Data analysis method and system for road construction safety engineering based on AI

Abstract

The invention discloses a data analysis method and system for road construction safety engineering based on AI, and relates to the field of road construction safety, and the method comprises the steps: obtaining an abnormal event image of the road construction safety engineering, and building a road construction safety database; an image acquisition device is arranged, and road construction information is acquired in different road section division areas; the image quality is improved according to a dark channel prior defogging algorithm; carrying out edge detection and identification on identifiable safety construction marks and protective article features in different areas; the recognizable capability of recognizable safety construction marks and protective article feature edge detection in different areas is further improved; establishing an AI road construction safety detection model, and judging whether the road construction has risks or not; a man-machine interaction platform is established, and monitoring data of road construction are received, stored and sorted. The identification efficiency of road construction safety is effectively improved, and the road construction risk caused by wrong operation of safety construction marks and protective articles is reduced.

Description

Technical Field

[0001] The present invention relates to the field of road construction safety, and in particular to a data analysis method and system for road construction safety engineering based on AI. Background Art

[0002] Road construction usually involves complex working environments and various potential dangers, especially in areas with complex traffic flow. The incorrect use of safety construction signs and protective equipment can easily lead to road construction safety accidents and personal injuries. Therefore, it is of great significance to ensure road construction safety and improve the recognition ability of road construction safety, thereby protecting the lives of construction workers and the public, ensuring the smooth completion of engineering projects, and reducing the occurrence of accidents and property losses. With the development of AI technology, applying AI technology to road construction safety detection can effectively improve the recognition efficiency of road construction safety and reduce road construction safety risks.

[0003] Existing road construction safety analysis mainly relies on manual inspection, which is done by arranging personnel to inspect the construction site according to time periods or randomly. However, manual inspection has high requirements for manpower and material resources, and it is difficult to achieve real-time inspection around the clock. The lack of detailed road construction safety analysis leads to low efficiency in identifying road construction safety, prone to road construction safety risks, and difficulty in comprehensively ensuring road construction safety. Summary of the invention

[0004] In order to solve the above technical problems, a data analysis method and system for road construction safety engineering based on AI are provided. This technical solution solves the problem that manual inspection proposed in the above background technology has high requirements for manpower and material resources, and it is difficult to achieve real-time inspection around the clock. There is a lack of detailed road construction safety analysis, which leads to low efficiency in identifying road construction safety, prone to road construction safety risks, and difficulty in comprehensively ensuring road construction safety.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is: A data analysis method for road construction safety engineering based on AI, characterized by comprising: Based on big data, obtain abnormal event images of road construction safety projects and establish a road construction safety database; An image acquisition device is provided to collect road construction information in different road sections according to the actual situation of road construction; Based on the dark channel prior defogging algorithm, the impact of dust and haze on the clarity of real-time collected images at the construction site is reduced, and the image quality is improved; Based on the edge detection algorithm, edge detection and recognition are performed on the identifiable safety construction signs and protective equipment features in different road construction areas; According to the Hough transform formula, the ability to detect the edges of safety construction signs and protective equipment features in different road construction areas is further improved; Establish an AI road construction safety detection model to determine whether there are risks in road construction by setting limited thresholds that can identify safety construction signs and protective equipment characteristics in different road construction areas; Based on cloud technology, a human-computer interaction platform is established to receive, store and organize road construction monitoring data, and to receive and view road construction safety risk information in real time.

[0006] Preferably, the acquisition of abnormal event images of road construction safety engineering based on big data and establishment of a road construction safety database specifically includes: Based on big data, obtain abnormal images of safety accidents during road construction and standard road construction safety images, and establish road construction abnormal atlas and road construction standard atlas; According to the road construction safety standards, extract identifiable safety construction signs and protective equipment features from the road construction abnormality atlas and road construction standard atlas; Based on the grayscale formula, the identifiable safety construction signs and protective equipment features in the road construction abnormal atlas and road construction standard atlas are grayscaled; Based on the Gaussian filtering algorithm, the grayscale identifiable safety construction signs and protective equipment feature data are filtered; Classify and organize the filtered identifiable safety construction signs and protective equipment feature data, and establish a road construction safety database; The grayscale formula is:

[0007] In the formula, To identify the safety construction signs and protective equipment feature images in the first Line The gray value of the pixel. , , are the weighted coefficients of human eye sensitivity to red, green and blue, respectively. , , They are respectively the identifiable safety construction signs and the protective equipment feature images. Line The red, green and blue color components of the column pixel; The formula for filtering based on the Gaussian filtering algorithm is:

[0008] In the formula, To identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, To identify the safety construction signs and protective equipment feature images in the first Line The pixel value of the column, is a Gaussian filter function that can identify the characteristic images of safety construction signs and protective equipment. Set the total number of window movement offsets for identifiable safety construction signs and protective equipment feature images.

[0009] Preferably, the image acquisition device is provided to collect road construction information in different road sections according to the actual situation of road construction, and specifically includes: According to the road construction plan and the actual situation on site, the road is divided into multiple areas, among which the divided areas are mainly divided into: intersections, main roads, and temporary passages; According to the division of road construction areas, an image acquisition device is set up to collect construction road information in real time and in all directions.

[0010] Preferably, the dark channel prior defogging algorithm is used to reduce the impact of dust and haze on the clarity of real-time collected images and improve the quality of images, which specifically includes: According to the road construction safety database and based on the dark channel prior theory, a dark channel map is generated to identify the safety construction signs and defogging of the protective equipment feature images; Determine the atmospheric light value based on the dark channel map of the defogging of the characteristic image of the identifiable safety construction signs and protective equipment; Based on big data, the transmission rate of light through fog at each pixel in the image that can identify safety construction signs and protective equipment features is obtained; Based on the dark channel prior defogging algorithm, an optimization model for identifying safety construction signs and protective equipment feature images is established to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site and improve the quality of images. The dark channel priori defogging algorithm expression is:

[0011] In the formula, The defogging process is used to identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, To identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, In order to identify the safety construction signs and protective equipment features, the Gaussian filtered image is Line The transmission rate of light through fog is is the atmospheric light value, In order to identify the safety construction signs and protective equipment features, the Gaussian filtered image is The coordinates of the window.

[0012] Preferably, the edge detection and identification of identifiable safety construction signs and protective equipment features in different road construction areas based on the edge detection algorithm specifically includes: According to the dehazed identifiable safety construction signs and protective equipment feature images, the gradient value and gradient direction of the first-order derivative are calculated based on the Canny operator; According to the gradient value and gradient direction of the first-order derivative, non-maximum suppression is used to remove some non-edge points of the identifiable safety construction signs and protective equipment feature images after defogging; According to the dual threshold method, a maximum threshold and a minimum threshold are set based on big data analysis; Construct an edge detection model for images of identifiable safety construction signs and protective equipment features, and perform edge detection and recognition on identifiable safety construction signs and protective equipment features in different road construction areas; The gradient value expression of the first-order derivative is:

[0013] In the formula, The defogging image shows the identifiable safety construction signs and protective equipment features. Line The gradient value of the first derivative of the column, To identify the safety construction signs and protective equipment feature images after defogging The first-order partial derivative matrix in the direction, To identify the safety construction signs and protective equipment feature images after defogging The first-order partial derivative matrix of the direction; The gradient direction expression of the first-order derivative is:

[0014] In the formula, It is the first-order derivative gradient direction of the dehazed identifiable safety construction signs and protective equipment feature images.

[0015] Preferably, the further improving the recognition capability of the safety construction signs and the edge detection of the protective equipment features in different road construction areas according to the Hough transform formula specifically includes: According to an edge detection model for identifiable safety construction signs and protective equipment feature images, edge detection data for identifiable safety construction signs and protective equipment feature images are obtained; The reference point coordinates are set according to the standard data in the edge detection data of the identifiable safety construction signs and protective equipment feature images; Select the coordinates with obvious corner features within the characteristic contours of identifiable safety construction signs and protective equipment as identifiable division points, and obtain the gradient direction of the point; Based on the Hough transform formula of arbitrary graphics, the edge information of identifiable safety construction signs and protective equipment features is optimized to further improve the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas; The Hough transform formula of the arbitrary graph is:

[0016] In the formula, To identify the safety construction signs and protective equipment feature outlines The coordinates of the identifiable partition points, To identify the safety construction signs and protective equipment features on the outline Point coordinates, For point To point The distance value of To identify the safety construction signs and protective equipment features on the outline The gradient direction of a point.

[0017] Preferably, the establishment of an AI road construction safety detection model, by setting a limited threshold for identifying safety construction signs and protective equipment features in different road construction areas, to determine whether there is a risk in road construction specifically includes: According to the identifiable safety construction signs and protective equipment features optimized by the Hough transform formula of any graphics, an AI road construction safety detection model is established based on the neural network algorithm; Based on big data, set the limited range and threshold for identifying safety construction signs and protective equipment characteristics in different road construction areas; Through real-time collection of construction sites by image acquisition devices and combined with AI road construction safety detection models, it is possible to determine whether there are abnormal conditions and risks in different road construction areas; Determine whether the overlap rate between the model calculation value and the limited range is greater than the threshold. If so, it means that there is no safety risk in the road construction in the area. If not, it means that there is a safety risk in the road construction in the area. Mark the abnormal feature points and issue alarms and warning messages.

[0018] Preferably, the cloud-based technology is used to establish a human-computer interaction platform for receiving, storing and collating road construction monitoring data, and receiving and viewing road construction safety risk information in real time, specifically including: Based on cloud technology, a human-computer interaction platform is established to receive, store and organize monitoring data of road construction; Based on the human-computer interaction platform, the hardware operating environment for building the AI ​​road construction safety detection model is used to ensure the normal operation of the AI ​​road construction safety detection model; Based on the human-computer interaction platform, distinguishable alarm devices are set up to issue alarms and warnings for road construction areas with safety risks; Based on the human-computer interaction platform, it is used to update and optimize the parameters and database information of the AI ​​road construction safety detection model.

[0019] Furthermore, the present invention proposes a data analysis system for road construction safety engineering based on AI, which is used to implement the data analysis method for road construction safety engineering based on AI as described above, including: A data acquisition module, which is used to acquire abnormal event images of road construction safety projects based on big data and establish a road construction safety database; A hardware setting module, which is used to set an image acquisition device to collect road construction information in different road sections according to the actual situation of road construction; A safety detection module is used to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site based on a dark channel priori defogging algorithm, thereby improving the quality of images; based on an edge detection algorithm, edge detection and identification of identifiable safety construction signs and protective equipment features in different road construction areas are performed; based on the Hough transform formula, the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas is further improved; an AI road construction safety detection model is established, and by setting a limited threshold for identifiable safety construction signs and protective equipment features in different road construction areas, it is determined whether there is a risk in road construction; The platform processing module is used to establish a human-computer interaction platform based on cloud technology, which is used to receive, store and organize monitoring data of road construction, and receive and view risk information of road construction safety in real time.

[0020] Preferably, the safety detection module specifically includes: A defogging optimization unit, which is used to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site based on a dark channel priori defogging algorithm, thereby improving the quality of the image; An edge detection unit, the edge detection unit is used to perform edge detection and recognition of identifiable safety construction signs and protective equipment features in different road construction areas based on an edge detection algorithm; An edge detection optimization unit, the edge detection optimization unit is used to further improve the ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas according to the Hough transform formula; A risk judgment unit is used to establish an AI road construction safety detection model, and to judge whether there is a risk in road construction by setting a limited threshold for identifying safety construction signs and protective equipment characteristics in different road construction areas.

[0021] Compared with the prior art, the present invention has the following beneficial effects: Based on big data, the features of identifiable safety construction signs and protective equipment are extracted, and the feature data of identifiable safety construction signs and protective equipment are grayed and Gaussian filtered. At the same time, through the dark channel prior defogging algorithm, an optimization model of identifiable safety construction signs and protective equipment feature images is established to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site and improve the quality of images. Secondly, through the Canny operator edge detection algorithm, an edge detection model of identifiable safety construction signs and protective equipment feature images is constructed to perform edge detection and recognition on the identifiable safety construction signs and protective equipment features in different road construction areas. Furthermore, through the Hough transform formula of any graphics, the edge information of the identifiable safety construction signs and protective equipment features is optimized to further improve the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas. Finally, based on the identifiable safety construction signs and protective equipment features optimized by the Hough transform formula of any graphics, an AI road construction safety detection model is established based on the neural network algorithm to determine whether there is a risk in road construction, thereby effectively improving the recognition efficiency of road construction safety and reducing the road construction risks caused by incorrect operation of safety construction signs and protective equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A flow chart of a data analysis method for road construction safety engineering based on AI according to the present invention; Figure 2 This is a flow chart of the present invention, which uses a dark channel priori defogging algorithm to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site and improve the quality of the image; Figure 3 The edge detection algorithm of the present invention is used to detect and identify the edge of identifiable safety construction signs and protective equipment features in different road construction areas; Figure 4This is a flow chart of the present invention for further improving the recognition capability of safety construction signs and protective equipment feature edge detection in different road construction areas based on the Hough transform formula. DETAILED DESCRIPTION

[0023] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0024] Reference Figure 1 As shown, a data analysis method for road construction safety engineering based on AI includes: Based on big data, obtain abnormal event images of road construction safety projects and establish a road construction safety database; An image acquisition device is provided to collect road construction information in different road sections according to the actual situation of road construction; Based on the dark channel prior defogging algorithm, the impact of dust and haze on the clarity of real-time collected images at the construction site is reduced, and the image quality is improved; Based on the edge detection algorithm, edge detection and recognition are performed on the identifiable safety construction signs and protective equipment features in different road construction areas; According to the Hough transform formula, the ability to detect the edges of safety construction signs and protective equipment features in different road construction areas is further improved; Establish an AI road construction safety detection model to determine whether there are risks in road construction by setting limited thresholds that can identify safety construction signs and protective equipment characteristics in different road construction areas; Based on cloud technology, a human-computer interaction platform is established to receive, store and organize road construction monitoring data, and to receive and view road construction safety risk information in real time.

[0025] It can be explained that this solution is based on big data to extract the features of identifiable safety construction signs and protective equipment, and grayscale and Gaussian filter the identifiable safety construction signs and protective equipment feature data. At the same time, a dark channel prior defogging algorithm is used to establish an optimization model for the feature image of identifiable safety construction signs and protective equipment to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site and improve the quality of the image. Secondly, a Canny operator edge detection algorithm is used to construct an edge detection model for the feature image of identifiable safety construction signs and protective equipment, and the identifiable safety construction signs and protective equipment features in different road construction areas are optimized. The edge detection and recognition of the features are carried out. Furthermore, the edge information of the identifiable safety construction signs and protective equipment features is optimized through the Hough transform formula of any graphics, so as to further improve the identifiable ability of the edge detection of the identifiable safety construction signs and protective equipment features in different road construction areas. Finally, according to the identifiable safety construction signs and protective equipment features optimized by the Hough transform formula of any graphics, an AI road construction safety detection model is established based on the neural network algorithm to judge whether there are risks in road construction, thereby effectively improving the recognition efficiency of road construction safety and reducing the road construction risks caused by incorrect operation of safety construction signs and protective equipment.

[0026] Reference Figure 2 As shown, the dark channel prior defogging algorithm is used to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site and improve the quality of the image, specifically including: According to the road construction safety database and based on the dark channel prior theory, a dark channel map is generated to identify the safety construction signs and defogging of the protective equipment feature images; Determine the atmospheric light value based on the dark channel map of the defogging of the characteristic image of the identifiable safety construction signs and protective equipment; Based on big data, the transmission rate of light through fog at each pixel in the image that can identify safety construction signs and protective equipment features is obtained; Based on the dark channel prior defogging algorithm, an optimization model for identifying safety construction signs and protective equipment feature images is established to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site and improve the quality of images. The dark channel priori defogging algorithm expression is:

[0027] In the formula, The defogging process is used to identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, In order to identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, In order to identify the safety construction signs and protective equipment features, the Gaussian filtered image is Line The transmission rate of light through fog is is the atmospheric light value, In order to identify the safety construction signs and protective equipment features, the Gaussian filtered image is The coordinates of the window.

[0028] It can be explained that during the road construction process, it is very easy to generate a large amount of dust or be affected by haze weather, resulting in unclear and blurred image data collected by the image acquisition device, resulting in inaccurate extraction of identifiable safety construction signs and protective equipment features at the road construction site, which can easily lead to system misjudgment. This scheme introduces a dark channel prior defogging algorithm to improve the clarity and contrast of identifiable safety construction signs and protective equipment feature images, making them easier to observe and identify, thereby reducing the impact of dust and haze on the construction site on the clarity of real-time collected images and improving image quality.

[0029] Reference Figure 3 As shown, the edge detection and identification of identifiable safety construction signs and protective equipment features in different road construction areas based on the edge detection algorithm specifically includes: According to the dehazed identifiable safety construction signs and protective equipment feature images, the gradient value and gradient direction of the first-order derivative are calculated based on the Canny operator; According to the gradient value and gradient direction of the first-order derivative, non-maximum suppression is used to remove some non-edge points of the identifiable safety construction signs and protective equipment feature images after defogging; According to the dual threshold method, a maximum threshold and a minimum threshold are set based on big data analysis; Construct an edge detection model for images of identifiable safety construction signs and protective equipment features, and perform edge detection and recognition on identifiable safety construction signs and protective equipment features in different road construction areas; The gradient value expression of the first-order derivative is:

[0030] In the formula, The defogging image shows the identifiable safety construction signs and protective equipment features. Line The gradient value of the first derivative of the column, To identify the safety construction signs and protective equipment feature images after defogging The first-order partial derivative matrix in the direction, To identify the safety construction signs and protective equipment feature images after defogging The first-order partial derivative matrix of the direction; The gradient direction expression of the first-order derivative is:

[0031] In the formula, It is the first-order derivative gradient direction of the dehazed identifiable safety construction signs and protective equipment feature images.

[0032] It can be explained that in road construction projects, according to road construction safety standards, identifiable safety construction signs and protective equipment features have obvious outer contour markings. According to the image data collected by the image acquisition device, by extracting the outer contour markings of identifiable safety construction signs and protective equipment features, road construction safety can be effectively identified. This scheme uses the Canny operator edge detection algorithm to construct an image edge detection model for identifiable safety construction signs and protective equipment features, and performs edge detection and identification on identifiable safety construction signs and protective equipment features in different road construction areas, thereby improving the detection capability of road construction safety.

[0033] Reference Figure 4 As shown, the Hough transform formula is used to further improve the recognition capability of safety construction signs and protective equipment feature edge detection in different road construction areas, specifically including: According to an edge detection model for identifiable safety construction signs and protective equipment feature images, edge detection data for identifiable safety construction signs and protective equipment feature images are obtained; The reference point coordinates are set according to the standard data in the edge detection data of the identifiable safety construction signs and protective equipment feature images; Select the coordinates with obvious corner features within the characteristic contours of identifiable safety construction signs and protective equipment as identifiable division points, and obtain the gradient direction of the point; Based on the Hough transform formula of arbitrary graphics, the edge information of identifiable safety construction signs and protective equipment features is optimized to further improve the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas; The Hough transform formula of the arbitrary graph is:

[0034] In the formula, To identify the safety construction signs and protective equipment feature outlines The coordinates of the identifiable partition points, To identify the safety construction signs and protective equipment features on the outline Point coordinates, For point To point The distance value of To identify the safety construction signs and protective equipment features on the outline The gradient direction of a point.

[0035] It can be explained that the Canny operator edge detection can extract the outer contours of identifiable safety construction signs and protective equipment feature images to a certain extent, but the outer contours of the identifiable safety construction signs and protective equipment feature images obtained in actual processing are not very clear, and some features are not obvious, which can easily lead to inaccurate final judgments by the system. This solution further extracts the outer contours of identifiable safety construction signs and protective equipment feature images by utilizing the Hough transform formula of arbitrary graphics, and optimizes the edge information of identifiable safety construction signs and protective equipment features, thereby ensuring that the outer contours of identifiable safety construction signs and protective equipment feature images can accurately identify the outer contours in the image even if they are interfered by noise or missing features. Therefore, the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas is further improved.

[0036] Further, based on the same inventive concept as the above-mentioned AI-based data analysis method for road construction safety engineering, this solution proposes an AI-based data analysis system for road construction safety engineering, including: A data acquisition module, which is used to acquire abnormal event images of road construction safety projects based on big data and establish a road construction safety database; A hardware setting module, which is used to set an image acquisition device to collect road construction information in different road sections according to the actual situation of road construction; A safety detection module is used to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site based on a dark channel priori defogging algorithm, thereby improving the quality of images; based on an edge detection algorithm, edge detection and identification of identifiable safety construction signs and protective equipment features in different road construction areas are performed; based on the Hough transform formula, the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas is further improved; an AI road construction safety detection model is established, and by setting a limited threshold for identifiable safety construction signs and protective equipment features in different road construction areas, it is determined whether there is a risk in road construction; A platform processing module, which is used to establish a human-computer interaction platform based on cloud technology, to receive, store and organize road construction monitoring data, and to receive and view road construction safety risk information in real time; The security detection module specifically includes: A defogging optimization unit, which is used to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site based on a dark channel priori defogging algorithm, thereby improving the quality of the image; An edge detection unit, the edge detection unit is used to perform edge detection and recognition of identifiable safety construction signs and protective equipment features in different road construction areas based on an edge detection algorithm; An edge detection optimization unit, the edge detection optimization unit is used to further improve the ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas according to the Hough transform formula; A risk judgment unit is used to establish an AI road construction safety detection model, and to judge whether there is a risk in road construction by setting a limited threshold for identifying safety construction signs and protective equipment characteristics in different road construction areas.

[0037] In summary, the advantages of the present invention are: effectively improving the recognition efficiency of road construction safety and reducing the road construction risks caused by incorrect operation of safety construction signs and protective equipment.

[0038] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A data analysis method for road construction safety engineering based on AI, characterized in that: include: Based on big data, obtain abnormal event images of road construction safety projects and establish a road construction safety database; An image acquisition device is provided to collect road construction information in different road sections according to the actual situation of road construction; Based on the dark channel prior defogging algorithm, the impact of dust and haze on the clarity of real-time collected images at the construction site is reduced, and the image quality is improved; Based on the edge detection algorithm, edge detection and recognition are performed on the identifiable safety construction signs and protective equipment features in different road construction areas; According to the Hough transform formula, the ability to detect the edges of safety construction signs and protective equipment features in different road construction areas is further improved; Establish an AI road construction safety detection model to determine whether there are risks in road construction by setting limited thresholds that can identify safety construction signs and protective equipment characteristics in different road construction areas; Based on cloud technology, a human-computer interaction platform is established to receive, store and organize road construction monitoring data, and to receive and view road construction safety risk information in real time.

2. According to the data analysis method of a road construction safety engineering based on AI in claim 1, it is characterized in that: The method of obtaining abnormal event images of road construction safety engineering based on big data and establishing a road construction safety database specifically includes: Based on big data, obtain abnormal images of safety accidents during road construction and standard road construction safety images, and establish road construction abnormal atlas and road construction standard atlas; According to the road construction safety standards, extract identifiable safety construction signs and protective equipment features from the road construction abnormality atlas and road construction standard atlas; Based on the grayscale formula, the identifiable safety construction signs and protective equipment features in the road construction abnormal atlas and road construction standard atlas are grayscaled; Based on the Gaussian filtering algorithm, the grayscale identifiable safety construction signs and protective equipment feature data are filtered; Classify and organize the filtered identifiable safety construction signs and protective equipment feature data, and establish a road construction safety database; The grayscale formula is: ； In the formula, In order to identify the safety construction signs and protective equipment feature images in the first Line The gray value of the pixel. , , are the weighted coefficients of human eye sensitivity to red, green and blue, respectively. , , They are respectively the identifiable safety construction signs and the protective equipment feature images. Line The red, green and blue color components of the column pixel; The formula for filtering based on the Gaussian filtering algorithm is: ； In the formula, In order to identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, In order to identify the safety construction signs and protective equipment feature images in the first Line The pixel value of the column, is a Gaussian filter function that can identify the characteristic images of safety construction signs and protective equipment. Set the total number of window movement offsets for identifiable safety construction signs and protective equipment feature images.

3. According to claim 2, a data analysis method for road construction safety engineering based on AI is characterized in that: The image acquisition device is provided to collect road construction information in different road sections according to the actual situation of road construction, and specifically includes: According to the road construction plan and the actual situation on site, the road is divided into multiple areas, among which the divided areas are mainly divided into: intersections, main roads, and temporary passages; According to the division of road construction areas, an image acquisition device is set up to collect construction road information in real time and in all directions.

4. According to the data analysis method of a road construction safety engineering based on AI in claim 3, it is characterized in that: The dark channel prior defogging algorithm is used to reduce the impact of dust and haze on the clarity of real-time collected images and improve the quality of images, specifically including: According to the road construction safety database and based on the dark channel prior theory, a dark channel map is generated to identify the safety construction signs and defogging of the protective equipment feature images; Determine the atmospheric light value based on the dark channel map of the defogging of the characteristic image of the identifiable safety construction signs and protective equipment; Based on big data, the transmission rate of light through fog at each pixel in the image that can identify safety construction signs and protective equipment features is obtained; Based on the dark channel prior defogging algorithm, an optimization model for identifying safety construction signs and protective equipment feature images is established to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site and improve the quality of images. The dark channel priori defogging algorithm expression is: ； In the formula, The defogging process is used to identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, In order to identify the safety construction signs and protective equipment feature images in the first Line The pixel values ​​after Gaussian filtering, In order to identify the safety construction signs and protective equipment features, the Gaussian filtered image is Line The transmission rate of light through fog is is the atmospheric light value, In order to identify the safety construction signs and protective equipment features, the Gaussian filtered image is The coordinates of the window.

5. The data analysis method for road construction safety engineering based on AI according to claim 4 is characterized in that: The edge detection and identification of identifiable safety construction signs and protective equipment features in different road construction areas based on the edge detection algorithm specifically includes: According to the dehazed identifiable safety construction signs and protective equipment feature images, the gradient value and gradient direction of the first-order derivative are calculated based on the Canny operator; According to the gradient value and gradient direction of the first-order derivative, non-maximum suppression is used to remove some non-edge points of the identifiable safety construction signs and protective equipment feature images after defogging; According to the dual threshold method, a maximum threshold and a minimum threshold are set based on big data analysis; Construct an edge detection model for images of identifiable safety construction signs and protective equipment features, and perform edge detection and recognition on identifiable safety construction signs and protective equipment features in different road construction areas; The gradient value expression of the first-order derivative is: ； In the formula, The defogging image shows the identifiable safety construction signs and protective equipment features. Line The gradient value of the first derivative of the column, To identify the safety construction signs and protective equipment feature images after defogging The first-order partial derivative matrix in the direction, To identify the safety construction signs and protective equipment feature images after defogging The first-order partial derivative matrix of the direction; The gradient direction expression of the first-order derivative is: ； In the formula, It is the first-order derivative gradient direction of the dehazed identifiable safety construction signs and protective equipment feature images.

6. The data analysis method for road construction safety engineering based on AI according to claim 5 is characterized in that: The Hough transform formula is used to further improve the ability to detect the edges of safety construction signs and protective equipment features in different road construction areas. Specifically, the following are included: According to an edge detection model for identifiable safety construction signs and protective equipment feature images, edge detection data for identifiable safety construction signs and protective equipment feature images are obtained; The reference point coordinates are set according to the standard data in the edge detection data of the identifiable safety construction signs and protective equipment feature images; Select the coordinates with obvious corner features within the characteristic contours of identifiable safety construction signs and protective equipment as identifiable division points, and obtain the gradient direction of the point; Based on the Hough transform formula of arbitrary graphics, the edge information of identifiable safety construction signs and protective equipment features is optimized to further improve the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas; The Hough transform formula of the arbitrary graph is: ； In the formula, To identify the safety construction signs and protective equipment feature outlines The coordinates of the identifiable partition points, To identify the safety construction signs and protective equipment features on the outline Point coordinates, For point To point The distance value of To identify the safety construction signs and protective equipment features on the outline The gradient direction of a point.

7. The data analysis method for road construction safety engineering based on AI according to claim 6 is characterized in that: The establishment of an AI road construction safety detection model, by setting a limited threshold for identifying safety construction signs and protective equipment features in different road construction areas, determines whether there is a risk in road construction, specifically including: According to the identifiable safety construction signs and protective equipment features optimized by the Hough transform formula of any graphics, an AI road construction safety detection model is established based on the neural network algorithm; Based on big data, set the limited range and threshold for identifying safety construction signs and protective equipment characteristics in different road construction areas; Through real-time collection of construction sites by image acquisition devices and combined with AI road construction safety detection models, it is possible to determine whether there are abnormal conditions and risks in different road construction areas; Determine whether the overlap rate between the model calculation value and the limited range is greater than the threshold. If so, it means that there is no safety risk in the road construction in the area. If not, it means that there is a safety risk in the road construction in the area. Mark the abnormal feature points and issue alarms and warning messages.

8. The data analysis method for road construction safety engineering based on AI according to claim 7 is characterized in that: The cloud-based technology is used to establish a human-computer interaction platform for receiving, storing and collating road construction monitoring data, and receiving and viewing road construction safety risk information in real time, including: Based on cloud technology, a human-computer interaction platform is established to receive, store and organize monitoring data of road construction; Based on the human-computer interaction platform, the hardware operating environment for building the AI ​​road construction safety detection model is used to ensure the normal operation of the AI ​​road construction safety detection model; Based on the human-computer interaction platform, distinguishable alarm devices are set up to issue alarms and warnings for road construction areas with safety risks; Based on the human-computer interaction platform, it is used to update and optimize the parameters and database information of the AI ​​road construction safety detection model.

9. A data analysis system for road construction safety engineering based on AI, characterized in that: A data analysis method for implementing an AI-based road construction safety engineering as described in any one of claims 1 to 8, comprising: A data acquisition module, which is used to acquire abnormal event images of road construction safety projects based on big data and establish a road construction safety database; A hardware setting module, which is used to set an image acquisition device to collect road construction information in different road sections according to the actual situation of road construction; A safety detection module is used to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site based on a dark channel priori defogging algorithm, thereby improving the quality of images; based on an edge detection algorithm, edge detection and identification of identifiable safety construction signs and protective equipment features in different road construction areas are performed; based on the Hough transform formula, the identifiable ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas is further improved; an AI road construction safety detection model is established, and by setting a limited threshold for identifiable safety construction signs and protective equipment features in different road construction areas, it is determined whether there is a risk in road construction; The platform processing module is used to establish a human-computer interaction platform based on cloud technology, which is used to receive, store and organize monitoring data of road construction, and receive and view risk information of road construction safety in real time.

10. The data analysis system for road construction safety engineering based on AI according to claim 9 is characterized in that: The safety detection module specifically includes: A defogging optimization unit, which is used to reduce the impact of dust and haze on the clarity of real-time collected images at the construction site based on a dark channel priori defogging algorithm, thereby improving the quality of the image; An edge detection unit, the edge detection unit is used to perform edge detection and recognition of identifiable safety construction signs and protective equipment features in different road construction areas based on an edge detection algorithm; An edge detection optimization unit, the edge detection optimization unit is used to further improve the ability of edge detection of identifiable safety construction signs and protective equipment features in different road construction areas according to the Hough transform formula; A risk judgment unit is used to establish an AI road construction safety detection model, and to judge whether there is a risk in road construction by setting a limited threshold for identifying safety construction signs and protective equipment characteristics in different road construction areas.

Images