Unmanned aerial vehicle monitoring system and method for unsafe behaviors on building construction site

Abstract

The invention relates to the technical field of unmanned aerial vehicle monitoring, and discloses an unmanned aerial vehicle monitoring system and method for unsafe behaviors on a building construction site, and the method comprises the steps: collecting related data of article stacking in a target region through an unmanned aerial vehicle, judging whether the article stacking is normal or not, and judging whether the article stacking has a prompt function or not; according to the unmanned aerial vehicle monitoring system and method for the unsafe behaviors on the building construction site, when building materials are stacked in the building construction site, whether the stacking position of the building materials is proper or not is judged, and if yes, the unsafe behaviors on the building construction site are monitored. According to the method, whether the building material stacks collapse or not is judged, whether the building material stacks have a certain prompting effect or not is judged, an alarm prompt is given in time for unreasonable building material stacking conditions, and dangerous conditions such as casualties caused by material collapse due to improper stacking of the building materials are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle monitoring, and in particular to a system and method for monitoring unsafe behaviors of unmanned aerial vehicles at a construction site. Background Art

[0002] There are various unsafe behaviors at construction sites, such as workers not wearing safety helmets, throwing tools around, and working at heights in violation of regulations. These behaviors not only affect construction efficiency, but may also lead to serious safety accidents. Traditional safety monitoring methods often have high labor costs and limited monitoring coverage. Therefore, an efficient and intelligent monitoring system is urgently needed. By adopting a drone monitoring system, the monitoring efficiency and accuracy of unsafe behaviors at construction sites can be effectively improved, providing a strong guarantee for ensuring construction safety. With the continuous development of drone technology, the application prospects of this system are broad and it is expected to be promoted in more fields.

[0003] The drone patrols along the construction site according to the preset path, takes videos and photos regularly, obtains real-time images of the site, and analyzes the images taken by the drone through machine learning algorithms and image processing technology to identify specific unsafe behaviors. The system can continuously optimize the recognition algorithm through training data to improve detection accuracy. After identifying unsafe behaviors, the drone will send the information back to the central control system in real time through the data transmission module. The central system analyzes the risks according to the set thresholds and automatically generates alarms or notifications to on-site managers. The system can generate detailed monitoring reports, including the time, location, and type of unsafe behaviors, for easy tracking and statistical analysis. Managers can view historical records through the user interface to analyze unsafe behaviors and formulate improvement measures;

[0004] The existing drone monitoring system and method for unsafe behavior at construction sites cannot determine whether the location of the stacked building materials is appropriate, whether there is a risk of collapse of the stacked building materials, or whether the stacked building materials have a certain warning function when there is stacked building materials at the construction site. It is impossible to issue an alarm in time for unreasonable stacking of building materials, which may easily lead to material collapse due to improper stacking of building materials, resulting in casualties and other dangerous situations. Its practicality has certain limitations. Summary of the invention

[0005] The present invention provides a drone monitoring system and method for unsafe behaviors at a construction site, which are used to promote the solution of the problems mentioned in the background technology.

[0006] The present invention provides the following technical solution: a method for monitoring unsafe behaviors at a construction site by using a drone, comprising:

[0007] Get the current monitoring area of ​​the target device;

[0008] Obtain the height data at various locations within the current monitoring area and define it as the monitoring height;

[0009] Obtain all the monitoring heights within the current monitoring area to form a height set;

[0010] Obtain the reference height data;

[0011] If the monitoring height = the reference height data, it is determined that there is no item, move the target device, update the current monitoring area, and repeat the above steps;

[0012] If the monitoring height ≠ the reference height data, it is determined that there is an item, obtain all the acquisition edge points of the item to form an acquisition point set;

[0013] Obtain the analysis edge points corresponding to each element in the acquisition point set to form an analysis point set;

[0014] If the number of elements in the analysis point set ≠ the number of elements in the acquisition point set, it is determined that the target device has error data;

[0015] If the number of elements in the analysis point set = the number of elements in the acquisition point set, it is determined that the target device has no error data;

[0016] Obtain the determination area;

[0017] If all the elements in the analysis point set are within the determination area, it is determined that there are fewer unsafe factors;

[0018] If there are elements in the analysis point set that are not within the determination area, it is determined that there are more unsafe factors.

[0019] As a method for monitoring unsafe behaviors of drones at a construction site according to the present invention, wherein: the obtaining of all the acquisition edge points of the item specifically includes:

[0020] A1. Obtain all the initial images of the item;

[0021] A2. Capture all the edge lines of the item to form an edge line set;

[0022] A3. Successively recognize each element in the edge line set as a target edge line;

[0023] A4. Obtain all the data points on the target edge line to form a data point set;

[0024] A5. Obtain the coordinates of each element in the data point set and define them as data coordinates;

[0025] A6. Make each data coordinate correspond one by one to each element in the data point set to form a coordinate set;

[0026] A7. One-to-one correspondence between each coordinate set and each element in the edge line set to form an edge data set;

[0027] A8. Sequentially extract each element in the edge data set and identify it as an analysis element;

[0028] A9. Define the edge data set after removing the analysis element as the analysis data set;

[0029] A10. Compare the analysis element with each element in the analysis data set in sequence;

[0030] If there is no element in the analysis data set that is the same as the analysis element, it is determined that the analysis element has no overlapping elements;

[0031] If there is an element in the analysis data set that is the same as the analysis element, it is determined that the analysis element has overlapping elements;

[0032] A11. Identify the analysis element determined to have overlapping elements as the collected edge point.

[0033] As a method for monitoring unsafe behaviors at a construction site using a drone according to the present invention, wherein: obtaining the analysis edge points corresponding to each element in the collected point set specifically includes:

[0034] Obtain all initial images of the item;

[0035] Obtain the device data of the target device when collecting each initial image, and define it as the target device data;

[0036] Obtain the position of the item in the current monitoring area and define it as the comparison position;

[0037] Obtain the limit data of the target device;

[0038] Adjust the images collected by the target device for the item according to the limit data, comparison position, and each target device data, and define it as the comparison image;

[0039] Adjust the images collected by the target device for the item according to the comparison position and target device data, and define it as the comparison image;

[0040] Execute A2 - A10;

[0041] Identify the analysis element determined to have overlapping elements as the analysis edge point.

[0042] As a method for monitoring unsafe behaviors at a construction site using a drone according to the present invention, wherein: if it is determined that the target device has error data, perform acquisition error analysis on the item:

[0043] Obtain all comparison images of the item;

[0044] Extract the number of obstacles in the comparison image;

[0045] Obtain the occupied area data of each obstacle, where the occupied area data includes volume data and area data;

[0046] Calculate the volume influence data in the comparison image, where the volume influence data = the sum of the volume data of each obstacle under the number of obstacles;

[0047] Calculate the area influence data in the comparison image, where the area influence data = the sum of the area data of each obstacle under the number of obstacles;

[0048] If the volume influence data ≥ 15% of the area influence data, it is determined that the item placement position is reliable and the number of unsafe factors is determined to be small;

[0049] If the volume influence data < 15% of the area influence data, it is determined that the item placement position is unreliable and an abnormal alarm is issued.

[0050] As a method for monitoring unsafe behaviors of drones at a construction site according to the present invention, where: if it is determined that the number of unsafe factors is small, stack analysis is performed on the items:

[0051] Obtain all the items in the current monitoring area to form an item set;

[0052] Obtain the type data of each item in the item set;

[0053] If the type data of all the items in the item set are of the same type data, randomly extract an item from the item set and identify it as the target item;

[0054] Obtain the center point of the target item and define it as the target center point;

[0055] Obtain all the items adjacent to the target item in the item set to form an analysis item set;

[0056] Successively identify each element in the analysis item set as an analysis item;

[0057] Obtain all the surfaces of the target item to form a first surface set;

[0058] Extract the surface in the first surface set that is closest to the analysis item and define it as the first plane;

[0059] Obtain all the surfaces of the analysis item to form a second surface set;

[0060] Extract the surface in the second surface set that is closest to the target item and define it as the second plane;

[0061] Set a number of data analysis points on the first plane;

[0062] Obtain the vertical distance between each data analysis point and the second plane, and define it as the analysis distance;

[0063] Obtain all the analysis distances to form a distance set;

[0064] Set the judgment distance;

[0065] If the analysis distance ≤ the judgment distance × (1 + 50%), it is determined that the judgment distance interval is small;

[0066] If the analysis distance > the judgment distance × (1 + 50%), it is determined that the judgment distance interval is large;

[0067] If all elements in the distance set are determined to have a small judgment distance interval, it is determined that the item interval is small;

[0068] If there is an element in the distance set determined to have a large judgment distance interval, it is determined that the item interval is large;

[0069] If all elements in the analyzed item set are determined to have a small item interval, it is determined that the items are stable;

[0070] If there is an element in the analyzed item set determined to have a large item interval, it is determined that the items are unstable;

[0071] Obtain the number of elements in the item set determined to be stable, and define it as the stable number;

[0072] Obtain the number of elements in the item set, and define it as the item number;

[0073] If the stable number ≥ 90% of the item number, it is determined that the item stacking is normal and no abnormal alarm is made;

[0074] If the stable number < 90% of the item number, it is determined that the item stacking is abnormal and an abnormal alarm is made.

[0075] As a method for monitoring unsafe behaviors of drones at a construction site according to the present invention, wherein: if it is determined that there are fewer unsafe factors, stack analysis is performed on the items, and it further includes:

[0076] Obtain all the items in the current monitoring area to form an item set;

[0077] Obtain the type data of each item in the item set;

[0078] If the type data of all items in the item set are not of the same type, each type data is sequentially identified as the target type;

[0079] Extract all the items corresponding to the target type in the item set to form a judgment set;

[0080] Arbitrarily extract an item from the judgment set and define it as the judgment item;

[0081] Define the determination set of the items to be removed as the determination analysis set;

[0082] Define each item in the determination analysis set as a score determination item in turn;

[0083] Obtain the number of items between the determination item and the score determination item, and define it as the determination quantity;

[0084] If the determination quantity = 0, then determine that the score determination item is a continuous item;

[0085] If the determination quantity ≠ 0, then extract all the items between the determination item and the score determination item, and define them as interval items;

[0086] Obtain the type data of each interval item, and define it as the analysis type;

[0087] If the analysis type = the target type, then determine that the interval items are the same items;

[0088] If the analysis type ≠ the target type, then determine that the interval items are different items;

[0089] If all the interval items are the same items, then determine that the score determination item is a continuous item;

[0090] If there are different items among all the interval items, then determine that the score determination item is a discontinuous item;

[0091] If all the elements in the determination analysis set are continuous items, then determine that the items are concentrated;

[0092] If there are elements in the determination analysis set that are discontinuous items, then determine that the items are dispersed;

[0093] If all the elements in the item set are determined to be concentrated items, then determine that the items are stacked normally and no abnormal alarm is made;

[0094] If there are elements in the item set that are determined to be dispersed items, then determine that the items are stacked abnormally and an abnormal alarm is made.

[0095] As a method for monitoring unsafe behaviors at a construction site using a drone according to the present invention, wherein: if it is determined that there are many unsafe factors, then environmental analysis is performed on the items:

[0096] B1. Obtain the target area;

[0097] B2. Obtain all the area data of the target area;

[0098] B3. Obtain all the items within the current monitoring area to form an item set;

[0099] B4. Obtain the spatial coordinates of each item in the item set and define them as item coordinates;

[0100] B5. Arbitrarily extract an item coordinate and identify it as the target coordinate;

[0101] B6. Obtain the distance between each item coordinate and the target coordinate and define it as the coordinate distance;

[0102] B7. Select the item coordinate corresponding to the smallest coordinate distance value and identify it as the selected coordinate;

[0103] B8. Connect the target coordinate and the selected coordinate to form a reference connection line;

[0104] B9. Remove the target coordinate from the item set and update the item set;

[0105] B10. Update the selected coordinate to the target coordinate;

[0106] B11. Repeat B6 - B11 until there are no elements in the updated item set and then stop, and identify the selected coordinate corresponding to the finally formed reference connection line as the termination coordinate;

[0107] B12. Identify the target coordinate in B5 as the initial coordinate;

[0108] B13. Connect the initial coordinate and the termination coordinate to form an analysis area;

[0109] B14. Obtain the regional data within the analysis area and define it as the analyzed regional data;

[0110] The regional data includes ordinary regional data and special regional data;

[0111] B15. If the analyzed regional data is ordinary regional data, then determine that there is no prompt for item stacking and perform an abnormal alarm;

[0112] If the analyzed regional data is special regional data, then determine that there is a prompt for item stacking and mark the analysis area as an abnormal area for key monitoring.

[0113] As a method for monitoring unsafe behaviors of drones at a construction site according to the present invention, wherein: key monitoring of the abnormal area includes:

[0114] Obtain the monitoring interval duration of the target device;

[0115] Obtain the device condition data of the target area and define it as the target condition data;

[0116] The device condition data includes the existence condition and the non - existence condition;

[0117] If the target condition data is an existence condition, cancel the monitoring interval duration;

[0118] Arbitrarily select a target device to monitor the abnormal area;

[0119] If the target condition data is a non-existence condition, calculate the adjusted interval duration, and the adjusted interval duration = monitoring interval duration × 20%;

[0120] Adjust the monitoring interval duration of the target device to the adjusted interval duration to monitor the abnormal area;

[0121] When an item is removed from the abnormal area, an abnormal alarm is issued.

[0122] As the present invention also discloses a system for implementing an unmanned aerial vehicle monitoring method for unsafe behaviors at a construction site, including:

[0123] Acquisition module: Collect relevant data on item stacking in the target area through an unmanned aerial vehicle;

[0124] Analysis module: Analyze the area where item stacking exists in the target area to determine whether the item stacking is normal;

[0125] Determination module: Analyze the situation of abnormal item stacking to determine whether the item stacking has a prompting effect;

[0126] Alarm module: For the situation of abnormal item stacking without a prompting effect and the situation where an item is removed from the abnormal area, an abnormal alarm is issued in a timely manner to avoid accidents.

[0127] The present invention has the following beneficial effects:

[0128] 1. The unmanned aerial vehicle monitoring system and method for unsafe behaviors at the construction site can determine whether there is item stacking by obtaining height data at various locations in the construction site, and determine whether there is data loss by comparing the data captured by the initial image of the item with the data captured by the currently collected comparison image, so as to determine whether there is an error in the data collected by the unmanned aerial vehicle. If there is no error, it is determined whether the placement position of the item is reliable and whether there is a risk. For the situation of unreasonable stacking of construction materials, an alarm prompt is issued in a timely manner, reducing the occurrence of dangerous situations such as casualties caused by the collapse of construction materials due to improper stacking, avoiding the influence of the stacking of construction materials on the operation of construction equipment, and avoiding the danger caused by the collapse of construction materials due to the operation of construction equipment.

[0129] 2. For the UAV monitoring system and method for unsafe behaviors at the construction site of this building, if there are errors in the data collected by the UAV, then by obtaining the position where the items are placed, it is judged whether the items are stacked in a suitable area. If they are in a suitable area, then by obtaining data such as the material, size of the items, and the spacing distance between each item, it is judged whether there are potential safety hazards in the stacking of the items, that is, it is judged whether there is a risk of collapse in the stacking of the items. If there are potential safety hazards, an alarm prompt is given in a timely manner. If there are no potential safety hazards, continuous monitoring can be carried out, reducing the occurrence of dangerous situations such as casualties caused by the collapse of building materials due to improper stacking.

[0130] 3. For the UAV monitoring system and method for unsafe behaviors at the construction site of this building, if the items are not stacked in a suitable area, then by obtaining the terrain data in the area and obtaining the spatial coordinates of each item, it is judged whether there is a prompting effect in the stacking of the items, that is, it is judged whether there is a situation where some people stack a number of items to form a closed area and enclose special terrains such as elevator shafts within this closed area. If there is a prompting effect, the closed area is marked and key monitored. If there is no prompting effect, an alarm prompt is given in a timely manner, reducing the occurrence of dangerous situations such as casualties caused by the collapse of building materials due to improper stacking, preventing workers from falling into danger due to not noticing this terrain during transportation and work, and at the same time avoiding the situation where some people do not notify other people in time after removing the stacked items, resulting in dangerous situations for other people due to untimely information transmission and not paying attention to the environmental situation. Description of the Drawings

[0131] Figure 1 It is a flowchart of the UAV monitoring method for unsafe behaviors at the construction site of the present invention;

[0132] Figure 2 It is a system block diagram for implementing the UAV monitoring method for unsafe behaviors at the construction site of the present invention. Detailed Embodiments

[0133] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0134] Embodiment 1. A UAV monitoring method for unsafe behaviors at the construction site of a building, referring to Figure 1 , including:

[0135] Obtain the current monitoring area of the target device, where the target device is a UAV;

[0136] Obtain the height data at various locations within the current monitoring area and define it as the monitoring height. The height data is the height distance between a certain point and the ground.

[0137] Obtain all the monitoring heights within the current monitoring area to form a height set.

[0138] Obtain the reference height data, where the reference height data is 0, i.e., the ground.

[0139] If the monitoring height = the reference height data, it is determined that there is no item, move the target device, update the current monitoring area, and repeat the above steps.

[0140] If the monitoring height ≠ the reference height data, it is determined that there is an item, obtain all the acquisition edge points of the item to form an acquisition point set.

[0141] Obtain the analysis edge points corresponding to each element in the acquisition point set to form an analysis point set.

[0142] If the number of elements in the analysis point set ≠ the number of elements in the acquisition point set, it is determined that the target device has error data.

[0143] If the number of elements in the analysis point set = the number of elements in the acquisition point set, it is determined that the target device has no error data.

[0144] Obtain the determination area, where the determination area is a pre-planned and set area where items can be placed.

[0145] If all elements in the analysis point set are within the determination area, it is determined that there are fewer unsafe factors, that is, all items are placed in the pre-planned and set area where items can be placed.

[0146] If there are elements in the analysis point set that are not within the determination area, it is determined that there are more unsafe factors, that is, there are items not placed in the determination area.

[0147] Among them, the specific method for obtaining all the acquisition edge points of the item is as follows:

[0148] A1. Obtain all the initial images of the item. The initial images are the images captured by the drone of the item when the item is in the area where the drone can collect complete data of the item before being placed in the current monitoring area.

[0149] A2. Capture all the edge lines of the item to form an edge line set.

[0150] A3. Successively identify each element in the edge line set as the target edge line.

[0151] A4. Obtain all the data points on the target edge line to form a data point set.

[0152] A5. Obtain the coordinates of each element in the data point set, which are defined as data coordinates;

[0153] A6. One-to-one correspond each data coordinate with each element in the data point set to form a coordinate set;

[0154] A7. One-to-one correspond each coordinate set with each element in the edge line set to form an edge data set. Since the item is in a stationary state when all the initial images of the item are collected, all the coordinates of the item collected at this time are within the same coordinate axis, that is, the data is referenceable and has less error;

[0155] A8. Sequentially extract each element in the edge data set and identify it as an analysis element;

[0156] A9. Define the edge data set after removing the analysis element as the analysis data set;

[0157] A10. Compare the analysis element with each element in the analysis data set in sequence;

[0158] If there is no element in the analysis data set that is the same as the analysis element, it is determined that the analysis element has no overlapping elements;

[0159] If there is an element in the analysis data set that is the same as the analysis element, it is determined that the analysis element has overlapping elements;

[0160] A11. Identify the analysis element determined to have overlapping elements as the collected edge point.

[0161] This embodiment also provides that obtaining the analysis edge point corresponding to each element in the collection point set specifically includes:

[0162] Obtain all the initial images of the item;

[0163] Obtain the device data when the target device collects each initial image, which is defined as the target device data. The device data are parameters such as the relative position and distance between the drone and the item when collecting the target image;

[0164] Obtain the position of the item in the current monitoring area, which is defined as the comparison position;

[0165] Obtain the limit data of the target device. The limit data are the limit conditions under which the drone cannot normally perform image collection, such as the presence of obstacles that prevent the drone from reaching the predetermined position for data collection;

[0166] According to the restricted data, comparison position, and each target device data, adjust the image collected by the target device for the item, and define it as the comparison image. The comparison image is the image of the item in the current monitoring area after the UAV adjusts its attitude and position to the data when collecting the target image, ensuring that the comparison image and the target image are two images with basically the same data and can be used for data comparison;

[0167] According to the comparison position and target device data, adjust the image collected by the target device for the item, and define it as the comparison image. The comparison image is the image of the item in the current monitoring area after the UAV adjusts its attitude and position to the data when collecting the target image, ensuring that the comparison image and the target image are two images with basically the same data and can be used for data comparison;

[0168] Execute A2 - A10;

[0169] Identify the analysis elements with overlapping elements as analysis edge points.

[0170] Through the above method, use machine vision technology to analyze and process the construction site images and videos, identify non - compliant manipulation behaviors and potential dangerous situations, construct a real - time warning system. Combining the results of construction danger behavior recognition and construction worker behavior analysis, it can timely detect dangerous behaviors and send warning information. The information is planned to be sent through alarms, vibration or whistling of the supporting safety bracelets, or connecting to the personal mobile phones of construction workers, and timely reminder through mobile phones. It can automatically identify non - standard manipulations and potential dangers, and monitor the behaviors of construction workers in real time, providing accurate warning information so as to take timely and effective measures to ensure the safety of the construction site.

[0171] Embodiment 2. This embodiment is an improvement based on Embodiment 2. For the UAV monitoring method of unsafe behaviors at the construction site, if it is determined that the target device has error data, then conduct an acquisition error analysis on the item:

[0172] Obtain all comparison images of the item;

[0173] Extract the number of obstacles in the comparison image;

[0174] Obtain the occupied area data of each obstacle. The occupied area data includes volume data and area data. The occupied area data is the volume and occupied area of the obstacle. The volume data is the volume of the obstacle, and the area data is the occupied area of the obstacle, so as to judge the type of the obstacle. For example, a wall has a large volume and a large occupied area, while a tower crane has a large volume and a small occupied area;

[0175] Calculate the volume influence data in the comparison image. The volume influence data = the sum of the volume data of each obstacle under the number of obstacles;

[0176] Calculate the area influence data in the comparison image, where the area influence data = the sum of the area data of each obstacle under the number of obstacles;

[0177] If the volume influence data ≥ 15% of the area influence data, it is determined that the placement position of the item is reliable and the unsafe factors are few, that is, when the building materials are placed near fixed objects such as walls, the probability of dangerous situations such as collisions of the items is small. As long as the building materials are stacked reasonably, the building materials can be placed at this position;

[0178] If the volume influence data < 15% of the area influence data, it is determined that the placement position of the item is unreliable and an abnormal alarm is given, that is, when the building materials are placed near equipment such as tower cranes, dangerous situations such as collisions of the items may occur when the equipment is operating, and the building materials need to be moved to a suitable position as soon as possible.

[0179] Example 3. This example is an improvement based on Example 3. In this example, if it is determined that the unsafe factors are few, then a stacking analysis is performed on the items:

[0180] Obtain all the items in the current monitoring area to form an item set;

[0181] Obtain the type data of each item in the item set. The type data is data such as the material and size of the item to determine whether the items stacked in this area are of the same type;

[0182] If the type data of all the items in the item set are of the same type data, then randomly extract an item from the item set and identify it as the target item;

[0183] Obtain the center point of the target item and define it as the target center point;

[0184] Obtain all the items adjacent to the target item in the item set to form an analysis item set;

[0185] Successively identify each element in the analysis item set as an analysis item;

[0186] Obtain all the surfaces of the target item to form a first surface set;

[0187] Extract the surface in the first surface set that is closest to the analysis item and define it as the first plane;

[0188] Obtain all the surfaces of the analysis item to form a second surface set;

[0189] Extract the surface in the second surface set that is closest to the target item and define it as the second plane;

[0190] Set a number of data analysis points on the first plane;

[0191] Obtain the vertical distance between each data analysis point and the second plane, and define it as the analysis distance;

[0192] Obtain all the analysis distances to form a distance set;

[0193] Set a determination distance, where the determination distance is the gap distance between the upper and lower items when the items are stacked normally. If the interval distances between the upper, lower, left, and right items and it are different when the items are stacked, it is likely to cause the center of gravity to be unstable, thus easily leading to dangerous situations such as the collapse of the items;

[0194] If the analysis distance ≤ the determination distance × (1 + 50%), then the determination distance interval is small;

[0195] If the analysis distance > the determination distance × (1 + 50%), then the determination distance interval is large;

[0196] If all elements in the distance set are determined to have a small determination distance interval, then it is determined that the item interval is small;

[0197] If there are elements in the distance set determined to have a large determination distance interval, then it is determined that the item interval is large;

[0198] If all elements in the analyzed item set are determined to have a small item interval, then it is determined that the items are stable;

[0199] If there are elements in the analyzed item set determined to have a large item interval, then it is determined that the items are unstable;

[0200] Obtain the number of elements in the item set determined to be stable, and define it as the stable number;

[0201] Obtain the number of elements in the item set, and define it as the item number;

[0202] If the stable number ≥ 90% of the item number, then it is determined that the item stacking is normal and no abnormal alarm is made;

[0203] If the stable number < 90% of the item number, then it is determined that the item stacking is abnormal and an abnormal alarm is made.

[0204] Among them, if it is determined that there are fewer unsafe factors, then for the item stacking analysis, it further includes:

[0205] Obtain all the items in the current monitoring area to form an item set;

[0206] Obtain the type data of each item in the item set, where the type data is data such as the material and size of the item to determine whether the items stacked in this area are of the same type;

[0207] If the type data of all items in the item set are not of the same type data, then each type data is sequentially recognized as the target type;

[0208] Extract all the items corresponding to the target type within the item set to form a determination set;

[0209] Arbitrarily extract an item from the determination set and define it as the determination item;

[0210] Recognize the determination set after removing the determination item as the determination analysis set;

[0211] Successively recognize each item in the determination analysis set as a score determination item;

[0212] Obtain the number of items between the determination item and the score determination item, and define it as the determination quantity;

[0213] If the determination quantity = 0, then determine that the score determination item is a continuous item;

[0214] If the determination quantity ≠ 0, then extract all the items between the determination item and the score determination item and define them as interval items;

[0215] Obtain the type data of each interval item and define it as the analysis type;

[0216] If the analysis type = the target type, then determine that the interval items are identical items;

[0217] If the analysis type ≠ the target type, then determine that the interval items are different items;

[0218] If all the interval items are identical items, then determine that the score determination item is a continuous item;

[0219] If there are different items among all the interval items, then determine that the score determination item is a discontinuous item;

[0220] If all the elements in the determination analysis set are continuous items, then determine that the items are concentrated;

[0221] If there are elements in the determination analysis set that are discontinuous items, then determine that the items are scattered;

[0222] If all the elements in the item set are determined to be concentrated items, then determine that the item stacking is normal and no abnormal alarm is made;

[0223] If there are elements in the item set that are determined to be scattered items, then determine that the item stacking is abnormal and an abnormal alarm is made.

[0224] This embodiment also provides that if there are many determined unsafe factors, then environmental analysis is performed on the items:

[0225] B1. Obtain the target area, where the target area is all the areas monitored by the UAV at the construction site;

[0226] B2. Obtain all regional data of the target area. The regional data is the structural distribution within this area, such as the distribution data of structures like stairs and walls within the area;

[0227] B3. Obtain all items within the current monitoring area to form an item set;

[0228] B4. Obtain the spatial coordinates of each item in the item set and define them as item coordinates;

[0229] B5. Arbitrarily extract one item coordinate and identify it as the target coordinate;

[0230] B6. Obtain the distance between each item coordinate and the target coordinate and define it as the coordinate distance;

[0231] B7. Select the item coordinate corresponding to the smallest coordinate distance value and identify it as the selected coordinate;

[0232] B8. Connect the target coordinate and the selected coordinate to form a reference connection line;

[0233] B9. Remove the target coordinate from the item set and update the item set;

[0234] B10. Update the selected coordinate to the target coordinate;

[0235] B11. Repeat B6 - B11 until there are no elements in the updated item set and then stop, and identify the selected coordinate corresponding to the finally formed reference connection line as the termination coordinate;

[0236] B12. Identify the target coordinate in B5 as the initial coordinate;

[0237] B13. Connect the initial coordinate and the termination coordinate to form an analysis area;

[0238] B14. Obtain the regional data within the analysis area and define it as the analysis regional data;

[0239] The regional data includes ordinary regional data and special regional data. The ordinary regional data is ordinary terrain such as the ground that is not likely to cause dangerous situations for personnel, and the special regional data is special terrain such as stairs and elevator shafts that are likely to cause dangerous situations for personnel;

[0240] B15. If the analysis regional data is ordinary regional data, then determine that there is no prompt for item stacking and conduct an abnormal alarm;

[0241] If the analyzed regional data is special regional data, it is determined that there is a prompt for item stacking, and the analyzed area is marked as an abnormal area for key monitoring. That is, there are terrains such as elevator shafts in this area that are likely to cause people to fall. Stacking items in this area to form an enclosed area may be to prevent workers from falling into danger due to not noticing this terrain during transportation and work. Mark this location and shorten the monitoring interval duration for this location. As much as possible, conduct real-time monitoring under permitted conditions to avoid some people not notifying others in time after removing the stacked items, resulting in dangerous situations for other people due to untimely information transmission and not paying attention to the environmental conditions.

[0242] Among them, key monitoring of the abnormal area includes:

[0243] Obtain the monitoring interval duration of the target device. The monitoring interval duration is the duration between two consecutive monitors of each monitoring area manually controlled according to the size and complexity of the target area;

[0244] Obtain the device condition data of the target area and define it as the target condition data. The device condition data is whether there are sufficient drones in this construction site to monitor the construction site;

[0245] The device condition data includes the existence condition and the non-existence condition. The existence condition means that there are more than two drones monitoring the target area of this construction site, indicating that there are sufficient drones in this construction site to monitor the construction site. The non-existence condition means that there are no more than two drones monitoring the target area of this construction site, indicating that there are not enough drones in this construction site to monitor the construction site;

[0246] If the target condition data is the existence condition, cancel the monitoring interval duration;

[0247] Arbitrarily select a target device to monitor the abnormal area;

[0248] If the target condition data is the non-existence condition, calculate the adjusted interval duration. The adjusted interval duration = monitoring interval duration × 20%;

[0249] Adjust the monitoring interval duration of the target device to the adjusted interval duration to monitor the abnormal area;

[0250] When there is a situation where an item is removed in the abnormal area, an abnormal alarm is issued. That is, promptly let the staff arrange warning signs and other facilities with warning effects to avoid some people not notifying others in time after removing the stacked items, resulting in dangerous situations for other people due to untimely information transmission and not paying attention to the environmental conditions.

[0251] Example 4. This example also discloses a system for implementing an unmanned aerial vehicle (UAV) monitoring method for unsafe behaviors at a construction site. Refer to Figure 2 , including:

[0252] Acquisition module: Collect relevant data on stacked items in the target area through a UAV;

[0253] Analysis module: Analyze the areas with stacked items in the target area to determine whether the stacked items are normal;

[0254] Judgment module: Analyze the situation of abnormal stacked items to determine whether the stacked items have a prompting effect;

[0255] Alarm module: Timely give an abnormal alarm for the situation of abnormal stacked items without a prompting effect and the situation of item removal in the abnormal area to avoid accidents.

[0256] In this example, aiming at the problems of accurately monitoring and timely determining irregular operations and potential dangers during the construction process, accurately judging the construction workers involved and giving direct, effective and timely early warning interventions, relying on the building construction site safety monitoring technology based on machine vision detection, it is planned to use devices such as cameras and sensors to obtain data on the construction site, to make up for the research of traditional construction site monitors that have no intelligent recognition of dangerous behaviors and no active intervention and early warning by artificial intelligence. By collecting a large amount of data, improve the dangerous behavior image recognition database of building construction safety technology research, establish a dangerous behavior sample library, and conduct machine learning through big data to improve the recognition and judgment accuracy rate of different construction dangerous behaviors. Based on this, process and analyze these data collected by machine vision to solve the problems of real-time intelligent monitoring of the safety status of the construction site, pre-judgment of dangerous behaviors and early warning interventions.

Claims

1. A drone monitoring method for unsafe behavior at a construction site, characterized by: include: Get the current monitoring area of ​​the target device; Obtain the height data of each location in the current monitoring area and set it as the monitoring height; Get all monitoring heights in the current monitoring area to form a height set; Obtaining reference height data; If the monitoring height = the reference height data, it is determined that there is no object, the target device is moved, the current monitoring area is updated, and the above steps are repeated; If the monitoring height ≠ the reference height data, it is determined that there is an object, and all the collection edge points of the object are obtained to form a collection point set; Obtain the analysis edge point corresponding to each element in the collection point set to form an analysis point set; If the number of elements in the analysis point set ≠ the number of elements in the acquisition point set, it is determined that the target device has error data; If the number of elements in the analysis point set = the number of elements in the acquisition point set, it is determined that the target device has no error data; Get the determination area; If all elements in the analysis point set are within the determination area, it is determined that there are fewer unsafe factors; If there are elements in the analysis point set that are not in the judgment area, it is judged that there are many unsafe factors.

2. The method for monitoring unsafe behaviors at construction sites by using drones according to claim 1, characterized in that: The method of obtaining all the collected edge points of the object is specifically as follows: A1. Obtain all initial images of the object; A2. Capture all edge lines of the object to form an edge line set; A3, identifying each element in the edge line set as a target edge line in turn; A4. Obtain all data points on the target edge line to form a data point set; A5. Get the coordinates of each element in the data point set and define them as data coordinates; A6. Match each data coordinate with each element in the data point set one by one to form a coordinate set; A7. Match each coordinate set with each element in the edge line set one by one to form an edge data set; A8. Extract each element in the edge data set in turn and identify it as an analysis element; A9. The marginal data set with the analysis elements removed is defined as the analysis data set; A10. Compare the analysis element with each element in the analysis data set in turn; If there is no element identical to the analysis element in the analysis data set, it is determined that there is no overlapping element for the analysis element; If there is an element identical to the analysis element in the analysis data set, it is determined that there is an overlapping element with the analysis element; A11. The analysis elements of the elements that are determined to be overlapped are identified as acquisition edge points.

3. The method for monitoring unsafe behaviors at construction sites by using drones according to claim 2, characterized in that: The step of obtaining the analysis edge point corresponding to each element in the collection point set is specifically: Get all the initial images of the item; Obtain the device data of the target device when it captures each initial image, and define it as the target device data; Obtain the location of the object in the current monitoring area and set it as the comparison location; Get restricted data of the target device; According to the restriction data, the comparison position and the data of each target device, the target device is adjusted to capture the image of the object and is defined as the comparison image; According to the comparison position and the target device data, the target device is adjusted to capture the image of the object and is defined as the comparison image; Execute A2-A10; Analysis elements where it is determined that there are overlapping elements are identified as analysis edge points.

4. The method for monitoring unsafe behaviors at a construction site by using a drone according to claim 1, characterized in that: If it is determined that the target device has error data, the item is collected for error analysis: Get all the comparison images of the object; Extract the number of obstacles in the comparison image; Acquire the occupation data of each obstacle, wherein the occupation data includes volume data and area data; Calculate volume impact data in the comparison image, where volume impact data = the sum of volume data of each obstacle under the number of obstacles; Calculate the area impact data in the comparison image, where the area impact data = the sum of the area data of each obstacle under the number of obstacles; If the volume impact data ≥ the area impact data × 15%, the item placement is determined to be reliable and the unsafe factors are less; If the volume impact data is less than the area impact data × 15%, the item placement position is determined to be unreliable and an abnormal alarm is issued.

5. A method for monitoring unsafe behaviors at a construction site by using a drone according to claim 1 or 4, characterized in that: If it is determined that there are fewer unsafe factors, the items are stacked and analyzed: Get all items in the current monitoring area to form an item set; Get the type data of each item in the item collection; If the type data of all items in the item set are the same type data, then any item in the item set is randomly selected and identified as the target item; Obtain the center point of the target object and set it as the target center point; Obtain all items in the item set that are adjacent to the target item to form an analysis item set; identifying each element in the analysis item set as an analysis item in turn; Obtain all surfaces of the target object to form a first surface set; Extract the surface closest to the analyzed object in the first face set and define it as the first plane; Acquire all surfaces of the analyzed object to form a second surface set; Extract the surface closest to the target object in the second face set and define it as the second plane; Setting a number of data analysis points on the first plane; Obtain the vertical distance between each data analysis point and the second plane, and define it as the analysis distance; Get all analyzed distances to form a distance set; Set the judgment distance; If the analysis distance ≤ the judgment distance × (1 + 50%), the judgment distance interval is small; If the analysis distance > judgment distance × (1 + 50%), the judgment distance interval is large; If all elements in the distance set are judged to have a small distance interval, then the item interval is judged to be small; If there is an element in the distance set that determines the distance interval to be large, then the item interval is determined to be large; If all elements in the analyzed item set are judged to have small item intervals, then the items are judged to be stable; If there is a large gap between elements and items in the analyzed item set, the item is considered unstable; Get the number of elements in the item set that determine the item is stable, and set it as the stable number; Get the number of elements in the item collection and set it as the number of items; If the stable quantity ≥ the number of items × 90%, the stacking of items is considered normal and no abnormal alarm is issued; If the stable quantity is less than the number of items × 90%, the stacking of items is judged to be abnormal and an abnormal alarm is issued.

6. The method for monitoring unsafe behaviors at a construction site by using a drone according to claim 5, characterized in that: If it is determined that there are few unsafe factors, the items are stacked and analyzed, including: Get all items in the current monitoring area to form an item set; Get the type data of each item in the item collection; If the type data of all items in the item set are not the same type data, each type data is identified as the target type in turn; Extract all items corresponding to the target type in the item set to form a judgment set; Randomly extract an item from the judgment set and define it as the judgment item; The decision set with the decision items removed is identified as a decision analysis set; Identify each item in the judgment analysis set as a judgment item in turn; Obtain the number of items between the determined items and the separated items, and define it as the determined number; If the determination quantity = 0, the item to be determined is a continuous item; If the judgment quantity ≠ 0, all the objects between the judgment object and the judged object are extracted and defined as the interval objects; Get the type data of each interval item and set it as the analysis type; If the analysis type = target type, the separated objects are determined to be the same object; If the analysis type ≠ the target type, the separated items are determined to be different items; If all the separated objects are the same object, the objects to be divided are judged as continuous objects; If there are different objects among all the separated objects, the objects to be judged are considered as discontinuous objects; If all elements in the judgment analysis set are continuous items, then the judgment item is concentrated; If there are elements in the judgment analysis set that are discontinuous items, the items are judged to be scattered; If all elements in the item set are determined to be in an item concentration, the item stacking is determined to be normal and no abnormal alarm is issued; If there are elements in the item set that determine that the items are scattered, the stacking of the items is determined to be abnormal and an abnormal alarm is issued.

7. The method for monitoring unsafe behaviors at a construction site by using a drone according to claim 1, characterized in that: If it is determined that there are many unsafe factors, an environmental analysis of the item will be conducted: B1. Obtain the target area; B2. Obtain all regional data of the target area; B3. Obtain all items in the current monitoring area to form an item set; B4. Obtain the spatial coordinates of each item in the item set and define them as item coordinates; B5. Randomly extract the coordinates of an object and identify them as the target coordinates; B6. Obtain the distance between each object coordinate and the target coordinate, and define it as the coordinate distance; B7. Select the coordinate distance with the smallest value corresponding to the item coordinate and identify it as the selected coordinate; B8. Connect the target coordinates and the selected coordinates to form a reference connection line; B9, remove the target coordinates from the item set and update the item set; B10, update the selected coordinates to the target coordinates; B11, repeat B6-B11 until there are no elements in the updated item set, and stop, and identify the selected coordinates corresponding to the last formed reference connection line as the end coordinates; B12, identify the target coordinates in B5 as the initial coordinates; B13, connecting the initial coordinates and the final coordinates to form an analysis area; B14. Obtain regional data within the analysis area and define it as analysis area data; The regional data includes general regional data and special regional data; B15. If the analyzed area data is ordinary area data, it is determined that there is no prompt for item stacking, and an abnormal alarm is issued; If the analysis area data is special area data, it is determined that there is a prompt for item stacking, and the analysis area is marked as an abnormal area for key monitoring.

8. The method for monitoring unsafe behaviors at a construction site by using a drone according to claim 7, characterized in that: Focus on monitoring abnormal areas, including: Get the monitoring interval duration of the target device; Acquire equipment condition data of the target area and set it as target condition data; The device condition data includes existence conditions and non-existence conditions; If the target condition data is an existing condition, the monitoring interval duration is cancelled; Randomly select a target device to monitor the abnormal area; If the target condition data is non-existent, the adjustment interval duration is calculated, and the adjustment interval duration = monitoring interval duration × 20%; Adjust the monitoring interval of the target device to the adjustment interval to monitor the abnormal area; When items are removed from the abnormal area, an abnormal alarm is issued.

9. A system for executing the method for monitoring unsafe behaviors at construction sites by using a drone according to claim 1, characterized in that: include: Collection module: Use drones to collect data related to the stacking of items in the target area; Analysis module: Analyze the areas where objects are stacked in the target area to determine whether the stacking of objects is normal; Judgment module: Analyzes abnormal item stacking and determines whether the item stacking has a prompt effect; Alarm module: Promptly issue abnormal alarms for abnormal stacking of items that do not have a warning function and for items removed from abnormal areas to avoid accidents.

Images