indicators corresponding to the environmental anomaly factors are calculated, and based on the environmental anomaly indicators, an early warning signal corresponding to the construction site is generated to realize abnormality identification in the construction site.

[0144] Specifically, the specific implementation method of the processor 30 for the above computer program can refer to Figure 1 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0145] Furthermore, if the module / unit integrated in the electronic device 1 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a non-volatile computer-readable storage medium. The storage medium can be volatile or non-volatile. For example, the computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0146] The present invention further provides a storage medium, wherein the readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement:

[0147] Acquire a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-label the image data to obtain image label data;

[0148] Performing image denoising on the image label data to obtain denoised image data, performing image enhancement on the denoised image data to obtain enhanced image data, and performing image segmentation on the enhanced image data to obtain segmented image data;

[0149] Dividing the segmented image data into a data training set, and performing feature extraction on the segmented image data to obtain image features, identifying feature vectors corresponding to the image features, calculating feature similarities corresponding to the segmented image data based on the feature vectors, and constructing an abnormality recognition model corresponding to the scene image set based on the feature similarities;

[0150] Performing data training on the anomaly recognition model using the data training set to obtain a trained anomaly recognition model, and deploying the tested anomaly recognition model to an environment corresponding to the construction site to obtain environmental parameters;

[0151] Anomalies are identified on the environmental parameters to obtain environmental anomaly parameters, environmental anomaly factors in the environmental anomaly parameters are extracted, environmental anomaly indicators corresponding to the environmental anomaly factors are calculated, and based on the environmental anomaly indicators, an early warning signal corresponding to the construction site is generated to realize abnormality identification in the construction site.

[0152] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0153] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0154] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0155] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0156] Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is therefore intended that all changes that fall within the meaning and range of equivalent elements of the claims are embraced in the present invention.

[0157] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0158] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for identifying abnormalities in construction based on machine vision, characterized in that: The method comprises: Acquire a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-label the image data to obtain image label data; Performing image denoising on the image label data to obtain denoised image data, performing image enhancement on the denoised image data to obtain enhanced image data, and performing image segmentation on the enhanced image data to obtain segmented image data; Dividing the segmented image data into a data training set, and performing feature extraction on the segmented image data to obtain image features, identifying feature vectors corresponding to the image features, calculating feature similarities corresponding to the segmented image data based on the feature vectors, and constructing an abnormality recognition model corresponding to the scene image set based on the feature similarities; Performing data training on the anomaly recognition model using the data training set to obtain a trained anomaly recognition model, and deploying the trained anomaly recognition model to an environment corresponding to the construction site to obtain environmental parameters; Performing abnormality identification on the environmental parameters to obtain environmental abnormality parameters, extracting environmental abnormality factors from the environmental abnormality parameters, calculating environmental abnormality indicators corresponding to the environmental abnormality factors, and generating early warning signals corresponding to the construction site based on the environmental abnormality indicators to achieve abnormality identification in the construction site; The calculating the environmental anomaly index corresponding to the environmental anomaly factor includes: The environmental anomaly index corresponding to the environmental anomaly factor is calculated using the following formula: Among them, Z represents the environmental abnormality index corresponding to the environmental abnormality factor, Fi represents the i-th environmental abnormality factor, Fj represents the j-th environmental abnormality factor, n represents the number of environmental abnormality factors, and m represents the total number of environmental abnormality factors.

2. The method for identifying abnormalities in construction based on machine vision according to claim 1, characterized in that: The step of acquiring a set of on-site images in a construction site and identifying image data corresponding to the set of on-site images includes: Determining the corresponding photographing equipment for the construction site; Acquiring a set of on-site images of a construction site using the photographing device; Storing the on-site image set in a preset cloud storage space; Performing data preprocessing on the data stored in the cloud storage space to obtain preprocessed data; Corresponding image data in the preprocessed data is identified.

3. The method for realizing abnormality identification in construction based on machine vision according to claim 1, characterized in that: The step of pre-labeling the image data to obtain image labeling data includes: Performing data adjustment on the image data to obtain adjusted image data; Defining a labeling category corresponding to the adjusted image data; Based on the annotation category, pre-annotate the adjusted image data to obtain initially-annotated image data; The initially marked image data is annotated and corrected to obtain image marked data.

4. The method for realizing abnormality identification in building construction based on machine vision according to claim 1, characterized in that: The performing image denoising on the image marked data to obtain denoised image data comprises: Importing the image marking data, and identifying the noise type corresponding to the image marking data; Based on the noise type, performing noise statistics on the image marking data to obtain noise statistical data; The noise statistical data is subjected to frequency domain denoising to obtain denoised image data.

5. The method for realizing abnormality identification in building construction based on machine vision as claimed in claim 1, characterized in that: The step of dividing the segmented image data into a data training set and performing feature extraction on the segmented image data to obtain image features includes: Proportionally dividing the segmented image data to obtain a data training set corresponding to the segmented image data; Normalizing the data training set to obtain normalized data; and determining a feature type corresponding to the segmented image data; Based on the feature type, feature extraction is performed on the segmented image data to obtain image features.

6. The method for identifying abnormalities in construction based on machine vision according to claim 1, characterized in that: The calculating the feature similarity corresponding to the segmented image data based on the feature vector includes: segmenting the feature vector into a first feature vector and a second feature vector; The feature similarity corresponding to the image feature is calculated using the following formula: Among them, CS represents the feature similarity corresponding to the image feature, and its value range is between [-1, 1]. The closer the value is to 1, the more similar it is, and the closer the value is to -1, the less similar it is. Ai represents the corresponding i-th element in the first feature vector, and Bi represents the corresponding i-th element in the second feature vector.

7. The method for identifying abnormalities in construction based on machine vision as claimed in claim 1, characterized in that: The step of deploying the trained anomaly recognition model to an environment corresponding to the construction site to obtain environmental parameters includes: Determining a development environment corresponding to the building construction site; Based on the development environment, deploy the model service corresponding to the trained anomaly recognition model; Based on the model service, the environment of the construction site is identified to obtain environmental parameters.

8. The method for identifying abnormalities in construction based on machine vision according to claim 1, characterized in that: The step of identifying abnormalities of the environmental parameters to obtain abnormal environmental parameters includes: Identify an environmental parameter list corresponding to the environmental parameter; determine a real-time value in the environmental parameter list; Setting a threshold range corresponding to the real-time value; Based on the threshold range, the environmental parameter is identified as abnormal to obtain an environmental abnormality parameter.

9. An abnormality recognition system under construction based on machine vision is characterized by: The system is used to implement the method for realizing abnormality recognition in building construction based on machine vision according to any one of claims 1 to 8, and comprises: An image marking module is used to obtain a set of on-site images in a construction site, identify image data corresponding to the set of on-site images, and pre-mark the image data to obtain image marking data; An image segmentation module, used for performing image denoising on the image tag data to obtain denoised image data, performing image enhancement on the denoised image data to obtain enhanced image data, and performing image segmentation on the enhanced image data to obtain segmented image data; A model building module, used to divide the segmented image data into a data training set, and perform feature extraction on the segmented image data to obtain image features, identify feature vectors corresponding to the image features, calculate feature similarities corresponding to the segmented image data based on the feature vectors, and build an abnormality recognition model corresponding to the scene image set based on the feature similarities; An environmental parameter module is used to perform data training on the anomaly recognition model using the data training set to obtain a trained anomaly recognition model, and deploy the trained anomaly recognition model to the environment corresponding to the construction site to obtain environmental parameters; The indicator calculation module is used to identify abnormalities in the environmental parameters, obtain environmental abnormality parameters, extract environmental abnormality factors in the environmental abnormality parameters, calculate environmental abnormality indicators corresponding to the environmental abnormality factors, and generate early warning signals corresponding to the construction site based on the environmental abnormality indicators to realize abnormality identification in the construction site.