A machine vision-based reinforced concrete silo construction safety monitoring system and method

Abstract

The application discloses a kind of based on machine vision's reinforced concrete silo construction safety monitoring system and method, related to the field of silo construction, solve the problem that existing reinforced concrete silo construction safety monitoring system has poor monitoring effect, including warehouse outside monitoring module: to be in visual structure monitoring period first to fifth visual monitoring plane periodic visual image analysis, and according to the analysis result obtains the structure visual monitoring deviation corresponding to target shallow circular silo, warehouse inside monitoring module: respectively to each warehouse inside monitoring area is damaged part monitoring, according to the monitoring result obtains the damaged area area ratio corresponding to each warehouse inside monitoring area, and according to this analysis obtains the warehouse inside area damage coefficient, monitoring early warning module: according to structure visual monitoring deviation and warehouse inside area damage coefficient, target shallow circular silo is carried out construction safety early warning, the application can provide guarantee for the construction safety of reinforced concrete silo.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of silo construction, and relates to a machine vision technology, in particular to a reinforced concrete silo construction safety monitoring system and method based on machine vision. BACKGROUND

[0002] The existing reinforced concrete silo construction safety monitoring system has the following defects when performing safety construction on a silo:

[0003] 1. The existing reinforced concrete silo construction safety monitoring system cannot perform periodic visual image monitoring on multiple different visual monitoring planes outside the target shallow silo cylinder, resulting in a lack of comprehensiveness in monitoring the structure of the shallow silo cylinder.

[0004] 2. The existing reinforced concrete silo construction safety monitoring system cannot monitor the damage positions of each in-silo monitoring area, cannot obtain the damage area ratio corresponding to each in-silo monitoring area, and cannot real-time grasp the construction loss inside the shallow silo cylinder, thereby failing to improve the construction safety.

[0005] Therefore, the application provides a reinforced concrete silo construction safety monitoring system and method based on machine vision. SUMMARY

[0006] In view of the deficiencies of the prior art, the application aims to provide a reinforced concrete silo construction safety monitoring system and method based on machine vision, which improves the monitoring comprehensiveness and accuracy of the reinforced concrete silo construction safety monitoring system.

[0007] To achieve the above-mentioned purpose, the application adopts the following technical solution: a reinforced concrete silo construction safety monitoring system based on machine vision, comprising:

[0008] An out-of-silo monitoring module: marking the first to fifth visual monitoring planes outside the target shallow silo cylinder, setting a visual structure monitoring period, performing periodic visual image analysis on the first to fifth visual monitoring planes in the visual structure monitoring period, and obtaining the structure visual monitoring deviation corresponding to the target shallow silo cylinder according to the analysis result;

[0009] An in-silo monitoring module: dividing the internal area of the target shallow silo cylinder into a plurality of in-silo monitoring areas, monitoring the damage positions of each in-silo monitoring area, obtaining the damage area ratio corresponding to each in-silo monitoring area according to the monitoring result, and analyzing the in-silo area damage coefficient based on the damage area ratio;

[0010] A monitoring and early warning module: performing construction safety early warning on the target shallow silo cylinder according to the structure visual monitoring deviation and the in-silo area damage coefficient.

[0011] Further, the structural visual monitoring deviation is acquired, and the acquisition is specifically as follows:

[0012] The reinforced concrete shallow silo cylinder in the construction state in the construction area is acquired, and one target shallow silo cylinder is randomly selected from the acquired plurality of reinforced concrete shallow silo cylinders;

[0013] The target shallow silo cylinder is divided into monitoring areas to obtain a first visual monitoring plane to a fifth visual monitoring plane;

[0014] In the process of monitoring the structure of the target shallow silo cylinder, a time point corresponding to the current time is marked as a cycle end time point to create a fixed time length visual structure monitoring cycle, and a plurality of time intervals are set in the visual structure monitoring cycle. Structure monitoring time points, and a sample structure monitoring time point is randomly selected from the acquired plurality of structure monitoring time points;

[0015] The first visual monitoring plane in the visual structure monitoring cycle is monitored to obtain a first plane structure visual deviation;

[0016] The second visual monitoring plane in the visual structure monitoring cycle is monitored to obtain a second plane structure visual deviation;

[0017] The acquisition process of the second plane structure visual deviation is repeated to acquire the plane structure visual deviations corresponding to the third visual monitoring plane to the fifth visual monitoring plane to obtain a third plane structure visual deviation to a fifth plane structure visual deviation;

[0018] The first plane structure visual deviation, the second plane structure visual deviation, the third plane structure visual deviation, the fourth plane structure visual deviation, and the fifth plane structure visual deviation are calculated to obtain the structure visual monitoring deviation corresponding to the target shallow silo cylinder;

[0019] The structure visual monitoring deviation corresponding to the target shallow silo cylinder is calculated.

[0020] Further, the first visual monitoring plane to the fifth visual monitoring plane is acquired, and the acquisition is specifically as follows:

[0021] In the process of monitoring the construction safety of the target shallow silo cylinder, the upper bottom surface of the target shallow silo cylinder is marked as the first visual monitoring plane, the silo center point corresponding to the target shallow silo cylinder is acquired in the first visual monitoring plane to obtain the target shallow silo cylinder center, a straight line is made through the target shallow silo cylinder center in the first visual monitoring plane to obtain a first top surface characteristic straight line, and a straight line is made through the target shallow silo cylinder center and perpendicular to the first top surface characteristic straight line to obtain a second top surface characteristic straight line;

[0022] The first visual monitoring plane is divided into four feature monitoring areas by the first top surface feature straight line and the second top surface feature straight line, and the four divided feature monitoring areas are respectively named as a first feature monitoring area, a second feature monitoring area, a third feature monitoring area and a fourth feature monitoring area;

[0023] In the target silo cylinder, the silo cylinder side corresponding to the first feature monitoring area is marked as a second visual monitoring plane, the silo cylinder side corresponding to the second feature monitoring area is marked as a third visual monitoring plane, the silo cylinder side corresponding to the third feature monitoring area is marked as a fourth visual monitoring plane, and the silo cylinder side corresponding to the fourth feature monitoring area is marked as a fifth visual monitoring plane.

[0024] Further, the first plane structure visual deviation is obtained, and the specific process is as follows:

[0025] A visual image corresponding to the first visual monitoring plane at a sample structure monitoring time point is obtained, to obtain a first monitoring plane visual image corresponding to the sample structure monitoring time point;

[0026] In the first monitoring plane visual image corresponding to the sample structure monitoring time point, an outer silo circumference edge corresponding to the target silo cylinder is obtained, and a plurality of edge feature points are selected on the outer silo circumference edge, and a diameter of the outer silo circumference is drawn through each edge feature point, to obtain a plurality of outer silo diameters;

[0027] A target outer silo diameter designed for the target silo cylinder is obtained, to obtain an outer silo preset diameter, a difference between each outer silo diameter and the outer silo preset diameter is calculated, and an absolute value of the obtained difference is taken, to obtain a plurality of outer silo diameter deviations;

[0028] The obtained plurality of outer silo diameter deviations are compared in numerical value, and the outer silo diameter deviation with the largest numerical value is marked as a structure monitoring deviation corresponding to the first visual monitoring plane, and is named as a plane structure visual deviation corresponding to the sample structure monitoring time point;

[0029] The plane structure visual deviation corresponding to each structure monitoring time point is obtained respectively, to obtain a plurality of plane structure visual deviations, and an average number of the obtained plurality of plane structure visual deviations is calculated, to obtain the first plane structure visual deviation.

[0030] Further, the second plane structure visual deviation is obtained, and the specific process is as follows:

[0031] A visual image corresponding to the second visual monitoring plane at a sample structure monitoring time point is obtained, to obtain a second monitoring plane visual image corresponding to the sample structure monitoring time point;

[0032] Fitting the target shallow round bin side surface area in the second monitoring plane visual image as an image plane to obtain a sample image plane, and selecting a plurality of side surface structure points in the target shallow round bin side surface area in the sample image plane, and selecting a sample side surface structure point from the plurality of side surface structure points obtained;

[0033] Performing deviation analysis on the sample side surface structure point, and obtaining a position structure deviation corresponding to the sample side surface structure point according to the analysis result;

[0034] Obtaining the position structure deviation corresponding to each side surface structure point respectively to obtain a plurality of position structure deviations, and comparing the numerical values of the plurality of position structure deviations obtained to mark the position structure deviation with the largest numerical value as a second monitoring image visual deviation corresponding to a sample structure monitoring time point;

[0035] Obtaining the second monitoring image visual deviation corresponding to each structure monitoring time point respectively, and performing average number calculation on the plurality of second monitoring image visual deviations obtained to obtain a second plane structure visual deviation.

[0036] Further, the position structure deviation corresponding to the sample side surface structure point is obtained, and the specific process is as follows:

[0037] A plane rectangular coordinate system is created in the sample image plane to obtain a sample image plane rectangular coordinate system, the coordinates corresponding to the sample side surface structure point are obtained in real time in the sample image plane rectangular coordinate system to obtain a first structure point coordinate, and a preset coordinate corresponding to the sample side surface structure point is obtained to obtain a second structure point coordinate;

[0038] The first structure point coordinate (x1, y1) and the second structure point coordinate (x2, y2) are calculated to obtain the position structure deviation corresponding to the sample side surface structure point;

[0039] The position structure deviation corresponding to the sample side surface structure point is calculated.

[0040] Further, the in-bin area damage coefficient is obtained, and the specific process is as follows:

[0041] The target shallow round bin internal area is obtained, the target shallow round bin internal area is divided into a plurality of in-bin monitoring areas, and a sample in-bin monitoring area is marked in the plurality of in-bin monitoring areas obtained;

[0042] The sample in-bin monitoring area is monitored for a damaged area, and the damaged area area ratio corresponding to the sample in-bin monitoring area is obtained according to the monitoring result;

[0043] The process of obtaining the area ratio of the damaged area corresponding to the in-warehouse monitoring area of the sample warehouse is repeated to obtain the area ratio of the damaged area corresponding to each in-warehouse monitoring area, and a plurality of area ratios of the damaged area are obtained. The plurality of area ratios of the damaged area are compared in value, and the area ratio of the damaged area with the largest value is marked as the in-warehouse area damage coefficient.

[0044] Further, the area ratio of the damaged area corresponding to the in-warehouse monitoring area of the sample warehouse is obtained, specifically as follows:

[0045] An historical damaged image corresponding to the internal area of the target shallow circular silo is obtained, and the historical damaged image is used to train an image recognition model to obtain a damaged area recognition model.

[0046] The in-warehouse monitoring area of the sample warehouse is subjected to regional image acquisition to obtain an in-warehouse monitoring image of the sample warehouse. The damaged area recognition model is used to identify the in-warehouse monitoring image of the sample warehouse. The damaged area in the in-warehouse monitoring image of the sample warehouse is marked as a first type image area, and the undamaged area in the in-warehouse monitoring image of the sample warehouse is marked as a second type image area to obtain an in-warehouse area marking image.

[0047] In the in-warehouse area marking image, the first type image area and the second type image area are respectively subjected to pixel point filling. The number of pixel points filled in the first type image area is counted to obtain a first pixel point number value. The number of pixel points filled in the second type image area is counted to obtain a second pixel point number value. The ratio of the first pixel point number value to the second pixel point number value is calculated to obtain the area ratio of the damaged area corresponding to the in-warehouse monitoring area of the sample warehouse.

[0048] The process of obtaining the area ratio of the damaged area corresponding to the in-warehouse monitoring area of the sample warehouse is repeated to obtain the area ratio of the damaged area corresponding to each in-warehouse monitoring area, and a plurality of area ratios of the damaged area are obtained. The plurality of area ratios of the damaged area are compared in value, and the area ratio of the damaged area with the largest value is marked as the in-warehouse area damage coefficient.

[0049] Further, the construction safety warning of the target shallow circular silo is obtained, specifically as follows:

[0050] The structural visual monitoring deviation and the in-warehouse area damage coefficient are obtained respectively.

[0051] The structural visual monitoring deviation safety interval and the in-warehouse area damage coefficient safety interval are obtained respectively.

[0052] If the structural visual monitoring deviation is in the structural visual monitoring deviation safety interval, and the in-warehouse area damage coefficient is in the in-warehouse area damage coefficient safety interval, it is determined that the target shallow circular silo does not have construction safety hazards.

[0053] If the structural visual monitoring deviation is not in the structural visual monitoring deviation safety interval, and the in-warehouse area damage coefficient is not in the in-warehouse area damage coefficient safety interval, it is judged that the target shallow circular silo cylinder has construction safety hidden dangers, and a construction safety hidden danger early warning is issued.

[0054] If the structural visual monitoring deviation is not in the structural visual monitoring deviation safety interval, and the in-warehouse area damage coefficient is not in the in-warehouse area damage coefficient safety interval, it is judged that the target shallow circular silo cylinder has construction safety hidden dangers, and a construction safety hidden danger early warning is issued.

[0055] If the structural visual monitoring deviation is not in the structural visual monitoring deviation safety interval, and the in-warehouse area damage coefficient is not in the in-warehouse area damage coefficient safety interval, it is judged that the target shallow circular silo cylinder has construction safety hidden dangers, and a construction safety hidden danger early warning is issued.

[0056] A reinforced concrete silo construction safety monitoring method based on machine vision, comprising:

[0057] Step S1: marking the first to fifth visual monitoring planes outside the target shallow circular silo cylinder, and setting a visual structure monitoring period, periodically analyzing the visual images of the first to fifth visual monitoring planes in the visual structure monitoring period, and obtaining the structural visual monitoring deviation corresponding to the target shallow circular silo cylinder according to the analysis result;

[0058] Step S2: dividing the internal area of the target shallow circular silo cylinder into a plurality of in-warehouse monitoring areas, monitoring the damage positions of each in-warehouse monitoring area respectively, obtaining the damage area ratio corresponding to each in-warehouse monitoring area according to the monitoring result, and analyzing to obtain the in-warehouse area damage coefficient;

[0059] Step S3: construction safety early warning of the target shallow circular silo cylinder according to the structural visual monitoring deviation and the in-warehouse area damage coefficient.

[0060] As described above, due to the adoption of the above technical scheme, the beneficial effects of the present application are:

[0061] 1. The present application periodically monitors the multiple different visual monitoring planes outside the target shallow circular silo cylinder respectively, which can effectively improve the comprehensiveness and accuracy of the external structure monitoring of the shallow circular silo cylinder;

[0062] 2. The present application monitors the damage positions of each in-warehouse monitoring area, can obtain the damage area ratio corresponding to each in-warehouse monitoring area, thereby grasping the construction loss inside the shallow circular silo cylinder in real time, and can provide construction guarantee for the target shallow circular silo cylinder. BRIEF DESCRIPTION OF DRAWINGS

[0063] In order to facilitate those skilled in the art to understand, the present application will be further described below in conjunction with the drawings.

[0064] Figure 1 is a whole system block diagram of the present application;

[0065] Figure 2 is an implementation step diagram of the present application. DETAILED DESCRIPTION

[0066] The technical solutions of the present application will be described clearly and completely in combination with the embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0067] Embodiment one

[0068] Please refer to Figure 1 The present application provides a technical solution: a reinforced concrete silo construction safety monitoring system based on machine vision, comprising an outside monitoring module, an inside monitoring module, a monitoring and early warning module and a server, the outside monitoring module, the inside monitoring module and the monitoring and early warning module are connected with the server respectively, and the server controls the outside monitoring module, the inside monitoring module and the monitoring and early warning module respectively.

[0069] The outside monitoring module marks the first to fifth visual monitoring planes outside the target shallow circular silo cylinder, and sets a visual structure monitoring period, periodically analyzes the visual images of the first to fifth visual monitoring planes in the visual structure monitoring period, and obtains the corresponding structural visual monitoring deviation of the target shallow circular silo cylinder according to the analysis result.

[0070] Specifically as follows:

[0071] The reinforced concrete shallow circular silos in the construction area in the construction state are obtained, and a target shallow circular silo is arbitrarily selected from the obtained multiple reinforced concrete shallow circular silos.

[0072] It should be noted here that:

[0073] The reinforced concrete shallow circular silo referred to here is a kind of reinforced concrete silo, and the implementation object of the present embodiment is a shallow circular silo.

[0074] In the process of construction safety monitoring of the target shallow silo cylinder, the upper bottom surface of the target shallow silo cylinder is marked as a first visual monitoring plane, the center point of the circle corresponding to the target shallow silo cylinder is obtained in the first visual monitoring plane, the center of the target shallow silo cylinder is obtained, a straight line is made through the center of the target shallow silo cylinder in the first visual monitoring plane, a first top surface characteristic straight line is obtained, a straight line is made through the center of the target shallow silo cylinder and perpendicular to the first top surface characteristic straight line, a second top surface characteristic straight line is obtained;

[0075] The first visual monitoring plane is divided into four characteristic monitoring areas through the first top surface characteristic straight line and the second top surface characteristic straight line, and the four divided characteristic monitoring areas are respectively named as a first characteristic monitoring area, a second characteristic monitoring area, a third characteristic monitoring area and a fourth characteristic monitoring area;

[0076] In the target shallow silo cylinder, the side surface of the shallow silo cylinder corresponding to the first characteristic monitoring area is marked as a second visual monitoring plane, the side surface of the shallow silo cylinder corresponding to the second characteristic monitoring area is marked as a third visual monitoring plane, the side surface of the shallow silo cylinder corresponding to the third characteristic monitoring area is marked as a fourth visual monitoring plane, and the side surface of the shallow silo cylinder corresponding to the fourth characteristic monitoring area is marked as a fifth visual monitoring plane;

[0077] It should be noted here that:

[0078] In the present application, the first visual monitoring plane referred to here is the upper bottom surface of the target shallow silo cylinder, and the second visual monitoring plane, the third visual monitoring plane, the fourth visual monitoring plane and the fifth visual monitoring plane are respectively four side surfaces of the target shallow silo cylinder.

[0079] In the process of structural monitoring of the target shallow silo cylinder, a time point corresponding to the current time is marked as a period end time point to create a fixed length visual structural monitoring period, a plurality of structural monitoring time points with equal time intervals are set in the visual structural monitoring period, and a sample structural monitoring time point is selected from the obtained plurality of structural monitoring time points;

[0080] The first visual monitoring plane in the visual structural monitoring period is subjected to structural monitoring to obtain a first plane structural visual deviation;

[0081] Specifically as follows:

[0082] A visual image of the first visual monitoring plane corresponding to the sample structural monitoring time point is obtained to obtain a first monitoring plane visual image corresponding to the sample structural monitoring time point;

[0083] It should be noted here that:

[0084] In the present application, the first visual monitoring plane is the same as the image acquisition device installation position and the shooting parameter corresponding to the first monitoring plane visual image corresponding to each structural monitoring time point;

[0085] The shooting parameters referred to herein include, but are not limited to, focal length, brightness, and contrast.

[0086] In the first monitoring plane visual image corresponding to the sample structure monitoring time point, the outer warehouse circumferential edge corresponding to the target shallow circular warehouse cylinder is acquired, and a plurality of edge feature points are selected on the outer warehouse circumferential edge, and a diameter of the outer warehouse circumference is drawn through each edge feature point, thereby obtaining a plurality of outer warehouse diameters;

[0087] The target outer warehouse diameter designed for the target shallow circular warehouse cylinder is acquired to obtain an outer warehouse preset diameter, the difference between each outer warehouse diameter and the outer warehouse preset diameter is calculated, and the absolute value of the obtained difference is taken to obtain a plurality of outer warehouse diameter deviations;

[0088] It should be noted here that:

[0089] In the present application, the target shallow circular warehouse cylinder referred to herein is the outer warehouse diameter determined during design.

[0090] The numerical size of the obtained plurality of outer warehouse diameter deviations is compared, the outer warehouse diameter deviation with the largest numerical value is marked as the structural monitoring deviation corresponding to the first visual monitoring plane, and is named as the plane structure visual deviation corresponding to the sample structure monitoring time point;

[0091] Each plane structure visual deviation corresponding to each structural monitoring time point is acquired to obtain a plurality of plane structure visual deviations, and the average of the obtained plurality of plane structure visual deviations is calculated to obtain a first plane structure visual deviation;

[0092] The second visual monitoring plane in the visual structure monitoring period is monitored to obtain a second plane structure visual deviation;

[0093] The visual image of the second visual monitoring plane corresponding to the sample structure monitoring time point is acquired to obtain a second monitoring plane visual image corresponding to the sample structure monitoring time point;

[0094] The target shallow circular warehouse cylinder side surface region plane in the second monitoring plane visual image is fitted as an image plane to obtain a sample image plane, and a plurality of side surface structure points are selected in the target shallow circular warehouse cylinder side surface region in the sample image plane, and a sample side surface structure point is selected from the obtained plurality of side surface structure points;

[0095] It should be noted here that:

[0096] The side structure points referred to herein can be side component connection points of the target shallow round silo, and the side component connection points referred to herein include but are not limited to the welding or anchoring nodes of the reinforcing ribs and the silo wall, the fixing points of the ladders and the guardrails.

[0097] In the sample image plane, a plane rectangular coordinate system is created to obtain a sample image plane rectangular coordinate system, in which the coordinates corresponding to the sample side structure points are acquired in real time to obtain first structure point coordinates, and the preset coordinates corresponding to the sample side structure points are acquired to obtain second structure point coordinates;

[0098] It should be noted that:

[0099] The preset coordinates referred to herein are the coordinates of the positions of the sample side structure points determined in the design process in the sample image plane rectangular coordinate system.

[0100] The first structure point coordinates (x1, y1) and the second structure point coordinates (x2, y2) are calculated to obtain the position structure deviation corresponding to the sample side structure points.

[0101] The position structure deviation corresponding to the sample side structure points is calculated, and the specific formula is as follows:

[0102]

[0103] The position structure deviation corresponding to each side structure point is obtained to obtain a plurality of position structure deviations, and the numerical values of the plurality of position structure deviations are compared, and the position structure deviation with the largest numerical value is marked as the second monitoring image visual deviation corresponding to the sample structure monitoring time point.

[0104] The second monitoring image visual deviation corresponding to each structure monitoring time point is obtained, and the average of the plurality of second monitoring image visual deviations is calculated to obtain a second plane structure visual deviation.

[0105] The acquisition process of the second plane structure visual deviation is repeated to obtain the plane structure visual deviations corresponding to the third visual monitoring plane to the fifth visual monitoring plane to obtain the third plane structure visual deviation to the fifth plane structure visual deviation.

[0106] The first plane structure visual deviation, the second plane structure visual deviation, the third plane structure visual deviation, the fourth plane structure visual deviation and the fifth plane structure visual deviation are calculated to obtain the structure visual monitoring deviation corresponding to the target shallow round silo.

[0107] The structure visual monitoring deviation corresponding to the target shallow round silo is calculated, and the specific formula is as follows:

[0108]

[0109] Wherein, Spc is the structural visual monitoring deviation corresponding to the target shallow circular silo cylinder, Jp1 is the first plane structural visual deviation, Jp2 is the second plane structural visual deviation, Jp3 is the third plane structural visual deviation, Jp4 is the fourth plane structural visual deviation, Jp5 is the fifth plane structural visual deviation;

[0110] It should be noted here that:

[0111] In this application, the structural visual monitoring deviation referred to here is specifically the average of the structural visual deviation of the upper bottom surface of the target shallow circular silo cylinder and the structural visual deviation of the side surface;

[0112] In specific implementation, there are the following test data:

[0113] It is known that Jp1 is 3 cm, Jp2 is 3.5 cm, Jp3 is 4.6 cm, Jp4 is 5.3 cm, and Jp5 is 3.9 cm, so Spc can be calculated to be 3.6625.

[0114] The out-of-silo monitoring module acquires the structural visual monitoring deviation and transmits it to the monitoring and early warning module;

[0115] The in-silo monitoring module divides the internal area of the target shallow circular silo cylinder into a plurality of in-silo monitoring areas, respectively monitors the damage position of each in-silo monitoring area, acquires the damage area ratio corresponding to each in-silo monitoring area according to the monitoring result, and analyzes the in-silo area damage coefficient accordingly;

[0116] The internal area of the target shallow circular silo cylinder is acquired, and the internal area of the target shallow circular silo cylinder is divided into a plurality of in-silo monitoring areas, and a sample in-silo monitoring area is marked in the acquired plurality of in-silo monitoring areas;

[0117] The damage area of the sample in-silo monitoring area is monitored, and the damage area ratio corresponding to the sample in-silo monitoring area is acquired according to the monitoring result;

[0118] Specifically as follows:

[0119] The historical damage image corresponding to the internal area of the target shallow circular silo cylinder is acquired, the historical damage image is used to train an image recognition model, and a damage area recognition model is obtained;

[0120] It should be noted here that:

[0121] The image recognition model referred to here is specifically a convolutional neural network model;

[0122] An area image of the monitoring area in the sample bin is collected to obtain a monitoring image in the sample bin. A damaged area identification model is used to identify the monitoring image in the sample bin. The damaged area in the monitoring image in the sample bin is marked as a first type of image area, and the undamaged area in the monitoring image in the sample bin is marked as a second type of image area to obtain a bin area marking image.

[0123] It should be noted here that:

[0124] In this application, the damaged area referred to here includes but is not limited to cracks, concrete surface spalling, and protective layer spalling caused by steel bar corrosion.

[0125] In the bin area marking image, the first type of image area and the second type of image area are filled with pixel points respectively. The number of pixel points filled in the first type of image area is counted to obtain a first pixel point number value. The number of pixel points filled in the second type of image area is counted to obtain a second pixel point number value. The ratio of the first pixel point number value to the second pixel point number value is calculated to obtain a damaged area area ratio corresponding to the monitoring area in the sample bin.

[0126] The process of obtaining the damaged area area ratio corresponding to the monitoring area in the sample bin is repeated to obtain the damaged area area ratio corresponding to each monitoring area in the bin. A plurality of damaged area area ratios are obtained, and the plurality of damaged area area ratios are compared in value. The damaged area area ratio with the largest value is marked as a bin area damage coefficient.

[0127] The monitoring and early warning module performs construction safety early warning on the target silo cylinder according to the structural visual monitoring deviation and the bin area damage coefficient.

[0128] Specifically as follows:

[0129] The structural visual monitoring deviation and the bin area damage coefficient are obtained respectively.

[0130] The structural visual monitoring deviation safety interval and the bin area damage coefficient safety interval are obtained respectively.

[0131] If the structural visual monitoring deviation is in the structural visual monitoring deviation safety interval, and the bin area damage coefficient is in the bin area damage coefficient safety interval, it is determined that the target silo cylinder does not have construction safety hazards.

[0132] It should be noted here that:

[0133] The structural visual monitoring deviation safety interval is obtained, specifically as follows:

[0134] The lower limit of the structural visual monitoring deviation safety interval is 0, i.e., the target silo cylinder does not have any structural deviation.

[0135] Obtaining a plurality of shallow silo cylinders without safety hazards, obtaining the structural visual monitoring deviation corresponding to each shallow silo cylinder respectively, calculating the average of the obtained multiple structural visual monitoring deviations to obtain the structural visual monitoring deviation average, calculating the standard deviation of the obtained multiple structural visual monitoring deviations to obtain the structural visual monitoring deviation standard deviation, calculating the difference between the structural visual monitoring deviation average and the structural visual monitoring deviation standard deviation to obtain the upper limit of the structural visual monitoring deviation safety interval;

[0136] The in-silo area damage coefficient safety interval is obtained as follows:

[0137] The lower limit of the in-silo area damage coefficient safety interval referred to here is 0, that is, the target shallow silo cylinder does not have any area damage;

[0138] Obtaining a plurality of shallow silo cylinders without safety hazards, obtaining the in-silo area damage coefficient corresponding to each shallow silo cylinder respectively, calculating the average of the obtained multiple in-silo area damage coefficients to obtain the in-silo area damage coefficient average, calculating the standard deviation of the obtained multiple in-silo area damage coefficients to obtain the in-silo area damage coefficient standard deviation, calculating the difference between the in-silo area damage coefficient average and the structural visual monitoring deviation standard deviation to obtain the upper limit of the in-silo area damage coefficient safety interval;

[0139] If the structural visual monitoring deviation is not in the structural visual monitoring deviation safety interval, and the in-silo area damage coefficient is not in the in-silo area damage coefficient safety interval, it is judged that the target shallow silo cylinder has construction safety hazards, and a construction safety hazard warning is issued;

[0140] If the structural visual monitoring deviation is in the structural visual monitoring deviation safety interval, and the in-silo area damage coefficient is not in the in-silo area damage coefficient safety interval, it is judged that the target shallow silo cylinder has construction safety hazards, and a construction safety hazard warning is issued;

[0141] If the structural visual monitoring deviation is not in the structural visual monitoring deviation safety interval, and the in-silo area damage coefficient is not in the in-silo area damage coefficient safety interval, it is judged that the target shallow silo cylinder has construction safety hazards, and a construction safety hazard warning is issued.

[0142] It should be noted here that

[0143] The target shallow silo cylinder without construction safety hazards includes the boundary of the structural visual monitoring deviation safety interval and the boundary of the in-silo area damage coefficient safety interval.

[0144] In the present application, if the corresponding calculation formula appears, the above calculation formula is to calculate the value without dimension, and the weight coefficient, the proportion coefficient and other coefficients existing in the formula are set to quantify the result value of each parameter. The size of the weight coefficient and the proportion coefficient can affect the proportional relationship between the parameter and the result value.

[0145] Embodiment two

[0146] Please refer to Figure 2 Based on another concept of the same invention, a reinforced concrete silo construction safety monitoring method based on machine vision is proposed, which is applied to a reinforced concrete silo construction safety monitoring system based on machine vision. The monitoring method comprises the following steps:

[0147] Step S1: Mark the first to fifth visual monitoring planes outside the target shallow silo cylinder, and set a visual structure monitoring period. Periodic visual image analysis is performed on the first to fifth visual monitoring planes in the visual structure monitoring period, and the structure visual monitoring deviation corresponding to the target shallow silo cylinder is obtained according to the analysis result;

[0148] The step S1 further comprises the following steps:

[0149] The reinforced concrete shallow silo in the construction state in the construction area is obtained, and a target shallow silo is arbitrarily selected from the obtained plurality of reinforced concrete shallow silos;

[0150] The target shallow silo is divided into a monitoring area to obtain the first visual monitoring plane to the fifth visual monitoring plane;

[0151] Specifically as follows:

[0152] In the process of monitoring the construction safety of the target shallow silo, the upper bottom surface of the target shallow silo is marked as the first visual monitoring plane. In the first visual monitoring plane, the silo center point corresponding to the target shallow silo is obtained to obtain the target shallow silo center. In the first visual monitoring plane, any straight line is drawn through the target shallow silo center to obtain a first top surface characteristic straight line. A straight line is drawn through the target shallow silo center and perpendicular to the first top surface characteristic straight line to obtain a second top surface characteristic straight line;

[0153] The first visual monitoring plane is divided into four characteristic monitoring areas through the first top surface characteristic straight line and the second top surface characteristic straight line, and the four divided characteristic monitoring areas are named as the first characteristic monitoring area, the second characteristic monitoring area, the third characteristic monitoring area and the fourth characteristic monitoring area respectively;

[0154] In the target shallow silo cylinder, the side surface of the shallow silo cylinder corresponding to the first characteristic monitoring area is marked as the second visual monitoring plane, the side surface of the shallow silo cylinder corresponding to the second characteristic monitoring area is marked as the third visual monitoring plane, the side surface of the shallow silo cylinder corresponding to the third characteristic monitoring area is marked as the fourth visual monitoring plane, and the side surface of the shallow silo cylinder corresponding to the fourth characteristic monitoring area is marked as the fifth visual monitoring plane.

[0155] In the process of monitoring the structure of the target shallow silo cylinder, a time point corresponding to the current time is marked as a cycle end time point, a fixed time length visual structure monitoring cycle is created, a plurality of structure monitoring time points with equal time intervals are set in the visual structure monitoring cycle, and a sample structure monitoring time point is selected from the plurality of structure monitoring time points.

[0156] The first plane structure visual deviation is obtained by monitoring the structure of the first visual monitoring plane in the visual structure monitoring cycle.

[0157] Specifically as follows:

[0158] The first monitoring plane visual image corresponding to the sample structure monitoring time point is obtained by acquiring the visual image of the first visual monitoring plane at the sample structure monitoring time point.

[0159] In the first monitoring plane visual image corresponding to the sample structure monitoring time point, the outer silo circumference edge corresponding to the target shallow silo cylinder is acquired, a plurality of edge feature points are selected on the outer silo circumference edge, and a diameter of the outer silo circumference is drawn through each edge feature point to obtain a plurality of outer silo diameters.

[0160] The target outer silo diameter designed for the target shallow silo cylinder is acquired to obtain an outer silo preset diameter, the difference between each outer silo diameter and the outer silo preset diameter is calculated, and the absolute value of the obtained difference is taken to obtain a plurality of outer silo diameter deviations.

[0161] The plurality of outer silo diameter deviations are compared in numerical value, the outer silo diameter deviation with the largest numerical value is marked as the structure monitoring deviation corresponding to the first visual monitoring plane, and is named as the plane structure visual deviation corresponding to the sample structure monitoring time point.

[0162] The plane structure visual deviation corresponding to each structure monitoring time point is acquired respectively to obtain a plurality of plane structure visual deviations, and the average of the plurality of plane structure visual deviations is calculated to obtain the first plane structure visual deviation.

[0163] The second plane structure visual deviation is obtained by monitoring the structure of the second visual monitoring plane in the visual structure monitoring cycle.

[0164] Specifically as follows:

[0165] acquiring a visual image corresponding to the second visual monitoring plane at the sample structure monitoring time point to obtain a second monitoring plane visual image corresponding to the sample structure monitoring time point;

[0166] fitting a target shallow circular silo side surface region plane in the second monitoring plane visual image as an image plane to obtain a sample image plane, and selecting a plurality of side surface structure points in the target shallow circular silo side surface region in the sample image plane, and selecting a sample side surface structure point from the plurality of side surface structure points;

[0167] performing deviation analysis on the sample side surface structure point, and acquiring a position structure deviation corresponding to the sample side surface structure point according to the analysis result;

[0168] Specifically as follows:

[0169] In the sample image plane, a plane rectangular coordinate system is created to obtain a sample image plane rectangular coordinate system, the coordinates corresponding to the sample side surface structure point are acquired in real time in the sample image plane rectangular coordinate system to obtain first structure point coordinates, and preset coordinates corresponding to the sample side surface structure point are acquired to obtain second structure point coordinates;

[0170] The first structure point coordinates (x1, y1) and the second structure point coordinates (x2, y2) are used to calculate the position structure deviation corresponding to the sample side surface structure point;

[0171] The position structure deviation corresponding to the sample side surface point is calculated, and the specific formula is as follows:

[0172]

[0173] The position structure deviation corresponding to each side surface structure point is acquired respectively to obtain a plurality of position structure deviations, and the numerical value of the plurality of position structure deviations is compared, and the position structure deviation with the largest numerical value is marked as the second monitoring image visual deviation corresponding to the sample structure monitoring time point;

[0174] The second monitoring image visual deviation corresponding to each structure monitoring time point is acquired respectively, and the average of the plurality of second monitoring image visual deviations is calculated to obtain a second plane structure visual deviation;

[0175] The acquisition process of the second plane structure visual deviation is repeated to acquire the plane structure visual deviations corresponding to the third visual monitoring plane to the fifth visual monitoring plane to obtain a third plane structure visual deviation to a fifth plane structure visual deviation;

[0176] The first plane structure visual deviation, the second plane structure visual deviation, the third plane structure visual deviation, the fourth plane structure visual deviation and the fifth plane structure visual deviation are calculated to obtain the structural visual monitoring deviation corresponding to the target shallow circular silo cylinder;

[0177] The structural visual monitoring deviation corresponding to the target shallow circular silo cylinder is calculated, and the specific formula is as follows:

[0178]

[0179] Wherein, Spc is the structural visual monitoring deviation corresponding to the target shallow circular silo cylinder, Jp1 is the first plane structure visual deviation, Jp2 is the second plane structure visual deviation, Jp3 is the third plane structure visual deviation, Jp4 is the fourth plane structure visual deviation, and Jp5 is the fifth plane structure visual deviation.

[0180] Step S2: the internal area of the target shallow circular silo cylinder is divided into a plurality of in-silo monitoring areas, and the damage position of each in-silo monitoring area is monitored, the damage area ratio corresponding to each in-silo monitoring area is obtained according to the monitoring result, and the in-silo area damage coefficient is obtained according to the analysis result.

[0181] The step S2 further includes the following steps:

[0182] The internal area of the target shallow circular silo cylinder is obtained, and the internal area of the target shallow circular silo cylinder is divided into a plurality of in-silo monitoring areas, and a sample in-silo monitoring area is marked in the obtained plurality of in-silo monitoring areas.

[0183] The damage area of the sample in-silo monitoring area is monitored, and the damage area ratio corresponding to the sample in-silo monitoring area is obtained according to the monitoring result.

[0184] Specifically as follows:

[0185] The historical damage image corresponding to the internal area of the target shallow circular silo cylinder is obtained, the image recognition model is trained using the historical damage image, and the damage area recognition model is obtained.

[0186] The sample in-silo monitoring area is regionally imaged to obtain a sample in-silo monitoring image, the sample in-silo monitoring image is identified using the damage area recognition model, the damage area in the sample in-silo monitoring image is marked as a first type image area, and the undamaged area in the sample in-silo monitoring image is marked as a second type image area, to obtain an in-silo area marking image.

[0187] In the bin area marking image, the pixel points of the first type image area and the second type image area are filled respectively, the number of the pixel points filled in the first type image area is counted to obtain a first pixel point number value, the number of the pixel points filled in the second type image area is counted to obtain a second pixel point number value, and the ratio of the first pixel point number value to the second pixel point number value is calculated to obtain a damaged area ratio corresponding to the sample bin monitoring area;

[0188] The process of obtaining the damaged area ratio corresponding to the sample bin monitoring area is repeated, and the damaged area ratio corresponding to each bin monitoring area is obtained respectively to obtain a plurality of damaged area ratios, and the obtained plurality of damaged area ratios are compared in value, and the damaged area ratio with the largest value is marked as a bin area damage coefficient;

[0189] Step S3: construction safety warning is performed on the target silo according to the structural visual monitoring deviation and the bin area damage coefficient;

[0190] In the step S3, the following steps are further included:

[0191] The structural visual monitoring deviation and the bin area damage coefficient are obtained respectively;

[0192] The structural visual monitoring deviation safety interval and the bin area damage coefficient safety interval are obtained respectively;

[0193] If the structural visual monitoring deviation is in the structural visual monitoring deviation safety interval and the bin area damage coefficient is in the bin area damage coefficient safety interval, it is determined that the target silo does not have construction safety hazards;

[0194] If the structural visual monitoring deviation is not in the structural visual monitoring deviation safety interval and the bin area damage coefficient is not in the bin area damage coefficient safety interval, it is determined that the target silo has construction safety hazards, and a construction safety hazard warning is issued;

[0195] If the structural visual monitoring deviation is in the structural visual monitoring deviation safety interval and the bin area damage coefficient is not in the bin area damage coefficient safety interval, it is determined that the target silo has construction safety hazards, and a construction safety hazard warning is issued;

[0196] If the structural visual monitoring deviation is not in the structural visual monitoring deviation safety interval and the bin area damage coefficient is not in the bin area damage coefficient safety interval, it is determined that the target silo has construction safety hazards, and a construction safety hazard warning is issued.

[0197] The preferred embodiments of the application disclosed above are only to facilitate the elucidation of the application. The preferred embodiments do not describe all the details of the application and limit the application to the specific embodiments. Obviously, many modifications and variations can be made in light of the teachings above. The description is chosen and described in order to provide the best illustration of the application and its practical application to those skilled in the art and to enable those skilled in the art to best utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. A machine vision-based safety monitoring system for reinforced concrete silo construction, characterized in that, include: External monitoring module: Mark the first to fifth visual monitoring planes outside the target shallow circular silo, and set a visual structure monitoring cycle. Perform periodic visual image analysis on the first to fifth visual monitoring planes within the visual structure monitoring cycle, and obtain the structural visual monitoring deviation corresponding to the target shallow circular silo based on the analysis results. The structural visual monitoring deviation is the average of the visual deviation of the top and bottom surfaces of the target shallow circular silo and the visual deviation of the side surfaces. The in-warehouse monitoring module divides the internal area of ​​the target shallow circular silo into several in-warehouse monitoring areas, monitors the damaged parts of each in-warehouse monitoring area, obtains the damaged area ratio of each in-warehouse monitoring area based on the monitoring results, and analyzes to obtain the damage coefficient of the in-warehouse area. Monitoring and early warning module: Provides construction safety warnings for target shallow circular silos based on structural visual monitoring deviations and the damage coefficient of the silo area; Obtain the structural visual monitoring deviation and the damage coefficient of the warehouse area respectively; The safe range for structural visual monitoring deviation and the safe range for damage coefficient in the warehouse area were obtained respectively. If the structural visual monitoring deviation is within the safe range of structural visual monitoring deviation, and the damage coefficient of the warehouse area is within the safe range of the damage coefficient of the warehouse area, then it is determined that there is no construction safety hazard in the target shallow circular silo. The safe range for structural visual monitoring deviations is obtained as follows: The lower limit of the safety range for structural visual monitoring deviation involved here is 0, meaning that the target shallow circular silo does not have any structural deviation. Obtain several shallow circular silos without safety hazards, and obtain the structural visual monitoring deviation corresponding to each shallow circular silo. Calculate the average of the obtained structural visual monitoring deviations to obtain the average value of the structural visual monitoring deviations. Calculate the standard deviation of the obtained structural visual monitoring deviations to obtain the standard deviation of the structural visual monitoring deviations. Calculate the difference between the average value of the structural visual monitoring deviations and the standard deviation of the structural visual monitoring deviations to obtain the upper limit of the safe range of the structural visual monitoring deviations. The safe range for damage coefficients within the warehouse area is obtained as follows: The lower limit of the safe range for the damage coefficient of the warehouse area involved here is 0, which means that there is no damage in any area of ​​the target shallow circular silo. Several shallow circular silos without safety hazards are obtained. The damage coefficient of the silo area corresponding to each shallow circular silo is obtained. The average of the multiple silo area damage coefficients is calculated to obtain the average value of the silo area damage coefficient. The standard deviation of the multiple silo area damage coefficients is calculated to obtain the standard deviation of the silo area damage coefficient. The difference between the average value of the silo area damage coefficient and the standard deviation of the visual monitoring deviation is calculated to obtain the upper limit of the safe range of the silo area damage coefficient. If the structural visual monitoring deviation is not within the safe range of structural visual monitoring deviation, and the damage coefficient of the warehouse area is not within the safe range of the damage coefficient of the warehouse area, then it is determined that there is a construction safety hazard in the target shallow circular silo, and a construction safety hazard warning is issued. If the structural visual monitoring deviation is within the safe range of structural visual monitoring deviation, and the damage coefficient of the warehouse area is not within the safe range of the damage coefficient of the warehouse area, then it is determined that there is a construction safety hazard in the target shallow circular silo, and a construction safety hazard warning is issued. If the structural visual monitoring deviation is not within the safe range of structural visual monitoring deviation, and the damage coefficient of the warehouse area is not within the safe range of the damage coefficient of the warehouse area, then it is determined that there is a construction safety hazard in the target shallow circular silo, and a construction safety hazard warning is issued.

2. The machine vision-based reinforced concrete silo construction safety monitoring system according to claim 1, characterized in that, The deviations in the structural visual monitoring are obtained as follows: Acquire the reinforced concrete shallow circular silos that are under construction within the construction area, and arbitrarily select one target shallow circular silo from the multiple acquired reinforced concrete shallow circular silos. The target shallow circular silo is divided into monitoring areas to obtain the first visual monitoring plane to the fifth visual monitoring plane; In the process of structural monitoring of the target shallow cylindrical silo, a visual structural monitoring cycle is created, and several structural monitoring time points with equal time intervals are set within the visual structural monitoring cycle. Then, a sample structural monitoring time point is randomly selected from the multiple structural monitoring time points obtained. Structural monitoring is performed on the first visual monitoring plane during the visual structure monitoring cycle to obtain the visual deviation of the first plane structure. Structural monitoring is performed on the second visual monitoring plane during the visual structure monitoring cycle to obtain the visual deviation of the second plane structure. Repeat the process of acquiring the visual deviation of the second planar structure, and acquire the visual deviation of the planar structure corresponding to the third to fifth visual monitoring planes respectively, to obtain the visual deviation of the third to fifth planar structures. The visual deviation of the first, second, third, fourth, and fifth planar structures is calculated to obtain the structural visual monitoring deviation corresponding to the target shallow circular silo. The structural visual monitoring deviation corresponding to the target shallow circular silo is calculated.

3. The machine vision-based reinforced concrete silo construction safety monitoring system according to claim 2, characterized in that, The data is acquired from the first visual monitoring plane to the fifth visual monitoring plane, as detailed below: During the construction safety monitoring of the target shallow circular silo, the top surface of the target shallow circular silo is marked as the first visual monitoring plane. Within the first visual monitoring plane, the center point of the circular silo corresponding to the target shallow circular silo is obtained to obtain the center of the target shallow circular silo. Within the first visual monitoring plane, any straight line is drawn through the center of the target shallow circular silo to obtain the first top surface feature line. A straight line perpendicular to the first top surface feature line is drawn through the center of the target shallow circular silo to obtain the second top surface feature line. The first visual monitoring plane is divided into four feature monitoring regions by the first top surface feature line and the second top surface feature line, and the four feature monitoring regions are named the first feature monitoring region, the second feature monitoring region, the third feature monitoring region and the fourth feature monitoring region, respectively. In the target shallow circular silo, the side of the shallow circular silo corresponding to the first feature monitoring area is marked as the second visual monitoring plane, the side of the shallow circular silo corresponding to the second feature monitoring area is marked as the third visual monitoring plane, the side of the shallow circular silo corresponding to the third feature monitoring area is marked as the fourth visual monitoring plane, and the side of the shallow circular silo corresponding to the fourth feature monitoring area is marked as the fifth visual monitoring plane.

4. The machine vision-based reinforced concrete silo construction safety monitoring system according to claim 2, characterized in that, The visual deviation of the first planar structure is obtained as follows: The visual image of the first visual monitoring plane at the time point of sample structure monitoring is acquired to obtain the visual image of the first monitoring plane at the time point of sample structure monitoring. In the first monitoring plane visual image corresponding to the time point of sample structure monitoring, the outer circumferential edge of the target shallow circular silo is acquired, and several edge feature points are selected on the outer circumferential edge. The diameter of the outer circumference is obtained through each edge feature point, resulting in multiple outer circumferential diameters. The target outer circle diameter of the target shallow circular silo is obtained, the preset outer circle diameter is obtained, the difference between each outer circle diameter and the preset outer circle diameter is calculated, and the absolute value of the obtained difference is taken to obtain multiple outer circle diameter deviations. The numerical values ​​of the multiple outer chamber circle diameter deviations are compared. The outer chamber circle diameter deviation with the largest value is marked as the structural monitoring deviation corresponding to the first visual monitoring plane, and it is named the planar structural visual deviation corresponding to the sample structural monitoring time point. The visual deviation of the planar structure at each monitoring time point is acquired to obtain multiple visual deviations of the planar structure. The average of the multiple visual deviations of the planar structure is calculated to obtain the first visual deviation of the planar structure.

5. The machine vision-based reinforced concrete silo construction safety monitoring system according to claim 2, characterized in that, The visual deviation of the second-plane structure is obtained as follows: The visual image of the second visual monitoring plane at the time point corresponding to the sample structure monitoring is obtained, thus obtaining the visual image of the second monitoring plane at the time point corresponding to the sample structure monitoring. The plane of the side region of the target shallow cylindrical silo in the second monitoring plane visual image is fitted to the image plane to obtain the sample image plane. Several side structure points are selected in the side region of the target shallow cylindrical silo in the sample image plane, and one sample side structure point is selected from the several side structure points obtained. Deviation analysis is performed on the side structural points of the sample, and the positional structural deviations corresponding to the side structural points of the sample are obtained based on the analysis results. The positional structural deviation corresponding to each side structural point is acquired to obtain multiple positional structural deviations. The numerical values ​​of the acquired multiple positional structural deviations are compared, and the positional structural deviation with the largest value is marked as the second monitoring image visual deviation corresponding to the sample structure monitoring time point. The visual deviation of the second monitoring image corresponding to each structural monitoring time point is acquired, and the average of the obtained visual deviations of multiple second monitoring images is calculated to obtain the visual deviation of the second planar structure.

6. The machine vision-based reinforced concrete silo construction safety monitoring system according to claim 1, characterized in that, The positional structural deviations corresponding to the structural points on the side of the sample are obtained, as follows: In the sample image plane, a Cartesian coordinate system is created to obtain the sample image Cartesian coordinate system. In the sample image Cartesian coordinate system, the coordinates corresponding to the side structure points of the sample are acquired in real time to obtain the coordinates of the first structure point. The preset coordinates corresponding to the side structure points of the sample are acquired to obtain the coordinates of the second structure point. The positional structural deviations of the sample side structural points are obtained by calculating the coordinates of the first structural point (x1, y1) and the second structural point (x2, y2). The positional structural deviations corresponding to the structural points on the side of the sample are calculated.

7. The machine vision-based reinforced concrete silo construction safety monitoring system according to claim 1, characterized in that, The damage coefficient of the warehouse area was obtained as follows: The internal region of the target shallow circular silo is acquired and divided into several monitoring areas. A sample monitoring area is marked within each of the acquired monitoring areas. Damaged areas are monitored within the sample chamber, and the area ratio of damaged areas corresponding to the monitored areas within the sample chamber is obtained based on the monitoring results. Repeat the process of obtaining the area ratio of the damaged area corresponding to the monitoring area in the sample warehouse. Obtain the area ratio of the damaged area corresponding to each monitoring area in the warehouse to obtain multiple area ratios of the damaged area. Compare the values ​​of the multiple area ratios of the damaged area and mark the area ratio of the damaged area with the largest value as the damage coefficient of the area in the warehouse.

8. The machine vision-based reinforced concrete silo construction safety monitoring system according to claim 7, characterized in that, The area ratio of the damaged area corresponding to the monitoring area within the sample chamber was obtained, as follows: Obtain historical damage images corresponding to the internal region of the target shallow circular silo, use the historical damage images to train an image recognition model, and obtain a damage area recognition model; Regional images are acquired in the monitoring area within the sample chamber to obtain monitoring images within the sample chamber. A damaged area recognition model is used to identify the monitoring images within the sample chamber. Damaged areas in the monitoring images within the sample chamber are marked as first-type image areas, and undamaged areas in the monitoring images within the sample chamber are marked as second-type image areas, resulting in labeled images of the areas within the chamber. In the marked image of the warehouse area, pixels are filled into the first type of image area and the second type of image area respectively. The number of pixels filled into the first type of image area is counted to obtain the first pixel count value. The number of pixels filled into the second type of image area is counted to obtain the second pixel count value. The ratio of the first pixel count value to the second pixel count value is calculated to obtain the area ratio of the damaged area corresponding to the monitoring area in the sample warehouse.

9. A machine vision-based method for monitoring the construction safety of reinforced concrete silos, applicable to the machine vision-based reinforced concrete silo construction safety monitoring system described in any one of claims 1-8, characterized in that, The construction safety monitoring method for reinforced concrete silos includes: Step S1: Mark the first to fifth visual monitoring planes outside the target shallow circular silo, and set a visual structure monitoring cycle. Perform periodic visual image analysis on the first to fifth visual monitoring planes within the visual structure monitoring cycle, and obtain the structural visual monitoring deviation corresponding to the target shallow circular silo based on the analysis results. Step S2: Divide the internal area of ​​the target shallow circular silo into several monitoring areas, monitor the damaged parts of each monitoring area, obtain the damaged area ratio of each monitoring area based on the monitoring results, and analyze the damage coefficient of the internal area accordingly. Step S3: Based on the structural visual monitoring deviation and the damage coefficient of the silo area, conduct construction safety early warning for the target shallow circular silo.

Images