Integrated intelligent monitoring system and method for intelligent construction site

Abstract

The invention discloses an integrated intelligent monitoring system and method for an intelligent construction site, and relates to the technical field of construction site monitoring, and the method comprises the steps: arranging a monitoring camera based on a construction scene of construction site construction, carrying out the image collection of the construction site based on the arranged monitoring camera, and obtaining the collected image data; performing image recognition on the acquired image data based on an image recognition technology to obtain article data, performing structural risk judgment on the article data, and marking an article with risk to obtain first article data; determining position information of the article data based on the acquired image data, performing state risk assessment on the article according to the position information, and marking the article with risk as second article data; according to the first article data and the second article data, determining the dangerousness of the article, and marking the article with the dangerousness as a dangerous object; and when the dangerous object is judged to exist, performing alarm marking on the dangerous object. The construction site safety monitoring method has the effect of improving the construction site safety monitoring accuracy.

Description

Technical Field

[0001] This application relates to the technical field of construction site monitoring, and in particular, to an integrated intelligent monitoring system and method for a smart construction site. Background Art

[0002] With the development of social economy and the implementation of a large number of infrastructure projects, how to ensure the safety of construction sites has become an urgent problem to be solved.

[0003] In the prior art, by setting up monitoring devices, such as cameras, at construction sites to dynamically monitor the personnel at the construction sites, when an accident is judged to occur, an alarm signal can be sent out in time to minimize the harm caused by the accident, thereby ensuring the safety of workers.

[0004] However, in the above-mentioned technology, it can only be dealt with when the risk occurs, and it is impossible to give an early warning of the risk to further reduce the harm caused by the risk, so there is room for improvement. Summary of the Invention

[0005] In order to improve the accuracy of construction site safety monitoring, this application provides an integrated intelligent monitoring system and method for a smart construction site.

[0006] In the first aspect, this application provides an integrated intelligent monitoring method for a smart construction site, adopting the following technical solutions: An integrated intelligent monitoring method for a smart construction site includes: Arranging monitoring cameras based on the construction scenarios of the construction site, and collecting images of the construction site based on the arranged monitoring cameras to obtain collected image data; Based on image recognition technology, performing image recognition on the collected image data to obtain item data, and performing a structural risk judgment on the item data, and marking the items with risks to obtain first item data; Determining the position information of the item data based on the collected image data, and performing a status risk assessment on the item according to the position information, and marking the items with risks as second item data; According to the first item data and the second item data, determining the danger of the item, and marking the items with danger as dangerous items; When it is determined that there are dangerous items, perform an alarm mark on the dangerous items.

[0007] Preferably, determine the construction scenario according to the construction characteristics of the construction site; the construction scenario includes planar construction and three-dimensional construction; If the construction scenario is three-dimensional construction, then divide each construction layer into planar construction with corresponding spatial relationships to obtain construction areas; Obtain the monitoring range of the surveillance cameras, and use the monitoring range as the spacing between the surveillance cameras. Arrange the surveillance cameras in the construction area according to the spacing between the surveillance cameras. Collect image data of the monitoring range based on the surveillance cameras to obtain the collected image data.

[0008] Preferably, collect arbitrary images of the construction area based on the surveillance cameras to obtain the image data to be analyzed. Based on image recognition technology, identify items in the image data to be analyzed to obtain the item data to be judged. Compare the item data to be judged with the actual item data to determine the recognition accuracy, and compare the recognition accuracy with the preset accuracy value to determine whether the recognition accuracy meets the standard. If the recognition accuracy does not meet the standard, adjust the monitoring range of the surveillance cameras until the recognition accuracy of the images collected by the image recognition technology is greater than the preset accuracy value.

[0009] Preferably, perform image recognition on the collected image data based on image recognition technology to determine all items included in the monitoring area and obtain the item data. Based on the collected image data, determine the position and shape and size of each item to obtain the first position information and shape information. Based on the first position information, screen the items to obtain the items whose positions are within the built-in space range, and mark the corresponding items as the first item items. Determine the effective area of the corresponding item based on the first item item. Obtain the physical properties of each item in the first item item, substitute the physical properties and the effective area into the built-in first evaluation formula to obtain the structural risk value of the item, and match the structural risk value with the built-in first evaluation index. Mark the items with a structural risk value greater than the built-in first evaluation index as the first item data.

[0010] Preferably, construct a three-dimensional space based on the monitoring area of the surveillance cameras and the image perspective relationship. Match the three-dimensional space with the area monitored by the surveillance cameras to determine the items that exceed the surface of the three-dimensional space in the monitoring area and mark them as the items to be recognized. Perform recognition on the items to be recognized based on image recognition technology to obtain the item data.

[0011] Preferably, mark the positions of the items based on the item data to obtain the position data. Compare the position data of the same item in multiple collected image data of the same surveillance camera with each other to determine whether the position of the item has changed. If it is determined that the position of the item has changed, then judge the reason for the position change of the item to determine whether there is human intervention; If it is determined to be human intervention, then re-mark and judge the position of the item; If it is determined to be non-human intervention, then determine the relationship between the displacement distance and time of the item based on the collected image data of the item, obtain the state change index, and determine the risk of the item according to the state change index.

[0012] Preferably, obtain the time period corresponding to the change in the position of the item, and perform face recognition on the collected image data of the corresponding time period according to the corresponding time period to determine whether there is a face in the corresponding collected image data; If it is determined that there is a face in the collected image data, then determine the position movement data of the face according to the collected image data, and match the position movement data of the face with the position change data when the position of the item changes to determine whether the position movement data of the face coincides with the position change data when the position of the item changes; If it is determined that the position movement data of the face coincides with the position data before the position of the item changes, then determine that the reason for the position change of the item is human intervention; otherwise, it is determined to be non-human intervention.

[0013] Preferably, based on the construction layer, establish a hierarchical relationship for the construction scene, and determine the hierarchical data of each monitoring camera according to the hierarchical relationship; Obtain the hierarchical data and regional location information of the current monitoring area, and according to the hierarchical data, regional location information and construction scene, obtain the collected image data located in the same regional location and with hierarchical data less than the current monitoring range, and perform statistical sorting on the collected image data and the collected image data corresponding to the current monitoring range according to the hierarchical relationship to obtain the group to be analyzed; If the hierarchical data of the current monitoring range is the first layer, then based on the built-in risk assessment model, perform risk assessment on the first item data and the second item data to determine whether the item is dangerous; If the hierarchical data of the current monitoring range is not the first layer, then judge the position state of the first item data, and divide the first item data into the first fixed items with unchanged item positions and the first moving items with changed item positions according to the position states of each item; Based on the built-in risk assessment model, perform item judgment on the first fixed items to determine whether there is a risk judgment criterion for the item in the risk assessment model. If there is a risk judgment criterion for the item, then perform risk assessment on the item based on the risk assessment model to determine whether the item is dangerous; If there is no risk judgment criterion for the item in the risk assessment model, an auxiliary signal is output, and the manual risk marking result after manual risk marking is received. The manual risk marking result is used as the risk assessment result of the item, and the risk assessment model is trained for the corresponding item based on the manual risk marking result to obtain the risk judgment criterion for the corresponding item; Based on the group to be analyzed, perform face recognition on the group to be analyzed to determine whether there is a human trajectory within the monitoring range at the lower level of the current monitoring range. If it is determined that there is a human trajectory, determine the risk of falling of the corresponding item according to the status change index of each item in the first moving item and the second item data to obtain the first index data; Perform risk assessment based on the built-in risk assessment model to obtain the second index data, and determine whether the item is dangerous based on the first index data and the second index data.

[0014] In a second aspect, the present application provides an integrated intelligent monitoring system for a smart construction site, adopting the following technical solutions: An integrated intelligent monitoring system for a smart construction site includes: a collection planning module, a first risk judgment module, a second risk judgment module, a danger judgment module, and an alarm module; The collection planning module arranges monitoring cameras based on the construction scenario of the construction site, and performs image collection on the construction site based on the arranged monitoring cameras to obtain collected image data; The first risk judgment module performs image recognition on the collected image data based on image recognition technology to obtain item data, and performs structural risk judgment on the item data, and marks the items with risks to obtain the first item data; The second risk judgment module determines the position information of the item data based on the collected image data, and performs status risk assessment on the item according to the position information, and marks the items with risks as the second item data; The danger judgment module determines the danger of the item according to the first item data and the second item data, and marks the items with danger as dangerous items; The alarm module makes an alarm mark for the dangerous item when it is determined that there is a dangerous item.

[0015] In summary, the present application includes at least one of the following beneficial technical effects: By arranging monitoring cameras in the construction scenes of the construction site, it is possible to comprehensively monitor the construction area, thereby improving the accuracy of safety judgment in the construction area. By performing object recognition on the collected images, a first risk judgment is made on the objects in the monitored area, and the dangerous objects are marked and monitored. Then, by judging and evaluating the position information of the objects, the stability of the objects is determined to further determine the danger of the objects in the monitored area, making the results obtained from the two judgments more comprehensive. As a result, when making subsequent risk judgments on the objects, the obtained results are more accurate, thereby improving the accuracy of safety monitoring of the construction site; Use the image recognition technology in the existing technology to perform image recognition on the images collected by the monitoring cameras, and compare the recognized results with the actual item data, so as to determine the compatibility between the adopted image recognition technology and the collection range of the current monitoring camera, thereby ensuring that the objects within the adjusted monitoring range can be accurately and effectively recognized, and further improving the accuracy of safety monitoring of this area; By marking the position of the same object for a long time, it is possible to determine whether the position of the object is likely to change, and determine the state stability of the object according to the difficulty of position change of the object. According to the relationship between the position change displacement of the object and time, the state stability value of the object is determined; when it is determined that the object has changed its position, by performing artificial intervention analysis on the reason for the position change of the object, performing face recognition on the time period when the position changes, and making a coincidence judgment on the action track of the face recognition and the position track when the position of the object changes, so as to improve the accuracy of judging the state stability of the object, and further improve the accuracy of the safety assessment of the construction site. Description of the Drawings

[0016] Figure 1 It is the step flow chart of the integrated intelligent monitoring method for the smart construction site in this embodiment; Figure 2 It is the module block diagram of the integrated intelligent monitoring system for the smart construction site in this embodiment.

[0017] Reference Signs: 1. Acquisition Planning Module; 2. First Risk Judgment Module; 3. Second Risk Judgment Module; 4. Danger Judgment Module; 5. Alarm Module. Detailed Embodiment

[0018] The following will further elaborate on this application in conjunction with the attached Figure 1 - Figure 2 Make a more detailed description of this application.

[0019] This application embodiment discloses an integrated intelligent monitoring system and method for a smart construction site.

[0020] Embodiment: As Figure 1As shown, an integrated intelligent monitoring method for a smart construction site according to the present invention includes: S1, arranging monitoring cameras based on the construction scenarios of the construction site, and collecting images of the construction site through the arranged monitoring cameras to obtain collected image data; S2, based on image recognition technology, performing image recognition on the collected image data to obtain item data, and making a structural risk judgment on the item data, marking the items with risks to obtain the first item data; by identifying the items and judging the hazards of the identified items, the hazards that the items themselves can bring are determined.

[0021] S3, determining the location information of the item data based on the collected image data, and performing a status risk assessment on the items according to the location information, marking the items with risks as the second item data; the risk assessment is used to determine the status change of the items, and the more easily the status changes, the higher the risk.

[0022] S4, determining the danger of the items according to the first item data and the second item data, and marking the items with danger as dangerous items; S5, when it is determined that there are dangerous items, making an alarm mark on the dangerous items.

[0023] In this embodiment, by arranging monitoring cameras for the construction scenarios of the construction site, it is possible to comprehensively monitor the construction area, thereby improving the accuracy of safety judgment for the construction area. By performing item recognition on the collected images, the first danger judgment is made on the items in the monitoring area, and the items with danger are marked and monitored; then, by judging and evaluating the location information of the items, the stability of the items is determined to further determine the danger of the items in the monitoring area, making the results obtained from the two judgments more comprehensive, so that when making subsequent danger judgments on the items, the obtained results are more accurate, and thus the safety monitoring of the construction site is more accurate.

[0024] Exemplarily, by comprehensively monitoring the construction area of the construction site through surveillance cameras, the result of the safety assessment of the construction area can be made more comprehensive and accurate when conducting a safety assessment of the construction area. When conducting a safety assessment of the construction site for the area monitored by the surveillance cameras, existing image recognition technology is used to identify items in the monitoring area of the surveillance cameras, and a structural risk judgment is made on the identified items to make a first judgment on the items. For example, if bricks, steel bars, gravel, ropes, etc. are identified in the area, a structural risk judgment is first made on the items to determine which items in the area can directly cause harm. For example, bricks and steel bars can directly cause harm, so it is determined that bricks and steel bars are risky, and thus bricks and steel bars are marked as the first item data. Since the more stable an item is, the lower the likelihood of danger, and conversely, the more unstable an item is, the more risk factors it brings. Therefore, after determining which items can directly cause harm, it is necessary to evaluate the state of the position information of the items to determine the stability of the items. For example, since a rope is not stable, the rope is marked as the second item data. By making a first judgment on the harmfulness of an item itself and then making a second judgment on the state of the item, the accuracy of the risk judgment of the item can be comprehensively improved. However, just because an item itself is dangerous and its state is prone to change does not mean that the item is dangerous. Therefore, further judgment is needed. By combining the dangerous action method when the corresponding item is dangerous, it is determined whether there is a possibility of danger triggering in the current environment. Similarly, the reason why danger is called danger is that it has an object that cannot be acted upon. For example, for a debris flow, if there are villages, households, etc. below the debris flow, it is determined that the occurrence of the debris flow is dangerous. Conversely, if there are no valuable products below the debris flow, it is determined that the debris flow is not dangerous. Therefore, based on the object of action, the item is rejudged to determine whether the occurrence of the danger will affect people or construction, so that the judgment of the danger is more accurate, and thus the accuracy of the construction site safety monitoring is improved.

[0025] In step S1, the surveillance cameras are arranged based on the construction scenario of the construction site, and image acquisition of the construction site is performed based on the arranged surveillance cameras to obtain the acquired image data, including the following steps: S11, determine the construction scenario according to the construction characteristics of the construction site; the construction scenario includes flat construction and three-dimensional construction; S12, if the construction scenario is three-dimensional construction, divide each construction layer into flat construction with corresponding spatial relationships to obtain the construction area; S13, obtain the monitoring range of the surveillance cameras, and use the monitoring range as the spacing between the surveillance cameras. Arrange the surveillance cameras for the construction area according to the spacing between the surveillance cameras. S131. Arbitrarily collect images of the construction area based on a surveillance camera to obtain image data to be analyzed; S132. Based on image recognition technology, perform item recognition on the image data to be analyzed to obtain item data to be judged; S133. Compare the item data to be judged with the actual item data to determine the recognition accuracy, and compare the recognition accuracy with a preset accuracy value to determine whether the recognition accuracy meets the standard; S134. If the recognition accuracy does not meet the standard, adjust the monitoring range of the surveillance camera until the recognition accuracy of the images collected by the image recognition technology is greater than the preset accuracy value.

[0026] S14. Collect images of the monitoring range based on a surveillance camera to obtain collected image data.

[0027] In this embodiment, due to the particularity of the construction area, multiple surveillance cameras are required to monitor the construction area. The installation positions of the surveillance cameras are determined by using the monitoring ranges of the surveillance cameras. At the same time, since the monitoring ranges of the surveillance cameras are infinite in depth, it is necessary to determine the monitoring ranges of the surveillance cameras. Image recognition is performed on the images collected by the surveillance cameras by using image recognition technology in the existing technology, and the recognized results are compared with the actual item data, so as to determine the adaptability between the adopted image recognition technology and the collection range of the current surveillance camera, and further ensure that the items within the adjusted monitoring range can be accurately and effectively recognized, thereby improving the accuracy of the safety monitoring of this area.

[0028] Exemplarily, since the construction scenario of the construction site not only has planar construction but also may have three-dimensional construction, in the case of three-dimensional construction, the sources of danger not only include the ground, etc., but also include the risk of falling objects from above. Therefore, it is necessary to arrange the surveillance cameras correspondingly according to the construction characteristics of the construction site of the construction site to ensure the comprehensiveness of the monitoring of the construction site.

[0029] If the monitoring range of the surveillance camera is 10 meters, determine the distribution of the surveillance cameras according to the structural characteristics of the construction area, so that the distance between the surveillance cameras is less than or equal to 10 meters, and arrange the construction area so that the surveillance cameras can comprehensively monitor the construction area.

[0030] Since the monitoring camera is limited in width but infinite in depth, it is necessary to limit the depth of the monitoring camera. By using the camera to collect the construction area once, and using the existing image recognition technology to identify the items in the image, and then comparing the identified items with the actual items in the area, the accuracy of the image recognition technology in item recognition can be determined. This accuracy includes whether an item can be recognized and whether the category of the item can be determined. By adjusting the monitoring range, the existing image recognition technology can achieve a good adaptation effect with the collected image, thereby improving the recognition accuracy.

[0031] In step S2, based on the image recognition technology, the collected image data is subjected to image recognition to obtain item data, and the structural risk of the item data is judged, and the items with risks are marked to obtain the first item data, including the following steps: S21, based on the image recognition technology, perform image recognition on the collected image data to determine all items included in the monitoring area and obtain item item data; S22, based on the collected image data, determine the position judgment and shape and size judgment of each item to obtain the first position information and shape information; S23, based on the first position information, screen the items to obtain the items whose positions are within the built-in space range, and mark the corresponding items as the first item items; S24, based on the first item items, determine the acting area of the corresponding item items; where the acting area is, when the position of the item does not change, the contact area when the surface formed by the horizontal line, vertical line and 45-degree inclined line in the built-in space range contacts the item.

[0032] S25, obtain the material characteristics of each item in the first item items, substitute the material characteristics and the acting area into the built-in first evaluation formula to obtain the structural risk value of the item, and match the structural risk value with the built-in first evaluation index, and mark the items with the structural risk value greater than the built-in first evaluation index as the first item data. Among them, the material characteristics of the item can be the hardness, sharpness, structural properties, etc. of the item. The first evaluation formula can be: z = ax + by, where z is the evaluated structural risk value, a is the material characteristic coefficient, x is the value of each material characteristic of the item, b is the acting area coefficient, and y is the acting area of the item.

[0033] Among them, the item characteristics determine the hardness and overall jointness of the item. For example, for bricks and gravel, bricks and gravel of the same volume can be regarded as having the same hardness, but in terms of overall jointness, the jointness of bricks is greater than that of gravel, so the danger of bricks is greater than that of gravel.

[0034] In this embodiment, by performing image recognition on the collected image data, the categories of items in the monitoring area are identified. At the same time, by judging the positions of each item, the items are screened, reducing the number of items to be evaluated, thereby improving the evaluation efficiency. Then, using the relationship between pressure and acting area and the relationship between action and reaction forces, a built-in first evaluation formula is constructed, and the relevant data of the corresponding item is substituted into this formula, making the final evaluation result more accurate and effective, and thus improving the accuracy of the safety evaluation of the construction site.

[0035] Exemplarily, first, the image is recognized once through image recognition technology to determine which items are included in the image. Then, according to the image, the shape, size, and position information of the corresponding item are determined. Assuming that the space of the monitoring range is a 3*3*3 cubic meter space, and the average height of a person is 1.8 meters, and the safety height is set to 2 meters, then only the items in the space below 2 meters need to be evaluated for structural risk, thereby reducing the amount of data for part of the evaluation. When judging the acting area of an item, assuming that the item is a cube, and assuming that its position state does not change, when a person collides with the item, it is mostly an inclined collision. To ensure the comprehensiveness of the collision situation, the possibility of collision of this item is comprehensively analyzed. By establishing a detection surface in the monitored space, for example, if the cube is placed upright on the ground, the contact area between the horizontal plane and the cube is the upper surface of the cube, and the contact area between the vertical plane and the cube is the two side surfaces of the cube. The contact between the inclined plane at 45 degrees and the cube is the edge or vertex of the cube. According to the action relationship between pressure and the force-bearing area, it can be judged that the smaller the acting area, the greater the impact force and the higher the harmfulness.

[0036] At the same time, according to the relationship between action and reaction forces, when the same acting force acts on objects of different materials, the softer the material of the item, the more easily its external form changes when the acting force acts. For the same cube, one is made of metal and the other is made of foam. When the same kind of collision occurs, the harm of the metal is greater than that of the foam. Therefore, according to the above characteristics, a first evaluation formula is constructed, making the result obtained more accurate when evaluating the structural risk of an item.

[0037] In step S21, based on image recognition technology, the collected image data is recognized to determine all the items included in the monitoring area, and item data is obtained, including the following steps: S211, based on the monitoring area of the monitoring camera and the image perspective relationship, construct a three-dimensional space. S212. Match the three-dimensional space with the area monitored by the surveillance camera, determine the items in the monitored area that exceed the surface of the three-dimensional space, and mark them as items to be recognized. S213. Recognize the items to be recognized based on image recognition technology to obtain item data.

[0038] In this embodiment, by judging the spatial perspective relationship of the images collected by the surveillance camera, the spatial range represented by the image data is clarified. By establishing a corresponding three-dimensional space model and matching the model with the actually collected acquisition image data, the additives existing in the acquisition image data are determined, and the items are recognized for the additives, improving the accuracy of item recognition.

[0039] Exemplarily, since the images collected by the surveillance camera are planar images, when recognizing planar images, it is necessary to determine the specific recognition range and clarify which things are the additives in this space, so that the image recognition technology can accurately recognize the additives and improve the accuracy.

[0040] Since the space is three-dimensional, when establishing the corresponding spatial relationship according to the acquisition image data, the spatial boundary of the monitored area coincides with the spatial boundary of the constructed three-dimensional space. When there are other items in the monitored area, the item will protrude from the plane. Therefore, when matching with the constructed three-dimensional space, the things protruding into the three-dimensional space are marked as items to be recognized, thereby reducing the difficulty of item recognition and improving the accuracy of item recognition.

[0041] In step S3, based on the acquisition image data, determine the position information of the item data, and perform a status risk assessment on the item according to the position information, and mark the items with risks as the second item data, including the following steps: S31. Mark the position of the item based on the item data to obtain position data. S32. Compare the position data of the same item in multiple acquisition image data of the same surveillance camera with each other to determine whether the position of the item has changed. S33. If it is determined that the position of the item has changed, judge the reason for the position change of the item to determine whether there is human intervention. S331. Obtain the time period corresponding to when the position of the item changes, and perform human face recognition on the acquisition image data corresponding to the corresponding time period to determine whether there is a human face in the corresponding acquisition image data. S332. If it is determined that there is a human figure in the collected image data, determine the position movement data of the human figure based on the collected image data, and match the position movement data of the human figure with the position change data when the position of the item changes to determine whether the position movement data of the human figure coincides with the position change data when the position of the item changes. S333. If it is determined that the position movement data of the human figure coincides with the position data before the position of the item changes, it is determined that the reason for the position change of the item is human intervention; otherwise, it is determined that it is non-human intervention.

[0042] S34. If it is determined to be human intervention, re-mark and judge the position of the item. S35. If it is determined to be non-human intervention, determine the relationship between the displacement distance and time of the item based on the collected image data of the item, obtain a state change index, and determine the riskiness of the item according to the state change index.

[0043] In this embodiment, by performing long-term position marking on the same item, it is determined whether the position of the item is likely to change, and the state stability of the item is determined according to the difficulty of the position change of the item. According to the relationship between the position change displacement and time of the item, the state stability value of the item is determined; when it is determined that the position of the item has changed, by analyzing the reason for the position change of the item for human intervention, performing human figure recognition on the time period when the position changes, and judging the coincidence of the action track of the human figure recognition and the position track when the position of the item changes, the accuracy of judging the state stability of the item is improved, and thus the accuracy of the safety assessment of the construction site is improved.

[0044] Exemplarily, after performing structural risk judgment on the physical properties and action area of the item, then performing item state change judgment on the items within the monitoring range. Since the less stable the item is, the greater the position change, and the more corresponding dangerous variables in the environment, by judging the item state change, the accuracy of the judgment is improved.

[0045] For example, there is an item A within the monitoring range. When performing the first collection, item A is at position a, and when performing the second collection, item A is at position b. Then it is determined that the position of the item has changed. Since the state of the item has changed, it indicates that the item is not stable and is thus more likely to generate danger. Therefore, the item needs to be marked. At the same time, since the marking of the item can only determine that the position of the item has changed but cannot determine the reason for the change. For example, assuming that item A is a tool in use, the position change of item A may be due to the staff using item A, resulting in the position change. Therefore, it is necessary to judge the reason for the change of item A to determine whether the change of item A is caused by human intervention.

[0046] By performing human face recognition on the time period during which the position of item A changes, it is determined whether there are people passing through this area during this time period. When it is determined that there are people arriving in this area, it is necessary to make an associated judgment on the position changes of this person and item A, so as to determine that the position change of item A is directly related to this person. By making a coincidence judgment on the movement trajectory of the person and the position data before and after the position change of the item, when the position change of item A is caused by human intervention, the movement trajectory of this person should coincide with the position of the position change of item A, so as to judge the reason for the position change of item A.

[0047] When it is determined that the reason for the position change of item A is a change caused by human intervention, the stability of the state of item A cannot be determined, so it is necessary to re-evaluate the stability of the state of item A. When it is determined that the reason for the position change of item A is not a change caused by human intervention, the stability of the state of item A can be determined, and then the variability of item A can be determined by determining the relationship between the position change of the item and time.

[0048] The more position changes item A makes in a shorter period of time, the less stable the state of item A is, and the more dangerous item A is accordingly.

[0049] In step S4, according to the first item data and the second item data, the danger of the item is determined, and the item with danger is marked as a dangerous item, including the following steps: S41, Based on the construction layer, establish a hierarchical relationship for the construction scene, and determine the hierarchical data of each monitoring camera according to the hierarchical relationship; among them, the hierarchical relationship takes the construction layer as the evaluation criterion, marks the construction layer on the ground as the first layer, and makes corresponding marks on other construction layers according to the change in height. For example, the second layer, the third layer, etc.

[0050] S42, Obtain the hierarchical data and regional location information of the current monitoring area, and according to the hierarchical data, regional location information and construction scene, obtain the acquisition image data located in the same regional location and with hierarchical data less than the current monitoring range, and perform statistical sorting on the acquisition image data corresponding to this acquisition image data and the current monitoring range according to the hierarchical relationship to obtain the group to be analyzed; S43, If the hierarchical data of the current monitoring range is the first layer, then based on the built-in danger assessment model, perform a danger assessment on the first item data and the second item data to determine whether the item is dangerous; S44, If the hierarchical data of the current monitoring range is not the first layer, then judge the position state of the first item data, and divide the first item data into the first fixed items with unchanged item positions and the first moving items with changed item positions according to the position states of each item; S45. Based on the built-in hazard assessment model, perform an item judgment on the first fixed item to determine whether there is a hazard judgment criterion for this item in the hazard assessment model. If there is a hazard judgment criterion for this item, then perform a hazard assessment on this item based on the hazard assessment model to determine whether the item is hazardous; S46. If there is no hazard judgment criterion for this item in the hazard assessment model, then output an auxiliary signal and receive the human hazard marking result after human hazard marking. Use the human hazard marking result as the hazard assessment result for this item, and based on the human hazard marking result, perform corresponding item training on the hazard assessment model to obtain the hazard judgment criterion for the corresponding item; among them, before applying the hazard assessment model, first perform a collection on this construction scenario, perform a human hazard marking on the collected image data, and obtain the human hazard marking result. Train the machine learning algorithm based on the human hazard marking result to obtain the hazard assessment model.

[0051] S47. Based on the group to be analyzed, perform a portrait recognition on the group to be analyzed to determine whether there is a human trajectory within the monitoring range at the lower level of the current monitoring range. If it is determined that there is a human trajectory, then based on the state change index of each item in the first moving item and the second item data, determine the risk of falling of the corresponding item to obtain the first index data; S48. Perform a hazard assessment based on the built-in hazard assessment model to obtain the second index data, and based on the first index data and the second index data, determine whether the item is hazardous.

[0052] In this embodiment, by establishing a hierarchical relationship for the construction scenario, the hierarchical data and regional location information corresponding to the current monitoring area are clarified. Then, according to different hierarchical data, corresponding evaluation methods are adopted to make the evaluation result more accurate. By evaluating the items without the risk of falling using the hazard assessment model, and evaluating the items with the risk of falling using the hazard assessment model and the falling risk assessment, a comprehensive judgment on the hazards brought by the items within the current monitoring area is made, thereby improving the accuracy of the safety monitoring of the construction site.

[0053] Exemplarily, for example, when building a building, after the foundation layer is completed, a platform needs to be built for relevant workers to work when building at a high place. And this built platform is the construction layer. By marking each construction layer, the layer marked on the ground is marked as 1, and the construction layer adjacent to the ground is marked as 2, and so on until all construction layers are marked.

[0054] If the current construction floor level is 1, it indicates that the construction floor is at ground level. Therefore, there is no risk of personnel injury caused by falling objects. Hence, it is only necessary to conduct corresponding risk assessments on each item through the built-in risk assessment model. Conversely, if the current construction floor level is not 1, for example, the construction floor level is 2, it indicates that there is a possibility of personnel injury caused by falling objects. Therefore, it is not possible to directly conduct risk assessments on each item within the monitoring range through the risk assessment model. Thus, it is necessary to divide the items within this range, conduct risk assessment model evaluations on the items with fixed positions, and conduct fall risk assessments and risk assessment model evaluations on the moving items, thereby making the final assessment results more accurate and improving the accuracy of the construction site safety assessment.

[0055] Based on the description of the above-mentioned embodiment of the integrated intelligent monitoring method for smart construction sites, the embodiment of the present invention also discloses an integrated intelligent monitoring system for smart construction sites: As Figure 2 shown, an integrated intelligent monitoring system for smart construction sites, by applying the above-mentioned integrated intelligent monitoring method for smart construction sites, includes: a collection planning module 1, a first risk judgment module 2, a second risk judgment module 3, a danger judgment module 4, and an alarm module 5; The collection planning module 1 arranges monitoring cameras based on the construction scenario of the construction site and collects image data of the construction site based on the arranged monitoring cameras to obtain the collected image data; The first risk judgment module 2 performs image recognition on the collected image data based on image recognition technology to obtain item data, and conducts structural risk judgment on the item data, marking the items with risks to obtain the first item data; The second risk judgment module 3 determines the location information of the item data based on the collected image data, and conducts status risk assessment on the items according to the location information, marking the items with risks as the second item data; The danger judgment module 4 determines the danger of the items according to the first item data and the second item data, marking the items with danger as dangerous objects; The alarm module 5 makes an alarm mark for the dangerous objects when it is determined that there are dangerous objects.

[0056] Compared with the existing integrated intelligent monitoring system and method for smart construction sites, the present invention improves the accuracy of construction site safety monitoring.

[0057] The above are all the preferred embodiments of this application. Without restricting the protection scope of this application accordingly, therefore: Any equivalent changes made according to the structure, shape, and principle of this application should be covered within the protection scope of this application.

Claims

1. An integrated intelligent monitoring method for a smart construction site, characterized in that, Including: Arrange surveillance cameras based on the construction scenarios at the construction site, and collect images of the construction site based on the arranged surveillance cameras to obtain the collected image data; Based on image recognition technology, perform image recognition on the collected image data to obtain item data, and perform structural risk judgment on the item data, and mark the items with risks to obtain the first item data; Determine the location information of the item data based on the collected image data, and perform status risk assessment on the item according to the location information, and mark the items with risks as the second item data; Determine the danger of the item according to the first item data and the second item data, and mark the items with danger as dangerous items; When it is determined that there are dangerous items, make an alarm mark for the dangerous items.

2. The integrated intelligent monitoring method for a smart construction site according to claim 1, characterized in that: The arrangement of the surveillance cameras based on the construction scenarios at the construction site, and the collection of images of the construction site based on the arranged surveillance cameras to obtain the collected image data are specifically as follows: Determine the construction scenario according to the construction characteristics of the construction site; the construction scenario includes planar construction and three-dimensional construction; If the construction scenario is three-dimensional construction, divide each construction layer into planar construction with corresponding spatial relationships to obtain the construction area; Obtain the monitoring range of the surveillance cameras, and use the monitoring range as the spacing between the surveillance cameras, and arrange the surveillance cameras in the construction area according to the spacing between the surveillance cameras; Collect images of the monitoring range based on the surveillance cameras to obtain the collected image data.

3. The integrated intelligent monitoring method for an intelligent construction site according to claim 2, characterized in that: The obtaining of the monitoring range of the surveillance cameras, and using the monitoring range as the spacing between the surveillance cameras, and arranging the surveillance cameras in the construction area according to the spacing between the surveillance cameras are specifically as follows: Arbitrarily collect images of the construction area based on the surveillance cameras to obtain the image data to be analyzed; Based on image recognition technology, perform item recognition on the image data to be analyzed to obtain the item data to be judged; Compare the item data to be judged with the actual item data to determine the recognition accuracy, and compare the recognition accuracy with the preset accuracy value to determine whether the recognition accuracy meets the standard; If the recognition accuracy does not meet the standard, adjust the monitoring range of the surveillance cameras until the recognition accuracy of the images collected by the image recognition technology is greater than the preset accuracy value.

4. An integrated intelligent monitoring method for a smart construction site according to claim 3, characterized in that: The performing of image recognition on the collected image data based on image recognition technology to obtain item data, and performing structural risk judgment on the item data, and marking the items with risks to obtain the first item data are specifically as follows: Perform image recognition on the collected image data based on image recognition technology to determine all items included in the monitoring area to obtain the item list data; Determine the position judgment and shape and size judgment of each item based on the collected image data to obtain the first position information and shape information; Based on the first position information, screen the items to obtain the items whose positions are within the built-in space range, and mark the corresponding items as the first item list; Determine the acting area of the corresponding item list based on the first item list; Obtain the material properties of each item in the first item, substitute the material properties and the action area into the built-in first evaluation formula to obtain the structural risk value of the item, and match the structural risk value with the built-in first evaluation index. Mark the items with a structural risk value greater than the built-in first evaluation index as the first item data.

5. An integrated intelligent monitoring method for a smart construction site according to claim 4, characterized in that: Based on the image recognition technology, perform image recognition on the collected image data to determine all items included in the monitoring area and obtain the item data. Specifically: Based on the monitoring area of the monitoring camera and the image perspective relationship, construct a three-dimensional space. Match the three-dimensional space with the area monitored by the monitoring camera to determine the items outside the surface of the three-dimensional space in the monitoring area and mark them as items to be recognized. Based on the image recognition technology, recognize the items to be recognized to obtain the item data.

6. The integrated intelligent monitoring method for an intelligent construction site according to claim 5, characterized in that: Based on the collected image data, determine the position information of the item data, and perform a status risk assessment on the item according to the position information. Mark the items with risks as the second item data. Specifically: Based on the item data, mark the position of the item to obtain the position data. Compare the position data of the same item in multiple collected image data of the same monitoring camera with each other to determine whether the position of the item has changed. If it is determined that the position of the item has changed, then judge the reason for the position change of the item to determine whether there is human intervention. If it is determined to be human intervention, then re-mark and judge the position of the item. If it is determined to be non-human intervention, then determine the relationship between the displacement distance and time of the item according to the collected image data of the item to obtain the status change index, and determine the risk of the item according to the status change index.

7. An integrated intelligent monitoring method for an intelligent construction site according to claim 6, characterized in that: If it is determined that the position of the item has changed, then judge the reason for the position change of the item to determine whether there is human intervention. Specifically: Obtain the time period corresponding to the change in the position of the item, and perform human face recognition on the collected image data of the corresponding time period according to the corresponding time period to determine whether there is a human face on the corresponding collected image data. If it is determined that there is a human face on the collected image data, then determine the position movement data of the human face according to the collected image data, and match the position movement data of the human face with the position change data when the position of the item changes to determine whether the position movement data of the human face coincides with the position change data when the position of the item changes. If it is determined that the position movement data of the human face coincides with the position data before the position change of the item, then determine that the reason for the position change of the item is human intervention; otherwise, determine it as non-human intervention.

8. The integrated intelligent monitoring method for a smart construction site according to claim 7, characterized in that: Based on the first item data and the second item data, determine the danger of the item, and mark the items with danger as dangerous items. Specifically: Based on the construction layer, establish a hierarchical relationship for the construction scene, and determine the hierarchical data of each monitoring camera according to the hierarchical relationship. Obtain the hierarchical data and regional location information of the current monitored area, and based on the hierarchical data, regional location information, and construction scenario, obtain the collected image data located at the same regional location and with hierarchical data smaller than the current monitoring range. Then, perform statistical sorting on the collected image data corresponding to the current monitoring range and the collected image data obtained, to obtain a group to be analyzed; If the hierarchical data of the current monitoring range is the first layer, then based on the built-in risk assessment model, perform risk assessment on the first item data and the second item data to determine whether the item is dangerous; If the hierarchical data of the current monitoring range is not the first layer, then judge the position status of the first item data, and divide the first item data into the first fixed items with unchanged item positions and the first moving items with changed item positions according to the position status of each item; Based on the built-in risk assessment model, perform item judgment on the first fixed items to determine whether there is a risk judgment criterion for the item in the risk assessment model. If there is a risk judgment criterion for the item, then based on the risk assessment model, perform risk assessment on the item to determine whether the item is dangerous; If there is no risk judgment criterion for the item in the risk assessment model, then output an auxiliary signal, receive the human risk marking result after human risk marking, use the human risk marking result as the risk assessment result of the item, and based on the human risk marking result, perform corresponding item training on the risk assessment model to obtain the risk judgment criterion for the corresponding item; Based on the group to be analyzed, perform human portrait recognition on the group to be analyzed to determine whether there is a human trajectory in the monitoring range at the lower level of the current monitoring range. If it is determined that there is a human trajectory, then according to the state change index of each item in the first moving items and the second item data, determine the risk of the corresponding item falling, to obtain the first index data; Perform risk assessment based on the built-in risk assessment model to obtain the second index data, and based on the first index data and the second index data, determine whether the item is dangerous.

9. An integrated intelligent monitoring system for a smart construction site, characterized in that, The system is used to implement an integrated intelligent monitoring method for a smart construction site described in any one of claims 1-8, including: a collection planning module, a first risk judgment module, a second risk judgment module, a danger judgment module, and an alarm module; The collection planning module arranges monitoring cameras based on the construction scenario of the construction site, and performs image collection on the construction site based on the arranged monitoring cameras to obtain the collected image data; The first risk judgment module performs image recognition on the collected image data based on image recognition technology to obtain item data, and performs structural risk judgment on the item data, and marks the items with risks to obtain the first item data; The second risk judgment module determines the position information of the item data based on the collected image data, and performs state risk assessment on the item according to the position information, and marks the items with risks as the second item data; The danger judgment module determines the danger of the item according to the first item data and the second item data, and marks the items with danger as dangerous items; Alarm module, when it is determined that there is a dangerous object, make an alarm mark for the dangerous object.

Images