Engineering safety supervision system and method based on artificial intelligence

Through an engineering safety supervision system based on artificial intelligence, using image analysis and multiple detection strategies, the problem of inefficient safety monitoring of traditional foundation pit support structures is solved, and the comprehensive, accurate and efficient foundation pit support structures is achieved. Safety monitoring ensures the safety of foundation pit projects.

CN118968313AActive Publication Date: 2024-11-15SHENZHEN YUMITIAN TECHNOLOGY CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411422816.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-12
Publication Date
2024-11-15
Estimated Expiration
2044-10-12

AI Technical Summary

Technical Problem

The traditional safety monitoring method of foundation pit support structures is inefficient and cannot achieve all-round and real-time monitoring, which poses safety hazards, especially in complex construction environments that are difficult to detect hidden dangers.

Method used

Adopt an engineering safety supervision system based on artificial intelligence to determine the effectiveness of the support structure by acquiring and analyzing the images of the support structure in the foundation pit and implementing multiple detection strategies (distance detection, contact detection, distance comparison, attachment detection, etc.) and stability to ensure that each support structure meets safety standards.

Benefits of technology

Achieve multi-level and multi-angle detection of foundation pit support structures, ensuring its stability and safety, reducing safety hazards, improving the accuracy and efficiency of detection, and avoiding missed inspections and errors caused by complex structural layout. Verdict.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118968313A_ABST
    Figure CN118968313A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of engineering safety supervision, and discloses an artificial intelligence-based engineering safety supervision system and method, and the method comprises the steps: obtaining and numbering a supporting structure in a foundation pit; each supporting structure comprises two stand columns and two cross braces which are marked as a first stand column, a second stand column, a first cross brace and a second cross brace respectively; marking an image acquisition point, and obtaining a support image acquisition position; collecting a detection image, executing distance detection, and reselecting structure detection if the detection image is marked; if not, contact detection is executed, and whether marking is conducted or not is judged; if not, acquiring a contact point, and executing a distance comparison strategy; if not, executing attachment detection; repairing and detecting the marked supporting structures until all the supporting structures are effective; according to the method, each part of the supporting structure is ensured to meet the safety standard through various detection strategies; by means of distance detection, contact detection and the like, the stand column is attached to the foundation pit wall, the cross brace is tightly connected with the stand column, and stability is guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of engineering safety supervision, and specifically to an engineering safety supervision system and method based on artificial intelligence. Background Art

[0002] Foundation pit engineering is a key link in building construction, which directly affects the stability and safety of buildings. Foundation pit construction usually adopts support structures to maintain the stability of pit walls and avoid collapse accidents. However, with the expansion of project scale and increase of complexity, traditional support structure safety supervision methods have many limitations, and it is difficult to efficiently monitor the safety status of support structures in foundation pits, which makes foundation pit engineering have certain safety hazards. Therefore, how to effectively monitor the safety status of support structures during foundation pit construction and discover safety hazards in time has become an important issue in engineering safety supervision.

[0003] At present, the safety monitoring of foundation pit support structures mainly relies on manual inspections and simple instrument monitoring; however, manual inspections are inefficient and are often unable to achieve all-round, real-time monitoring due to environmental and human resource limitations; in addition, the accuracy of manual inspections is affected by the experience of inspectors, and there is subjectivity and risk of omissions; in complex construction environments, inspectors often cannot access certain hidden parts of the foundation pit, which can easily lead to missed inspections of safety hazards.

[0004] Although traditional monitoring instruments, such as displacement sensors and stress strain gauges, can monitor local locations, these instruments need to be deployed in advance and cannot fully cover the support structure of the entire foundation pit, and their detection range has certain limitations.

[0005] To this end, this method proposes an engineering safety supervision system and method based on artificial intelligence. By collecting images of the support structure in the foundation pit and analyzing the images, it is possible to detect whether the support structure provides effective support to the foundation pit wall, thereby ensuring the safety of the project. Summary of the invention

[0006] The present invention provides an engineering safety supervision system and method based on artificial intelligence, which promotes solving the problems mentioned in the above background technology.

[0007] The present invention provides the following technical solution: an engineering safety supervision method based on artificial intelligence, comprising:

[0008] Obtain each supporting structure in the foundation pit, and number the supporting structures in sequence according to the arrangement order of the supporting structures in the foundation pit;

[0009] Each support structure in the foundation pit consists of two columns and two cross braces;

[0010] For each supporting structure, its two columns are respectively recorded as the first column and the second column, and its two cross braces are respectively recorded as the first cross brace and the second cross brace;

[0011] Mark the image acquisition points according to the support structure;

[0012] According to the position of the supporting structure in the foundation pit and the image acquisition points on the supporting structure, the supporting image acquisition position is obtained;

[0013] For each supporting structure, an image of the supporting structure is collected at a supporting image collection position, and recorded as a detection image of the supporting structure;

[0014] According to the detection image of the support structure, the distance detection strategy is executed, and it is determined whether to mark the support structure;

[0015] If the supporting structure is marked, select a new supporting structure for testing;

[0016] If the supporting structure is not marked, the contact detection strategy is executed according to the detection image of the supporting structure;

[0017] According to the contact detection strategy, determine whether to mark the supporting structure;

[0018] If the supporting structure is marked, select a new supporting structure for testing;

[0019] If the supporting structure is not marked, the contact point of the supporting structure is obtained according to the detection image of the supporting structure, and a distance comparison strategy is performed on the supporting structure;

[0020] According to the distance comparison strategy, determine whether to mark the supporting structure;

[0021] If the supporting structure is marked, select a new supporting structure for testing;

[0022] If the supporting structure is not marked, an attachment detection strategy is performed on the supporting structure, and according to the attachment detection strategy, it is determined whether to mark the supporting structure;

[0023] The marked support structures shall be repaired and tested again until all the support structures in the foundation pit are recorded as effective supports.

[0024] Optionally, obtaining each supporting structure in the foundation pit and sequentially numbering the supporting structures according to the arrangement order of the supporting structures in the foundation pit includes:

[0025] Select one end of the foundation pit as the starting end, and the other end as the ending end;

[0026] The supporting structures in the foundation pit are numbered in sequence from the starting end to the ending end;

[0027] Each supporting structure in the foundation pit shall be inspected in turn according to the numbering order of the supporting structures.

[0028] Optionally, marking image acquisition points on the supporting structure according to the supporting structure includes:

[0029] For each supporting structure:

[0030] Obtain the first column and the second column and the vertical axis respectively, and record them as the first axis and the second axis;

[0031] Intercepting a portion of the first axis within the first column to form a first line segment;

[0032] Cutting off the portion of the second axis within the second column to form a second line segment;

[0033] Obtain the midpoints of the first line segment and the second line segment respectively, and connect them to form a reference line;

[0034] Get the midpoint of the reference line and record it as the reference midpoint;

[0035] The location of the reference midpoint is used as the image acquisition point of the support structure.

[0036] Optionally, acquiring the support image acquisition position according to the position of the support structure in the foundation pit and the image acquisition point on the support structure includes:

[0037] For each supporting structure in the foundation pit, the bottom surface of the foundation pit between every two adjacent supporting structures is intercepted;

[0038] Obtain the center point of the intercepted foundation pit bottom surface as a marking point;

[0039] For each supporting structure, the marking points located on both sides of the supporting structure and closest to the supporting structure are used as image marking points of the supporting structure;

[0040] Obtain a plane that passes through the image acquisition point of the support structure and is parallel to the bottom surface of the foundation pit, and record the plane as a reference plane;

[0041] Draw perpendicular lines to the bottom of the foundation pit through the image marking points respectively, and intersect with the reference plane respectively. The two intersection points obtained are used as the support image acquisition positions of the support structure.

[0042] Optionally, for the support structure, collecting the support structure image at the support image collection position, recorded as the detection image of the support structure, includes:

[0043] For each supporting structure:

[0044] At the support image acquisition positions on both sides of the support structure, respectively, an image of the support structure is acquired in a direction pointing to the image acquisition point of the support structure;

[0045] The two acquired images are used as detection images of the support structure.

[0046] Optionally, executing the distance detection strategy according to the detection image of the support structure and determining whether to mark the support structure includes:

[0047] A distance detection strategy is applied to each detection image of the support structure;

[0048] The two sides of the foundation pit wall are marked as the left foundation pit wall and the right foundation pit wall, with the direction from the first column to the second column;

[0049] According to the detection image of the support structure, the edge contour line of the first column close to the left foundation pit wall is taken as the first column line;

[0050] Take any point on the first column line, and use the point as the starting point to draw a line segment parallel to the bottom surface of the foundation pit, connected to the left foundation pit wall and perpendicular to the left foundation pit wall, which is recorded as the first connecting vertical line;

[0051] Obtain the length of the first connecting perpendicular line, recorded as the first connecting line distance;

[0052] If there is a first connection line distance that is not zero, the support structure is marked;

[0053] If all the first connection line distances that exist are zero, then based on the detection image of the support structure, the edge contour line of the second column close to the right foundation pit wall is taken as the second column line;

[0054] Take any point on the second column line, and use the point as the starting point to draw a line segment parallel to the bottom of the foundation pit and perpendicular to the right side of the foundation pit wall, which is recorded as the second connecting vertical line;

[0055] Obtain the length of the second connecting perpendicular line, recorded as the second connecting line distance;

[0056] If there is a second connection line with a distance not equal to zero, the support structure is marked;

[0057] If all the first connection distances that exist are zero, the support structure is recorded as the first effective support;

[0058] If the distance detection strategy performed according to the two detection images of the support structure both records the support structure as the first effective support, the contact detection strategy is executed on the support structure.

[0059] Optionally, if the supporting structure is not marked, executing a contact detection strategy according to a detection image of the supporting structure includes:

[0060] A contact detection strategy is applied to each detection image of the support structure;

[0061] According to the detection image of the supporting structure, identifying and detecting whether the two ends of the first cross brace are in contact with the first column and the second column respectively;

[0062] If there is an end of the first cross brace that is not in contact with the first column or the second column, the supporting structure is marked;

[0063] If the two ends of the first cross brace are in contact with the first column and the second column respectively, then according to the detection image of the support structure, identifying and detecting whether the two ends of the second cross brace are in contact with the first column and the second column respectively;

[0064] If there is an end of the second cross brace that is not in contact with the first column or the second column, the supporting structure is marked;

[0065] If the two ends of the second cross brace are in contact with the first column and the second column respectively, the support structure is recorded as the second effective support;

[0066] If the contact detection strategies performed according to the two detection images of the support structure both record the support structure as the second effective support, the distance comparison strategy is executed on the support structure.

[0067] Optionally, if the supporting structure is not marked, obtaining the contact point of the supporting structure according to the detection image of the supporting structure, and executing a distance comparison strategy for the supporting structure, including:

[0068] A distance comparison strategy is performed on each detection image of the support structure;

[0069] In the detection image, the line segment where the horizontal brace and the vertical column are in contact is recorded as the contact line between the horizontal brace and the vertical column;

[0070] According to the detection image, contact lines between the first cross brace and the first column and the second column are respectively obtained and recorded as a first contact line and a second contact line;

[0071] According to the detection image, the contact lines between the second cross brace and the first column and the second column are respectively obtained and recorded as the third contact line and the fourth contact line respectively;

[0072] Obtaining midpoints of the first contact line, the second contact line, the third contact line, and the fourth contact line, respectively, and recording them as the first contact point, the second contact point, the third contact point, and the fourth contact point, respectively;

[0073] Obtain vertical distances between the first contact point, the second contact point, the third contact point and the fourth contact point and the bottom surface of the foundation pit respectively, and record them as the first distance, the second distance, the third distance and the fourth distance respectively;

[0074] comparing the first distance and the second distance;

[0075] If the first distance is not equal to the second distance, the support structure is marked;

[0076] If the first distance is equal to the second distance, then compare the third distance and the fourth distance;

[0077] If the third distance is not equal to the fourth distance, the support structure is marked;

[0078] If the third distance is equal to the fourth distance, the support structure is recorded as the third effective support;

[0079] If the support structure is recorded as the third effective support according to the contact distance comparison strategy performed on the two detection images of the support structure, the attachment detection strategy is executed on the support structure.

[0080] Optionally, if the supporting structure is not marked, performing an attachment detection strategy on the supporting structure, and judging whether to mark the supporting structure according to the attachment detection strategy, includes:

[0081] The attachment detection strategy is applied to each detection image of the support structure;

[0082] According to the detection image, identify whether there are objects other than the supporting structure components attached to the supporting structure;

[0083] If present, the supporting structure is marked;

[0084] If it does not exist, the support structure is recorded as the fourth effective support;

[0085] If the attachment detection strategies performed according to the two detection images of the support structure both record the support structure as the fourth effective support, the support structure is recorded as the effective support structure, and the next support structure is selected for detection.

[0086] A system for implementing the engineering safety supervision method based on artificial intelligence, comprising:

[0087] Image acquisition module: used to acquire detection images of support structures;

[0088] Point marking module: used to calibrate image acquisition points and contact points;

[0089] Distance acquisition module: used to obtain the distance between the contact point and the bottom surface of the foundation pit, and the distance between the column and the foundation pit wall;

[0090] Numbering module: Sequentially number the supporting structures in the foundation pit to determine the inspection order;

[0091] Image recognition module: determines whether the horizontal brace and the column of the supporting structure are in contact based on the detection image;

[0092] Distance comparison module: used to compare the first distance, the second distance, the third distance and the fourth distance to determine whether the distances are equal;

[0093] Intelligent recognition module: Through artificial intelligence technology, the detection image is recognized to identify whether there are objects other than support structure components attached to the support structure.

[0094] The present invention has the following beneficial effects:

[0095] 1. The support structure is composed of multiple parts such as columns and cross braces. This method uses a variety of detection strategies to detect each component to ensure that each part meets safety standards. The combined effect of multiple strategies such as distance detection, contact detection, and distance comparison can ensure the stability of the support structure from multiple angles. Whether the columns are in contact with the foundation pit wall and whether the cross braces are tightly connected to the columns are important factors affecting the stability of the support structure. Through multi-level and multi-angle detection methods, it is comprehensively guaranteed that every detail of the support structure is in a safe state.

[0096] 2. By obtaining each supporting structure in the foundation pit and numbering them according to the arrangement order of the supporting structures, the entire inspection process is organized and traceable; this sequential numbering can help supervisors clearly identify the location of each supporting structure and conduct inspections in sequence without missing any supporting structure; for large-scale foundation pit construction projects, there are a large number of supporting structures, and sequential numbering can effectively prevent confusion caused by complex structural layouts, ensuring that each supporting structure can be inspected and recorded; in addition, sequential numbering also helps with data management and subsequent maintenance work. Problem areas can be quickly located through numbering, improving maintenance efficiency.

[0097] 3. By marking the image acquisition points for each support structure, it can be ensured that the acquisition of the detection image has a unified standard and reduce the deviation caused by human operation; the marking of the image acquisition points not only enables the detection image to clearly capture the key parts of the support structure, but also improves the accuracy of subsequent image analysis. The marked image acquisition points are usually based on the midpoint of the geometric center of the columns and cross braces of the support structure, which can better reflect the overall state of the structure, effectively avoid the problem of local information loss when acquiring images, and improve the reliability of the detection results.

[0098] 4. By collecting images on both sides of the support structure and analyzing them, the status information of the support structure can be fully obtained from different angles to ensure the accuracy and completeness of the detection. There may be different stress concentrations or force conditions on both sides of the support structure. The image on a single side may not fully reflect the overall structure. By collecting images on both sides, more details can be captured, especially when detecting the contact and distance between the cross brace and the column, and the support structure and the foundation pit wall. The dual-sided perspective can effectively avoid blind spots and missed detections. In addition, the images on both sides verify each other, which improves the reliability of detection and reduces misjudgments caused by factors such as angle or light. At the same time, this dual-sided acquisition method can more comprehensively detect the offset, deformation or looseness of the support structure, provide a full range of detection perspectives, ensure the stability of the support structure in the foundation pit, and improve the accuracy and efficiency of engineering safety supervision.

[0099] 5. The stability of the support structure is directly related to the safety of the foundation pit project. By executing the distance detection strategy on the detection image, the distance change between the support structure and the foundation pit wall can be discovered in time, so as to know whether there is a risk of displacement or tilt of the support structure. The distance detection strategy can detect subtle changes between the columns and the foundation pit wall, ensuring that the columns of the support structure are in full contact with the foundation pit wall, and that the pillars provide effective support to the foundation pit wall, thereby ensuring the safety of the foundation pit project.

[0100] 6. The contact detection strategy ensures that the various components of the support structure are closely connected to each other by analyzing the contact between the cross braces and the columns. This strategy effectively avoids the failure risks of the components in the support structure due to improper installation or looseness, ensures that the cross braces are in contact with the pillars, supports the pillars, and effectively supports the foundation pit wall, ensuring the safety of the foundation pit project. If looseness or non-contact between the cross braces and the columns is detected, the system will mark the structure and perform repair operations in the subsequent process. Through this strategy, improper installation problems can be discovered and corrected in a timely manner, ensuring the overall stability of the support structure and reducing the probability of accidents.

[0101] 7. Distance comparison strategy ensures the stability of the support structure by comparing the vertical distance between the contact point of the support structure and the bottom of the foundation pit; by comparing the distance, it can be ensured that the cross brace and the column of the support structure are basically in a vertical state, so that the force generated by the column on the cross brace is on a horizontal line, so that the cross brace can play the best supporting role for the column, and the support structure can play the greatest supporting effect on the foundation pit wall, ensuring the continuous stability of the support system.

[0102] 8. The attachment detection strategy mainly detects whether there are any redundant attachments or other objects that affect the safety of the structure on the support structure. This strategy effectively ensures the safety and stability of the support structure and avoids heavy objects being carried or hung on the support structure, which may interfere with or damage the support structure. If there are attachments on the support structure, the attachments may affect the bearing capacity of the support structure or cause stress concentration, thereby affecting the overall stability. Through attachment detection, these potential hidden dangers can be effectively eliminated to ensure that the support structure plays its supporting role under ideal conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0103] Figure 1 It is a schematic diagram of the support structure of the present invention.

[0104] Figure 2 It is a line segment schematic diagram of the contact part of the present invention.

[0105] Figure 3 Schematic diagram of marking points of the present invention. DETAILED DESCRIPTION

[0106] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0107] Embodiment 1, an engineering safety supervision method based on artificial intelligence, comprising:

[0108] Obtain each supporting structure in the foundation pit, and number the supporting structures in sequence according to the arrangement order of the supporting structures in the foundation pit;

[0109] Each support structure in the foundation pit consists of two columns and two cross braces;

[0110] Reference Figure 1 , for each supporting structure, its two columns are respectively recorded as the first column and the second column, and its two cross braces are respectively recorded as the first cross brace and the second cross brace;

[0111] Mark the image acquisition points according to the support structure;

[0112] According to the position of the supporting structure in the foundation pit and the image acquisition points on the supporting structure, the supporting image acquisition position is obtained;

[0113] For each supporting structure, an image of the supporting structure is collected at a supporting image collection position, and recorded as a detection image of the supporting structure;

[0114] According to the detection image of the support structure, the distance detection strategy is executed, and it is determined whether to mark the support structure;

[0115] If the supporting structure is marked, a new supporting structure is selected for detection; the reselection of a supporting structure for detection means selecting the next supporting structure according to the serial number sequence of the supporting structures, and performing distance detection strategy, contact detection strategy, distance detection strategy and attachment detection strategy on the supporting structure in sequence;

[0116] If the supporting structure is not marked, the contact detection strategy is executed according to the detection image of the supporting structure;

[0117] According to the contact detection strategy, determine whether to mark the supporting structure;

[0118] If the supporting structure is marked, a new supporting structure is selected for detection; the reselection of a supporting structure for detection means selecting the next supporting structure according to the serial number sequence of the supporting structures, and performing distance detection strategy, contact detection strategy, distance detection strategy and attachment detection strategy on the supporting structure in sequence;

[0119] If the supporting structure is not marked, the contact point of the supporting structure is obtained according to the detection image of the supporting structure, and a distance comparison strategy is performed on the supporting structure;

[0120] According to the distance comparison strategy, determine whether to mark the supporting structure;

[0121] If the supporting structure is marked, a new supporting structure is selected for detection; the reselection of a supporting structure for detection means selecting the next supporting structure according to the serial number sequence of the supporting structures, and performing distance detection strategy, contact detection strategy, distance detection strategy and attachment detection strategy on the supporting structure in sequence;

[0122] If the supporting structure is not marked, an attachment detection strategy is performed on the supporting structure, and according to the attachment detection strategy, it is determined whether to mark the supporting structure;

[0123] The marked support structures are repaired and tested again until all the support structures in the foundation pit are recorded as effective supports; the support structure is composed of multiple parts such as columns and cross braces, and this method uses a variety of detection strategies to detect each component to ensure that each part meets safety standards; the combined effect of multiple strategies such as distance detection, contact detection, and distance comparison can ensure the stability of the support structure from multiple angles; whether the columns are in contact with the foundation pit wall and whether the cross braces are tightly connected to the columns are important factors affecting the stability of the support structure. Through multi-level and multi-angle detection methods, it is comprehensively guaranteed that every detail of the support structure is in a safe state.

[0124] The obtaining of each supporting structure in the foundation pit and sequentially numbering the supporting structures according to the arrangement order of the supporting structures in the foundation pit includes:

[0125] Select one end of the foundation pit as the starting end, and the other end as the ending end;

[0126] The supporting structures in the foundation pit are numbered in sequence from the starting end to the ending end;

[0127] Each supporting structure in the foundation pit is inspected in sequence according to the numbering sequence of the supporting structures. By obtaining each supporting structure in the foundation pit and numbering them according to the arrangement sequence of the supporting structures, the entire inspection process is organized and traceable. This sequential numbering can help supervisors identify the location of each supporting structure and inspect them in sequence without missing any supporting structure. For large-scale foundation pit construction projects, there are a large number of supporting structures, and sequential numbering can effectively prevent confusion caused by complex structural layouts, ensuring that every supporting structure can be inspected and recorded. In addition, sequential numbering also helps with data management and subsequent maintenance work. Problem areas can be quickly located through numbering, improving maintenance efficiency.

[0128] The step of marking image acquisition points on the supporting structure according to the supporting structure includes:

[0129] For each supporting structure:

[0130] Obtain the first column and the second column and the vertical axis respectively, and record them as the first axis and the second axis;

[0131] Intercepting a portion of the first axis within the first column to form a first line segment;

[0132] Cutting off the portion of the second axis within the second column to form a second line segment;

[0133] Obtain the midpoints of the first line segment and the second line segment respectively, and connect them to form a reference line;

[0134] Get the midpoint of the reference line and record it as the reference midpoint;

[0135] The location of the reference midpoint is used as the image acquisition point of the support structure.

[0136] The step of obtaining the support image acquisition position according to the position of the support structure in the foundation pit and the image acquisition point on the support structure includes:

[0137] For each supporting structure in the foundation pit, the bottom surface of the foundation pit between each two adjacent supporting structures is intercepted; the bottom surface of the foundation pit is regarded as a horizontal plane;

[0138] Reference Figure 3, obtain the center point of the intercepted foundation pit bottom surface as a marking point;

[0139] For each supporting structure, the marking points located on both sides of the supporting structure and closest to the supporting structure are used as image marking points of the supporting structure;

[0140] Obtain a plane that passes through the image acquisition point of the support structure and is parallel to the bottom surface of the foundation pit, and record the plane as a reference plane;

[0141] Draw perpendicular lines to the bottom of the foundation pit through the image marking points respectively, and intersect with the reference plane respectively, and use the two intersection points as the support image acquisition positions of the support structure; by marking the image acquisition points for each support structure, it can be ensured that the acquisition of the detection image has a unified standard and reduce the deviation caused by human operation; the marking of the image acquisition points not only enables the detection image to clearly capture the key parts of the support structure, but also improves the accuracy of subsequent image analysis. The marked image acquisition points are usually based on the midpoints of the geometric centers of the columns and cross braces of the support structure, which can better reflect the overall state of the structure, effectively avoid the problem of local information loss when acquiring images, and improve the reliability of the detection results.

[0142] The step of collecting an image of the support structure at the support image collection position, which is recorded as a detection image of the support structure, includes:

[0143] For each supporting structure:

[0144] At the support image acquisition positions on both sides of the support structure, respectively, an image of the support structure is acquired in a direction pointing to the image acquisition point of the support structure;

[0145] The two images collected are used as the detection images of the support structure; by collecting images on both sides of the support structure and analyzing them, the status information of the support structure can be fully obtained from different angles to ensure the accuracy and completeness of the detection; there may be different stress concentrations or force conditions on both sides of the support structure, and the image on a single side may not be able to fully reflect the situation of the overall structure, but by collecting images on both sides, more details can be captured, especially when detecting the contact and distance between the cross brace and the column, and the support structure and the foundation pit wall, the dual-sided perspective can effectively avoid blind spots and missed detections; in addition, the images on both sides verify each other, which improves the reliability of the detection and reduces misjudgment caused by factors such as angle or light; at the same time, this dual-sided acquisition method can more comprehensively detect the offset, deformation or looseness of the support structure, provide a full range of detection perspectives, ensure the stability of the support structure in the foundation pit, and improve the accuracy and efficiency of engineering safety supervision.

[0146] The method of executing the distance detection strategy according to the detection image of the support structure and determining whether to mark the support structure includes:

[0147] A distance detection strategy is applied to each detection image of the support structure;

[0148] Reference Figure 1 , for the two sides of the foundation pit, with the direction from the first column to the second column, they are marked as the left foundation pit wall and the right foundation pit wall respectively;

[0149] According to the detection image of the support structure, the edge contour line of the first column close to the left foundation pit wall is taken as the first column line;

[0150] Reference Figure 2 , select any point on the first column line, and use the selected point as the starting point to draw a line segment parallel to the bottom surface of the foundation pit, connected to the left foundation pit wall and perpendicular to the left foundation pit wall, which is recorded as the first connecting vertical line;

[0151] Obtain the length of the first connecting perpendicular line, recorded as the first connecting line distance;

[0152] If the distance between the first connection line is not zero, it means that there is a gap between the first pillar of the supporting structure and the left foundation pit wall, and the supporting effect of the pillar on the foundation pit wall is poor, then the supporting structure is marked;

[0153] If all the distances of the first connecting lines are zero, it means that the first pillar of the supporting structure is closely attached to the left foundation pit wall, and the supporting effect of the pillar on the foundation pit wall is good. Then, according to the detection image of the supporting structure, the edge contour line of the second pillar close to the right foundation pit wall is taken as the second pillar line;

[0154] Take any point on the second column line, and use the point as the starting point to draw a line segment parallel to the bottom of the foundation pit and perpendicular to the right side of the foundation pit wall, which is recorded as the second connecting vertical line;

[0155] Obtain the length of the second connecting perpendicular line, recorded as the second connecting line distance;

[0156] If the distance between the second connection line is not zero, it means that there is a gap between the second pillar of the supporting structure and the right foundation pit wall, and the supporting effect of the pillar on the foundation pit wall is poor, then the supporting structure will be marked;

[0157] If all the first connection line distances are zero, it means that the second pillar of the support structure is closely fitted with the right foundation pit wall, and the support effect of the pillar on the foundation pit wall is good, then the support structure is recorded as the first effective support;

[0158] If the distance detection strategy based on the two detection images of the support structure records the support structure as the first effective support, the contact detection strategy will be implemented for the support structure; the stability of the support structure is directly related to the safety of the foundation pit project. By implementing the distance detection strategy on the detection image, the distance change between the support structure and the foundation pit wall can be discovered in time, so as to know whether there is a risk of displacement or tilt of the support structure. The distance detection strategy can detect subtle changes between the columns and the foundation pit wall, ensuring that the columns of the support structure are in full contact with the foundation pit wall, and that the pillars provide effective support to the foundation pit wall, thereby ensuring the safety of the foundation pit project.

[0159] If the supporting structure is not marked, a contact detection strategy is executed according to the detection image of the supporting structure, including:

[0160] A contact detection strategy is applied to each detection image of the support structure;

[0161] According to the detection image of the supporting structure, identifying and detecting whether the two ends of the first cross brace are in contact with the first column and the second column respectively;

[0162] If there is one end of the first cross brace that is not in contact with the first column or the second column, it means that the first cross brace does not support the column, and the supporting structure is marked;

[0163] If the two ends of the first cross brace are in contact with the first column and the second column respectively, it means that the first cross brace has a supporting effect on the pillar, then according to the detection image of the supporting structure, it is identified and detected whether the two ends of the second cross brace are in contact with the first column and the second column respectively;

[0164] If there is one end of the second cross brace that is not in contact with the first column or the second column, it means that the second cross brace does not support the column, and the supporting structure is marked;

[0165] If the two ends of the second cross brace are in contact with the first column and the second column respectively, it means that the second cross brace has a supporting effect on the column, and the supporting structure is recorded as the second effective support;

[0166] If the contact detection strategy based on the two detection images of the support structure records the support structure as the second effective support, the distance comparison strategy is executed for the support structure; the contact detection strategy ensures that the various components of the support structure are closely connected to each other by analyzing the contact between the cross brace and the column; this strategy effectively avoids the hidden dangers of failure caused by improper installation or looseness of the components in the support structure, ensures that the cross brace is in contact with the column, supports the column, ensures effective support for the foundation pit wall, and ensures the safety of the foundation pit project; if looseness or non-contact between the cross brace and the column is detected, the system will mark the structure and perform repair operations in the subsequent process; through this strategy, improper installation problems can be discovered and corrected in a timely manner, ensuring the overall stability of the support structure and reducing the probability of accidents.

[0167] If the supporting structure is not marked, the contact point of the supporting structure is obtained according to the detection image of the supporting structure, and a distance comparison strategy is performed on the supporting structure, including:

[0168] A distance comparison strategy is performed on each detection image of the support structure;

[0169] Reference Figure 2 , in the detection image, the line segment where the horizontal brace and the vertical column keep in contact is recorded as the contact line between the horizontal brace and the vertical column;

[0170] According to the detection image, contact lines between the first cross brace and the first column and the second column are respectively obtained and recorded as a first contact line and a second contact line;

[0171] According to the detection image, the contact lines between the second cross brace and the first column and the second column are respectively obtained and recorded as the third contact line and the fourth contact line respectively;

[0172] Obtaining midpoints of the first contact line, the second contact line, the third contact line, and the fourth contact line, respectively, and recording them as the first contact point, the second contact point, the third contact point, and the fourth contact point, respectively;

[0173] Obtain vertical distances between the first contact point, the second contact point, the third contact point and the fourth contact point and the bottom surface of the foundation pit respectively, and record them as the first distance, the second distance, the third distance and the fourth distance respectively;

[0174] comparing the first distance and the second distance;

[0175] If the first distance is not equal to the second distance, it means that although there is contact between the first cross brace and the column, the pressure on the first cross brace from the first column and the second column is not maintained on the same horizontal line, the supporting effect is average and there is a risk, then the supporting structure is marked;

[0176] If the first distance is equal to the second distance, then compare the third distance and the fourth distance;

[0177] If the third distance is not equal to the fourth distance, it means that although the first cross brace is in contact with the column, the pressure from the first column and the second column on the first cross brace is not kept on the same horizontal line, and the supporting effect is average and there is a risk, then the supporting structure will be marked;

[0178] If the third distance is equal to the fourth distance, the support structure is recorded as the third effective support;

[0179] If the support structure is recorded as the third effective support according to the distance comparison strategy of the two detection images of the support structure, the attachment detection strategy will be executed on the support structure; the distance comparison strategy ensures the stability of the support structure by comparing the vertical distance between the contact point of the support structure and the bottom surface of the foundation pit; by comparing the distance, it can be ensured that the cross brace and the column of the support structure are basically in a vertical state, so that the force generated by the column on the cross brace is on a horizontal line, so that the cross brace can play the best supporting role on the column, and the support structure can play the maximum supporting effect on the foundation pit wall, thereby ensuring the continuous stability of the support system.

[0180] If the supporting structure is not marked, performing an attachment detection strategy on the supporting structure, and judging whether to mark the supporting structure according to the attachment detection strategy, including:

[0181] The attachment detection strategy is applied to each detection image of the support structure;

[0182] According to the detection image, identify whether there are objects other than the supporting structure components attached to the supporting structure;

[0183] If present, the supporting structure is marked;

[0184] If it does not exist, the support structure is recorded as the fourth effective support;

[0185] If the attachment detection strategy based on the two detection images of the support structure records the support structure as the fourth effective support, the support structure will be recorded as a valid support structure, and the next support structure will be selected for detection; the attachment detection strategy mainly detects whether there are redundant attachments or other objects that affect the safety of the structure on the support structure. This strategy effectively ensures the safety and stability of the support structure and avoids heavy objects being carried or hung on the support structure, which may interfere with or damage the support structure; if there are attachments on the support structure, the attachments may affect the bearing capacity of the support structure or cause stress concentration, thereby affecting the overall stability. Through attachment detection, these potential hidden dangers can be effectively eliminated to ensure that the support structure plays its supporting role under ideal conditions.

[0186] Embodiment 2, a system for implementing the engineering safety supervision method based on artificial intelligence, comprising:

[0187] Image acquisition module: used to acquire detection images of support structures;

[0188] Point marking module: used to calibrate image acquisition points and contact points;

[0189] Distance acquisition module: used to obtain the distance between the contact point and the bottom surface of the foundation pit, and the distance between the column and the foundation pit wall;

[0190] Numbering module: Sequentially number the supporting structures in the foundation pit to determine the inspection order;

[0191] Image recognition module: determines whether the horizontal brace and the column of the supporting structure are in contact based on the detection image;

[0192] Distance comparison module: used to compare the first distance, the second distance, the third distance and the fourth distance to determine whether the distances are equal;

[0193] Intelligent recognition module: Through artificial intelligence technology, the detection image is recognized to identify whether there are objects other than support structure components attached to the support structure.

[0194] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0195] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An engineering safety supervision method based on artificial intelligence, characterized in that: include: Obtain each supporting structure in the foundation pit, and number the supporting structures in sequence according to the arrangement order of the supporting structures in the foundation pit; Each support structure in the foundation pit consists of two columns and two cross braces; For each supporting structure, its two columns are respectively recorded as the first column and the second column, and its two cross braces are respectively recorded as the first cross brace and the second cross brace; Mark the image acquisition points according to the support structure; According to the position of the supporting structure in the foundation pit and the image acquisition points on the supporting structure, the supporting image acquisition position is obtained; For each supporting structure, an image of the supporting structure is collected at a supporting image collection position, and recorded as a detection image of the supporting structure; According to the detection image of the support structure, the distance detection strategy is executed, and it is determined whether to mark the support structure; If the supporting structure is marked, select a new supporting structure for testing; If the supporting structure is not marked, the contact detection strategy is executed according to the detection image of the supporting structure; According to the contact detection strategy, determine whether to mark the supporting structure; If the supporting structure is marked, select a new supporting structure for testing; If the supporting structure is not marked, the contact point of the supporting structure is obtained according to the detection image of the supporting structure, and a distance comparison strategy is performed on the supporting structure; According to the distance comparison strategy, determine whether to mark the supporting structure; If the supporting structure is marked, select a new supporting structure for testing; If the supporting structure is not marked, an attachment detection strategy is performed on the supporting structure, and according to the attachment detection strategy, it is determined whether to mark the supporting structure; The marked support structures shall be repaired and tested again until all the support structures in the foundation pit are recorded as effective supports.

2. According to claim 1, an engineering safety supervision method based on artificial intelligence is characterized in that: The obtaining of each supporting structure in the foundation pit and sequentially numbering the supporting structures according to the arrangement order of the supporting structures in the foundation pit includes: Select one end of the foundation pit as the starting end, and the other end as the ending end; The supporting structures in the foundation pit are numbered in sequence from the starting end to the ending end; Each supporting structure in the foundation pit shall be inspected in turn according to the numbering order of the supporting structures.

3. According to the artificial intelligence-based engineering safety supervision method of claim 1, it is characterized in that: The step of marking image acquisition points on the supporting structure according to the supporting structure includes: For each supporting structure: Obtain the first column and the second column and the vertical axis respectively, and record them as the first axis and the second axis; Intercepting a portion of the first axis within the first column to form a first line segment; Cutting off the portion of the second axis within the second column to form a second line segment; Obtain the midpoints of the first line segment and the second line segment respectively, and connect them to form a reference line; Get the midpoint of the reference line and record it as the reference midpoint; The location of the reference midpoint is used as the image acquisition point of the support structure.

4. The engineering safety supervision method based on artificial intelligence according to claim 3 is characterized in that: The step of obtaining the support image acquisition position according to the position of the support structure in the foundation pit and the image acquisition point on the support structure includes: For each supporting structure in the foundation pit, the bottom surface of the foundation pit between every two adjacent supporting structures is intercepted; Obtain the center point of the intercepted foundation pit bottom surface as a marking point; For each supporting structure, the marking points located on both sides of the supporting structure and closest to the supporting structure are used as image marking points of the supporting structure; Obtain a plane that passes through the image acquisition point of the support structure and is parallel to the bottom surface of the foundation pit, and record the plane as a reference plane; Draw perpendicular lines to the bottom of the foundation pit through the image marking points respectively, and intersect with the reference plane respectively. The two intersection points obtained are used as the support image acquisition positions of the support structure.

5. The engineering safety supervision method based on artificial intelligence according to claim 1 is characterized in that: The step of collecting an image of the support structure at the support image collection position, which is recorded as a detection image of the support structure, includes: For each supporting structure: At the support image acquisition positions on both sides of the support structure, respectively, an image of the support structure is acquired in a direction pointing to the image acquisition point of the support structure; The two acquired images are used as detection images of the support structure.

6. The engineering safety supervision method based on artificial intelligence according to claim 1 is characterized in that: The method of executing the distance detection strategy according to the detection image of the support structure and determining whether to mark the support structure includes: A distance detection strategy is applied to each detection image of the support structure; The two sides of the foundation pit wall are marked as the left foundation pit wall and the right foundation pit wall, with the direction from the first column to the second column; According to the detection image of the support structure, the edge contour line of the first column close to the left foundation pit wall is taken as the first column line; Take any point on the first column line, and use the point as the starting point to draw a line segment parallel to the bottom surface of the foundation pit, connected to the left foundation pit wall and perpendicular to the left foundation pit wall, which is recorded as the first connecting vertical line; Obtain the length of the first connecting perpendicular line, recorded as the first connecting line distance; If there is a first connection line distance that is not zero, the support structure is marked; If all the first connection line distances that exist are zero, then based on the detection image of the support structure, the edge contour line of the second column close to the right foundation pit wall is taken as the second column line; Take any point on the second column line, and use the point as the starting point to draw a line segment parallel to the bottom of the foundation pit and perpendicular to the right side of the foundation pit wall, which is recorded as the second connecting vertical line; Obtain the length of the second connecting perpendicular line, recorded as the second connecting line distance; If there is a second connection line with a distance not equal to zero, the support structure is marked; If all the first connection distances that exist are zero, the support structure is recorded as the first effective support; If the distance detection strategy performed according to the two detection images of the support structure both records the support structure as the first effective support, the contact detection strategy is executed on the support structure.

7. The engineering safety supervision method based on artificial intelligence according to claim 1 is characterized in that: If the supporting structure is not marked, a contact detection strategy is executed according to the detection image of the supporting structure, including: A contact detection strategy is applied to each detection image of the support structure; According to the detection image of the supporting structure, identifying and detecting whether the two ends of the first cross brace are in contact with the first column and the second column respectively; If there is an end of the first cross brace that is not in contact with the first column or the second column, the supporting structure is marked; If the two ends of the first cross brace are in contact with the first column and the second column respectively, then according to the detection image of the support structure, identifying and detecting whether the two ends of the second cross brace are in contact with the first column and the second column respectively; If there is an end of the second cross brace that is not in contact with the first column or the second column, the supporting structure is marked; If the two ends of the second cross brace are in contact with the first column and the second column respectively, the support structure is recorded as the second effective support; If the contact detection strategies performed according to the two detection images of the support structure both record the support structure as the second effective support, the distance comparison strategy is executed on the support structure.

8. The engineering safety supervision method based on artificial intelligence according to claim 1 is characterized in that: If the supporting structure is not marked, the contact point of the supporting structure is obtained according to the detection image of the supporting structure, and a distance comparison strategy is performed on the supporting structure, including: A distance comparison strategy is performed on each detection image of the support structure; In the detection image, the line segment where the horizontal brace and the vertical column are in contact is recorded as the contact line between the horizontal brace and the vertical column; According to the detection image, contact lines between the first cross brace and the first column and the second column are respectively obtained and recorded as a first contact line and a second contact line; According to the detection image, the contact lines between the second cross brace and the first column and the second column are respectively obtained and recorded as the third contact line and the fourth contact line respectively; Obtaining midpoints of the first contact line, the second contact line, the third contact line, and the fourth contact line, respectively, and recording them as the first contact point, the second contact point, the third contact point, and the fourth contact point, respectively; Obtain vertical distances between the first contact point, the second contact point, the third contact point and the fourth contact point and the bottom surface of the foundation pit respectively, and record them as the first distance, the second distance, the third distance and the fourth distance respectively; comparing the first distance and the second distance; If the first distance is not equal to the second distance, the support structure is marked; If the first distance is equal to the second distance, then compare the third distance and the fourth distance; If the third distance is not equal to the fourth distance, the support structure is marked; If the third distance is equal to the fourth distance, the support structure is recorded as the third effective support; If the support structure is recorded as the third effective support according to the contact distance comparison strategy performed on the two detection images of the support structure, the attachment detection strategy is executed on the support structure.

9. The engineering safety supervision method based on artificial intelligence according to claim 1 is characterized in that: If the supporting structure is not marked, performing an attachment detection strategy on the supporting structure, and judging whether to mark the supporting structure according to the attachment detection strategy, including: The attachment detection strategy is applied to each detection image of the support structure; According to the detection image, identify whether there are objects other than the supporting structure components attached to the supporting structure; If present, the supporting structure is marked; If it does not exist, the support structure is recorded as the fourth effective support; If the attachment detection strategies performed according to the two detection images of the support structure both record the support structure as the fourth effective support, the support structure is recorded as the effective support structure, and the next support structure is selected for detection.

10. A system using the engineering safety supervision method based on artificial intelligence as claimed in claim 1, characterized in that: include: Image acquisition module: used to acquire detection images of support structures; Point marking module: used to calibrate image acquisition points and contact points; Distance acquisition module: used to obtain the distance between the contact point and the bottom surface of the foundation pit, and the distance between the column and the foundation pit wall; Numbering module: Sequentially number the supporting structures in the foundation pit to determine the inspection order; Image recognition module: determines whether the horizontal brace and the column of the supporting structure are in contact based on the detection image; Distance comparison module: used to compare the first distance, the second distance, the third distance and the fourth distance to determine whether the distances are equal; Intelligent recognition module: Through artificial intelligence technology, the detection image is recognized to identify whether there are objects other than support structure components attached to the support structure.

Citation Information

Patent Citations

  • Intelligent foundation pit supporting construction method based on BIM technology

    CN115238343A

  • Foundation pit displacement detection device and monitoring method

    CN117803025A

  • Building deep foundation pit remote monitoring method and processing system

    CN118581930A

  • System and method of monitoring vibration of a blasting model test for a jointed rock mass

    US20220244155A1