Building house and facade hollowing unmanned aerial vehicle AI detection method, system and terminal

Abstract

The invention relates to a building house and facade hollowing unmanned aerial vehicle AI detection method, system and terminal, and relates to the technical field of unmanned aerial vehicles, and the method comprises the steps: obtaining a building range of a house building; determining a flight height according to the building range and a preset shooting range; a preset unmanned aerial vehicle is controlled to fly to the building range from the preset reference parking position, and image detection information is acquired according to the flight height; determining a detected building specification according to the image detection information and preset building features; determining a routing inspection path according to the detected building specification, controlling a preset unmanned aerial vehicle to fly along the routing inspection path, and obtaining routing inspection image information; and forming a building model according to the inspection image information, determining a crack position according to the building model and a preset crack feature, and outputting preset abnormal information according to the crack position. The method has the effect of improving the accuracy of the inspection result.

Description

Technical Field

[0001] The present invention relates to the field of drone technology, and in particular to a drone AI detection method, system, and terminal for hollowing of buildings and facades. Background Art

[0002] A drone is an unmanned aerial vehicle controlled by a radio remote control device or its own program control device.

[0003] During the construction process, operators need to conduct inspections. When inspecting the facade of a house, the operator first needs to observe the bottom of the facade, then perform aerial work to inspect the facade from a high position, and record the inspection data. The recorded data can be used to determine whether there are cracks or hollows on the facade.

[0004] During the inspection of the facade of a house, manual inspection may result in incomplete data records, leading to errors in the inspection results. Summary of the Invention

[0005] In order to improve the accuracy of inspection results, the present invention provides a drone AI detection method, system and terminal for hollowing out of buildings and facades.

[0006] In a first aspect, the present invention provides a drone AI detection method for hollowing of buildings and facades, which adopts the following technical solutions: A drone AI detection method for hollowing of buildings and facades, comprising: Get the building range of the house building; Determine the flight altitude based on the building range and the preset shooting range; Control the preset UAV to fly from the preset reference docking position to the building area and obtain image detection information based on the flight altitude; Determine the specifications of the building to be inspected based on the image detection information and the preset building features; Determine the inspection route based on the specifications of the building to be inspected, control the preset drone to fly along the inspection route, and obtain inspection image information; A building model is formed based on the inspection image information, the crack location is determined based on the building model and preset crack characteristics, and preset abnormality information is output based on the crack location.

[0007] By adopting the above technical solution, a drone is flown to the building range and photographed at the flying height to obtain the building specifications for inspection. The inspection path is then obtained by inspecting the building specifications to control the flight of the drone to obtain a building model, and the crack characteristics are analyzed to obtain the crack location and output abnormal information. In this way, drones are used instead of personnel to inspect buildings, thereby improving the safety and efficiency of inspections, reducing the probability of incomplete data records during manual inspections, reducing labor costs and time costs, and improving the accuracy of inspection results.

[0008] Optionally, the method for controlling the preset drone to fly along the inspection path includes: Determine a vertical scanning position according to the inspection building specifications, and obtain scanning information according to the vertical scanning position; Determine the position of the protrusion based on the scanning information; Determine the bulge scanning path according to the bulge position and the inspection path, and obtain the bulge scanning information; Update the building model based on the convex scan information; Determine the marked raised position according to the updated building model and the preset raised features; The preset prompt information is output according to the raised position of the mark.

[0009] By adopting the above technical solution, the drone is controlled to scan the vertical scanning position vertically to obtain the raised position, and then the raised position is scanned through the raised scanning path to obtain the marked raised position to output prompt information, so that the hollowing condition of the facade of the building can be analyzed by the drone, thereby improving the accuracy of the inspection results.

[0010] Optionally, also include: When the preset raised features do not appear in the updated building model, obtaining solar radiation information and current temperature at the location of the building; Determine the detection range and exposure time based on solar radiation information and the updated building model; Acquire thermal image information according to the detection range; Determine the base thermal color based on solar radiation information, radiation time, preset detection time point and current temperature; The abnormal color position is determined based on the thermal image information and the reference thermal color, and a preset prompt message is output.

[0011] By adopting the above technical solution, when no protruding features appear on the building model, the detection range and exposure time are obtained through the solar radiation information, the current temperature and the building model. The reference thermal color is then determined through the solar radiation information, the exposure time and the manufacturing specifications. Based on the reference thermal color, the abnormal color position is selected from the thermal image information to output prompt information, thereby identifying the non-protruding hollowing conditions on the facade of the building and improving the accuracy of the inspection results.

[0012] Optionally, the method after determining the abnormal color position includes: Determine the irradiation temperature based on the current temperature, irradiation time and solar irradiation information; To obtain the thermal conductivity according to the manufacturing specifications; Determine the heat consumption time based on the current air temperature, irradiation temperature and preset heat conductivity coefficient; Determine the heat dissipation range based on the detection range and solar radiation information; Obtain the detection emission time based on the heat emission range; When the detection emission time is consistent with the heat consumption time, the thermal image information is updated according to the heat emission range; The abnormal color position is updated according to the updated thermal image information and the reference thermal color.

[0013] By adopting the above technical solution, the current temperature, exposure time, solar exposure information and manufacturing specifications are analyzed to obtain the heat consumption time, and then the detection and distribution time of the heat distribution range is obtained by the detection range and solar exposure information. When the detection and distribution time is consistent with the heat consumption time, new thermal image information and new abnormal color positions are obtained, thereby reducing the probability of deviation in the inspection results due to the temperature in the hollow and the temperature of the facade being the same.

[0014] Optionally, after determining the detection range, the method further includes: Determine the remaining detection range based on the detection range and the updated building model; Determine the marked crack position based on the remaining detection range and the crack position; Determine crack prediction parameters based on the marked crack positions and image detection information; Determine the baseline air pressure value and blowing position based on the crack prediction parameters; Controlling the air pressure detection device preset on the drone to blow air at the blowing position with the reference air pressure value and obtain the detection air pressure value; When the detected air pressure value is less than the reference air pressure value, the marked crack position of the detected air pressure value less than the reference air pressure value is defined as the target crack position, and preset prompt information is output according to the target crack position.

[0015] By adopting the above technical solution, the drone is controlled to blow air on the cracks and obtain the detection air pressure value. When the detection air pressure value is inconsistent with the reference air pressure value, prompt information is output according to the target crack position, so that the hollowness of the cracks on the facade can be detected, thereby improving the accuracy of the inspection results.

[0016] Optionally, also include: When the remaining detection range does not include the crack position, the knocking path and the initial detection position are determined according to the remaining detection range and the updated building model; According to the knocking path and the preset reference suction force, the preset detection device is controlled to be adsorbed at the initial detection position and moved; Controlling the air pressure extension device between the tires preset on the detection device to perform interval tapping according to the preset inflation pressure and obtaining sound detection information; Updating sound detection information according to preset noise characteristics; When the updated sound detection information is consistent with the preset reference hollow drum sound, the hollow drum position is determined according to the tapping path, the updated sound detection information and the reference hollow drum sound, and the preset prompt information is output according to the hollow drum position.

[0017] By adopting the above technical solution, the detection device is controlled to detect the remaining detection range, and the detection device is controlled to perform interval tapping, and sound detection information is obtained, and the sound detection information is updated according to the noise characteristics. When the updated sound detection information is consistent with the reference hollow drum sound, prompt information is output according to the hollow drum position, so that the hollow drum can be detected by sound, thereby improving the accuracy of the inspection results.

[0018] Optionally, the method for determining the inflation pressure includes: Determine the specifications of the knocked wall according to the knocking path and the preset manufacturing specifications; Determine the knocking force according to the knocking wall specifications and the knocking specifications of the knocking device preset on the detection device; Determine the extension distance based on the force of the tap; Determine the inflation pressure based on the extended distance; The interval detection time is determined according to the knocking force and the preset moving speed, and the air pressure extension device is controlled to inflate and deflate according to the interval detection time and the inflation pressure.

[0019] By adopting the above technical solution, the knocking path, manufacturing specifications and knocking specifications are used to obtain the knocking force and inflation pressure, and then the knocking force and moving speed are used to obtain the interval detection time, and the air pressure extension device is controlled to inflate and deflate according to the interval detection time and inflation pressure, so that hollowness on the facade can be detected by sound.

[0020] Optionally, after determining the hollowing position, the method further includes: Determine the tapping interval based on the interval detection time and movement speed; Determine the location of the sound source based on the knock path and knock interval; The sound source position where no sound detection information appears is defined as a concave position; Determine an offset distance based on the depression position and a preset tire position, and control the drone to move the detection device by the offset distance; Obtain pressure detection information on the gas extension device; Calculating a force deviation value based on the pressure detection information and a preset baseline pressure value, and controlling the drone to move the detection device to a recessed position by an offset distance; The extension length is determined according to the force deviation value, and the knocking device preset on the detection device is controlled to extend, and the sound detection information is re-acquired.

[0021] By adopting the above technical solution, the sound source position and sound detection information are analyzed to obtain the offset distance and control the movement of the detection device. The pressure detection information and the reference pressure value are used to obtain the extension length, and the knocking device is controlled to extend to the extension length, so that the depressions on the facade can be knocked, and the depressions and depressions caused by hollowing can be analyzed, thereby improving the accuracy of the inspection results.

[0022] In a second aspect, the present application provides a drone AI detection system for hollowing of buildings and facades, which adopts the following technical solutions: A drone AI detection system for hollowing of buildings and facades, including: An acquisition module is used to obtain building range, image detection information, inspection image information, scanning information, convex scanning information, solar radiation information, current temperature, thermal image information, detection emission time, detection pressure value, sound detection information, and pressure detection information; A memory for storing a drone AI detection method for hollowing of buildings and facades; The processor is configured to load, execute, and implement the program stored in the memory.

[0023] In a third aspect, the present application provides a terminal that adopts the following technical solution: A terminal includes a memory and a processor, wherein the memory stores a method for detecting hollow walls of buildings and facades using drones using AI technology that can be loaded and executed by the processor.

[0024] In summary, this application includes at least one of the following beneficial technical effects: 1. Use drones to inspect houses, identify cracks, and output abnormal information. This allows drones to replace human inspections of buildings, improving safety and efficiency, reducing the probability of incomplete data records during manual inspections, reducing labor and time costs, and improving the accuracy of inspection results. 2. The heat consumption time is obtained by analyzing the current temperature, exposure time, solar radiation information, and manufacturing specifications. The detection and dissipation time of the heat dissipation range is then obtained by combining the detection range and solar radiation information. When the detection and dissipation time are consistent with the heat consumption time, new thermal image information and new abnormal color positions are obtained, thereby reducing the probability of the temperature in the hollow being the same as the temperature of the facade, which may cause deviations in the inspection results. 3. By analyzing the sound source position and sound detection information to obtain the offset distance and control the movement of the detection device, and then using the pressure detection information and the reference pressure value to obtain the extension length, and controlling the knocking device to extend to the extension length, it is possible to knock on the depressions on the facade, analyze the depressions and depressions caused by hollowing, and improve the accuracy of the inspection results. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 4. This is a flowchart of a method for detecting hollowing of buildings and facades using a drone AI in accordance with an embodiment of the present invention. Figure 2 This is the method flow of controlling a preset drone to fly along an inspection path according to an embodiment of the present invention. Figure 1 ; Figure 3 This is the method flow of controlling a preset drone to fly along an inspection path according to an embodiment of the present invention. Figure 2 ; Figure 4 is a flow chart of a method after determining the position of an abnormal color according to an embodiment of the present invention; Figure 5 This is the method flow after determining the detection range in the embodiment of the present invention Figure 1 ; Figure 6 This is the method flow after determining the detection range in the embodiment of the present invention Figure 2 ; Figure 7 is a flow chart of a method for determining inflation pressure according to an embodiment of the present invention; Figure 8 4 is a flow chart of a method after determining the hollowing position according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0027] The AI drone detection method for hollowing out buildings and facades controls drones to inspect buildings, and further analyzes the sunlight conditions and the detection results of the detection device. This allows drones to replace personnel in inspecting buildings, improving safety and efficiency during inspections, reducing the probability of incomplete data records during manual inspections, reducing labor costs and time costs, and improving the accuracy of inspection results.

[0028] Reference Figure 1 The present application discloses an AI-powered drone method for detecting hollows in buildings and facades, comprising the following steps: Step S100: Acquire the building range of the building.

[0029] The building range refers to the range of the building that the drone needs to inspect, which can be obtained through pre-input by the operator.

[0030] Step S101: Determine the flight altitude based on the building range and the preset shooting range.

[0031] The capture range is the maximum range set by the drone's camera. The flight altitude is the minimum height at which the drone can capture the entire building area. The flight altitude is calculated by analyzing the building area and the capture range. The higher the flight altitude, the larger the area captured by the drone's camera. The analysis method for the image capture range is well known to those skilled in the art and will not be elaborated here.

[0032] Step S102: Control the preset UAV to fly from the preset reference docking position to the building range, and obtain image detection information according to the flight altitude.

[0033] The baseline docking position is the initial docking location set by the technician. The drone is equipped with an air pressure detection device and a robotic arm for blowing air. The air pressure detection device includes pipes and an air pump. Image detection information is captured by the drone's camera at flight altitude within the building area.

[0034] Step S103: Determine the specifications of the detected building according to the image detection information and the preset building features.

[0035] Architectural features are the shape characteristics of a building during construction, as determined by technicians. Detected building specifications refer to the shape and dimensions of a building during construction. These specifications are determined by identifying the corresponding shapes and dimensions of architectural features from image detection information. Image recognition technology is well-known to those skilled in the art and will not be elaborated upon here.

[0036] Step S104: Determine an inspection path according to the specifications of the building to be inspected, control a preset UAV to fly along the inspection path, and obtain inspection image information.

[0037] The inspection path is the route the drone takes to inspect a building. It follows a spiral flight path around the building, from bottom to top. The inspection path is mapped to a pre-set inspection database by inspecting the building's specifications. The database contains the mapping between inspected building specifications and inspection paths. The inspection database is manually configured and will not be detailed here.

[0038] Step S105: forming a building model based on the inspection image information, determining the crack location based on the building model and preset crack characteristics, and outputting preset abnormality information based on the crack location.

[0039] The facade refers to the visible wall surface of a building. Crack characteristics are the shape and size of cracks on the facade, as defined by the technician. Abnormal information is information set by the technician to alert the operator of cracks on the facade.

[0040] The architectural model is a three-dimensional model formed by the virtual parameters of a building. It is obtained by analyzing inspection image information. The analysis methods for architectural models are common knowledge among those skilled in the art and will not be detailed here. Crack location refers to the location of cracks on the facade of a building. The crack location is determined by identifying the location corresponding to the crack characteristics in the architectural model. The crack location and abnormality information are then output to the terminal held by the operator.

[0041] Reference Figure 2 ,The methods for controlling the preset UAV to fly along the inspection path include: Step S200: determining a vertical scanning position according to the specifications of the inspected building, and acquiring scanning information according to the vertical scanning position.

[0042] The vertical scanning position refers to the top of a building's facade. The vertical scanning position is calculated by retrieving the building's facade's base height from the building specifications and calculating the ground level and height. Scanning information refers to the vertical dimension of the facade scanned by the drone, using vertical distance parameters detected by the drone's pre-installed infrared device.

[0043] Step S201: determining the protrusion position according to the scanning information.

[0044] The raised position refers to the position where there is a hollow on the facade. The vertical distance is retrieved from the scanning information, and the position where the vertical distance is less than the reference height is used as the raised position.

[0045] Step S202: determining a protrusion scanning path according to the protrusion position and the inspection path, and obtaining protrusion scanning information.

[0046] The "raised scanning path" refers to the inspection path used by the drone to scan raised locations. This path uses the raised exterior surface at the raised location as a basis, retrieves a path from the inspection path that can scan the raised exterior surface, and then calculates the distance between the raised location and each retrieved path. The path with the smallest distance perpendicular to the raised location is used as the raised scanning path. The "raised scanning information" refers to the path taken by the drone to scan the raised location along the raised scanning path. When the drone performs a second inspection along the inspection path and is located on the raised scanning path, the distance parameter obtained by controlling the infrared device to scan the raised location is used as the raised scanning information.

[0047] Step S203: updating the building model according to the convex scanning information.

[0048] A new building model is obtained by analyzing the protrusion scanning information. In this embodiment, if a hollow protrusion appears when the drone scans the protrusion position, the detected distance value will show a trend of changing in the form of an arc.

[0049] Step S204: determining the marked raised position according to the updated building model and the preset raised features.

[0050] Raised features are the dimensional features of the raised areas that appear on the building model, as defined by technicians. Marking a raised location refers to the location where the raised area appears on the facade. The location corresponding to the raised feature is identified in the updated building model as the marked raised location.

[0051] The warning message is set by the technician to inform the operator when the facade is tilted. If the building model does not have a raised feature, it indicates that the raised facade is tilted, and the warning message is output to the operator's terminal.

[0052] Step S205: outputting preset prompt information according to the marked raised position.

[0053] The prompt message is set by the technician to inform the operator of the presence of hollowing on the facade. The position of the raised mark and the prompt message are output to the terminal held by the operator.

[0054] Reference Figure 3 , the method for controlling the preset UAV to fly along the inspection path also includes: Step S300: When the preset raised feature does not appear in the updated building model, the solar radiation information and the current temperature of the location of the building are obtained.

[0055] Solar radiation information refers to the altitude, azimuth, and irradiance of the sun at the building's location. This information is obtained by monitoring the sun at the building's location using a sun tracker and a pyranometer. Current temperature refers to the atmospheric temperature at the building's location. The system uses the temperature value at the meteorological observatory at the building's location to obtain the current temperature.

[0056] Step S301: Determine the detection range and exposure time according to the solar exposure information and the updated building model.

[0057] The detection range refers to the area of a building exposed to sunlight. This range is determined by analyzing the solar radiation information and the updated building model. The analysis method for the detection range is well-known to those skilled in the art and will not be detailed here. The exposure time refers to the maximum duration that the detection range is exposed to sunlight. The detection range is measured and the resulting time is used as the exposure time.

[0058] Step S302: Acquire thermal image information according to the detection range.

[0059] Thermal image information refers to the thermal image of the facade of the building within the detection range, and the image obtained by detecting the detection range with a thermal imager preset on the drone is used as the thermal image information.

[0060] Step S303: determining a reference thermal color according to solar radiation information, radiation time, a preset detection time point, and current temperature.

[0061] The inspection time point is the time point set by the technician for the drone to perform thermal imaging inspection on the facade. In this embodiment, the system will dispatch drones with different inspection functions to perform inspection operations. When performing thermal imaging inspection on the facade, multiple drones will be dispatched for inspection.

[0062] The baseline thermal color refers to the color of the facade in the thermal image at the time of detection. This color is determined by analyzing the solar radiation, exposure time, and current temperature through heat conduction analysis. The color corresponding to the detection time is then used as the baseline thermal color. Heat conduction analysis methods are well known to those skilled in the art and will not be detailed here.

[0063] Step S304: determining the abnormal color position according to the thermal image information and the reference thermal color, and outputting a preset prompt message.

[0064] The abnormal color position refers to the position where a color that is inconsistent with the reference thermographic color appears in the thermographic image information. The position of the color that is inconsistent with the reference thermographic color is identified as the abnormal color position from the thermographic image information, and a prompt message is output to the operator's handheld terminal.

[0065] Reference Figure 4 , the method after determining the abnormal color location includes: Step S400: Determine the irradiation temperature according to the current temperature, irradiation time and solar irradiation information.

[0066] Manufacturing specifications are the materials, manufacturing processes, and specific heat capacity specifications for building construction, as determined by technical personnel. Irradiation temperature refers to the temperature of the facade after the irradiation period. This temperature is determined by analyzing the manufacturing specifications for heat conduction based on the current temperature, irradiation time, and solar radiation information. Heat conduction analysis methods are well-known to those skilled in the art and will not be detailed here.

[0067] Step S401: determining heat consumption time according to the current air temperature, the irradiation temperature and a preset heat conductivity coefficient.

[0068] The thermal conductivity coefficient is the coefficient of heat conduction for the facade, as set by technicians. Heat dissipation time refers to the length of time it takes for the facade to reach its current temperature from the irradiated temperature. This is calculated by calculating the temperature difference between the current temperature and the irradiated temperature. This temperature difference and the thermal conductivity coefficient are then input into a preset consumption database to match the heat dissipation time. The consumption database contains the corresponding relationship between temperature difference, thermal conductivity, and heat dissipation time, which will not be detailed here.

[0069] Step S402: Determine the heat dissipation range according to the detection range and solar radiation information.

[0070] The heat dissipation range refers to the range within which heat begins to be dissipated on the facade per unit time. The detection range is updated by referring to step S301 for solar radiation information at different time points, and the range corresponding to the difference between the detection ranges before and after the update is used as the heat dissipation range.

[0071] Step S403: Obtaining the heat dissipation detection time according to the heat dissipation range.

[0072] The detection emission time refers to the time length from the heat emission range starting to emit heat. When the heat emission range appears, the timing starts and the timing result is used as the detection emission time.

[0073] Step S404: When the heat emission time is detected to be consistent with the heat consumption time, the thermal image information is updated according to the heat emission range.

[0074] When the detection emission time is consistent with the heat consumption time, it means that the facade can be thermally imaged and the thermal image information of the heat emission range is reacquired.

[0075] Step S405: updating the abnormal color position according to the updated thermographic image information and the reference thermographic color.

[0076] The updated thermal image information is referred to step S304 to obtain a new abnormal color position.

[0077] Reference Figure 5 , the method after determining the detection range also includes: Step S500: determining a remaining detection range according to the detection range and the updated building model.

[0078] The remaining detection range refers to the range on the facade that is not exposed to the sun. The range that does not correspond to the detection range is retrieved from the facade of the updated building model as the remaining detection range.

[0079] Step S501: Determine and mark the crack position according to the remaining detection range and the crack position.

[0080] The marked crack position refers to the crack position within the remaining detection range, and the crack position falling within the remaining detection range is retrieved from the crack positions as the marked crack position.

[0081] Step S502: determining crack prediction parameters based on the marked crack positions and image detection information.

[0082] Crack prediction parameters refer to the estimated shape and size parameters of the crack. The shape and size parameters of the cracks marked at the crack locations are identified from the image detection information and used as the crack prediction parameters. The image recognition method is well known to those skilled in the art and will not be described in detail here.

[0083] Step S503: determining a reference air pressure value and an air blowing position according to the crack prediction parameters.

[0084] In this embodiment, when hollowing occurs on the facade, cracks may form. These cracks penetrate the facade and the hollow, allowing air to circulate with the outside world. The reference pressure value refers to the pressure value when the drone's air blowing device blows air into a crack that is not caused by hollowing. This reference pressure value is matched from a preset air blowing database using crack prediction parameters. The air blowing position refers to the center of the crack. The center of the crack shape is determined by analyzing the crack prediction parameters and serves as the air blowing position.

[0085] The air blowing database contains the correspondence between the crack prediction parameters and the reference air pressure values. The air blowing database is set manually and will not be described in detail here.

[0086] Step S504: controlling the air pressure detection device preset on the drone to blow air at the blowing position at the reference air pressure value, and obtaining the detected air pressure value.

[0087] The detected air pressure value refers to the air pressure value detected by the air pressure detection device when blowing air into the crack. When the air pressure detection device is controlled to blow air into the blowing position at the reference air pressure value, the parameter obtained by the pressure sensor preset on the air pressure detection device is used as the detected air pressure value.

[0088] Step S505: When the detected air pressure value is less than the reference air pressure value, the marked crack position of the detected air pressure value less than the reference air pressure value is defined as a target crack position, and preset prompt information is output according to the target crack position.

[0089] The target crack position refers to the position of the crack on the facade caused by hollowing. When the detected air pressure value is lower than the reference air pressure value, it indicates that the air in the hollow is circulating with the outside world. Therefore, the crack position marked with a detected air pressure value lower than the reference air pressure value is defined as the target crack position, and the target crack position and prompt information are output to the terminal held by the operator.

[0090] Reference Figure 6 , the method after determining the detection range also includes: Step S600 : When the remaining detection range does not include the crack position, determining the knocking path and the initial detection position according to the remaining detection range and the updated building model.

[0091] The initial detection position refers to the highest point farthest from the ground in the remaining detection range. When the remaining detection range does not include the crack position, it means that the hollowing on the facade is not obvious. The highest point farthest from the ground in the remaining detection range is taken as the initial detection position.

[0092] Step S601: controlling a preset detection device to be adsorbed at an initial detection position and move according to the knocking path and a preset reference suction force.

[0093] The detection device refers to a device used to detect inconspicuous hollows on the facade. The detection device includes a main body, a suction cup provided on the main body and used for the main body to be adsorbed on the facade, tires evenly provided on the main body and used to drive the main body to move on the facade, and a knocking device provided on the main body and used to knock on the facade. The knocking device is a hollow telescopic rod. A hollow telescopic rod is installed between the suction cup and the main body, and a marking telescopic rod is provided between the tire and the main body. The marking telescopic rod includes a hollow fixed rod and a driving rod. The fixed rod is provided on the main body, and the driving rod is provided at the end of the fixed rod away from the main body. A spring is provided between the tire and the fixed rod.

[0094] A plurality of air pumps are provided in the main body, one air pump provides suction for the suction cup, one air pump is used to control the movement of the main body in the normal direction away from the facade, and one air pump is used to control the extension of the knocking device.

[0095] The reference suction force is the force set by the technician for the suction cup to perform adsorption. The detection device is controlled to adsorb at the initial detection position with the reference suction force and move along the tapping path.

[0096] Step S602: controlling the air pressure extension device between the tires preset on the detection device to perform interval tapping according to the preset inflation pressure, and obtaining sound detection information.

[0097] The air pressure extension device is a fixed rod and an air pump. The inflation pressure refers to the air pressure that controls the air pump connected to the fixed rod to inflate the air. The inflation pressure controls the air pressure extension device between the tires preset on the detection device to perform interval tapping. The sound detection information is the sound parameters collected by the sound collection device preset on the main body. The sound collection device is a microphone. The reference continuous air pressure is the air pressure set by the technician to drive the air pressure extension device to continuously adhere to the outer surface. In this embodiment, when the detection device is adsorbed on the external force surface, the air pressure extension device will output the reference continuous air pressure to keep the tire in continuous contact with the outer surface.

[0098] Step S603: Update the sound detection information according to the preset noise characteristics.

[0099] The noise feature is a feature set by the technician that is not the timbre of the sound of the knocking device knocking on the facade. New sound detection information is formed by removing the noise feature from the sound detection information.

[0100] Step S604: When the updated sound detection information is consistent with the preset reference hollow drum sound, the hollow drum position is determined according to the striking path, the updated sound detection information and the reference hollow drum sound, and the preset prompt information is output according to the hollow drum position.

[0101] The baseline hollowing sound is the timbre of the sound produced by a percussion device set by a technician striking a facade with a hollow. The hollowing location refers to the location on the facade where the hollowing occurs. When the updated sound detection information matches the baseline hollowing sound, it indicates that a hollow has occurred on the facade. The location where the updated sound detection information that matches the baseline hollowing sound appears along the percussion path is selected as the hollowing location, and the hollowing location and prompt information are output to the terminal held by the operator.

[0102] Reference Figure 7 , the method for determining the inflation pressure includes: Step S700: Determine specifications for knocking the wall according to the knocking path and preset manufacturing specifications.

[0103] The specifications for knocking walls refer to the specifications that need to be knocked, and the specifications of the facade included in the knocking path are retrieved from the manufacturing specifications as the specifications for knocking walls.

[0104] Step S701: Determine the knocking force according to the knocking wall specifications and the knocking specifications of the knocking device preset on the detection device.

[0105] The striking specifications are the material specifications of the striking device set by the technician. The striking force refers to the force required to strike the facade. The striking force is determined from a pre-set striking database by matching the striking wall specifications with the striking specifications. The striking database contains the correspondence between striking wall specifications, striking specifications, and striking force. The striking database is manually set and will not be detailed here.

[0106] Step S702: Determine the extension distance according to the tapping force.

[0107] The elastic coefficient is the spring's coefficient value set by a technician. The extension distance refers to the distance the spring is required to extend. This distance is calculated by analyzing the impact force and the preset elastic coefficient. The analysis method for the extension distance is common knowledge among those skilled in the art and will not be detailed here.

[0108] Step S703: Determine the inflation pressure according to the extension distance, and control the air pressure extension device between the tires preset on the detection device to operate at the inflation pressure.

[0109] The inflation pressure is the air pressure required by the gas extension device to extend the spring. This pressure is matched to the extension distance from a preset inflation database, and the gas extension device is controlled to operate at this inflation pressure. The inflation database contains the correspondence between extension distance and inflation pressure. The inflation database is manually set and will not be detailed here.

[0110] Step S704: Determine the interval detection time according to the tapping force and the preset movement speed, and control the air pressure extension device to inflate and deflate according to the interval detection time and the inflation pressure.

[0111] The movement speed is the speed at which the detection device moves, as set by the technician. The interval detection time refers to the duration of the interval detections performed by the tapping device. The diffusion distance refers to the range over which the vibration generated by the tapping is diffused. The diffusion distance is matched to the tapping force in the tapping database, and the quotient of the diffusion distance and the movement speed is calculated as the interval detection time. The tapping database also contains the corresponding relationship between tapping force and diffusion distance, which will not be detailed here.

[0112] Reference Figure 8 , the method after determining the hollowing position also includes: Step S800: Determine the tapping interval according to the interval detection time and the movement speed.

[0113] The striking interval refers to the distance that the striking device moves when performing interval striking, and the striking interval is the diffusion distance.

[0114] Step S801: Determine the location of the sound source according to the tapping path and the tapping distance.

[0115] The sound source position refers to the position point where the knocking device knocks on the knocking path. The knocking path is separated by knocking intervals, and each interval position point is used as the sound source position.

[0116] Step S802: defining the sound source position where no sound detection information appears as a recessed position.

[0117] The sunken position refers to a position where a depression occurs within the remaining detection range of the facade, and the sunken position is defined by defining the sound source position where no sound detection information occurs.

[0118] Step S803: determining an offset distance according to the recessed position and a preset tire position, and controlling the UAV to move the detection device by the offset distance.

[0119] The tire position is the point on the vehicle where the tire is installed, as determined by the technician. The offset distance is the horizontal distance between the depression and the tire. This distance is calculated as the offset distance, and the suction cup is deflated. The robotic arm on the drone then moves the gripping detection device by the offset distance.

[0120] Step S804: Obtain pressure detection information on the gas extension device.

[0121] The pressure detection information refers to the air pressure value on the gas extension device, and the parameters obtained by detecting the pressure sensor preset on the driving rod close to the inner wall of the tire are used as the pressure detection information.

[0122] Step S805: Calculate a force deviation value based on the pressure detection information and a preset reference pressure value, and control the drone to move the detection device to a recessed position by an offset distance.

[0123] When the tire is in a recessed position, the baseline pressure value causes the tire to extend, increasing the space between the fixed rod and the driving rod. The more gas contained, the lower the air pressure acting on the pressure sensor. The baseline pressure value is the baseline continuous air pressure in step S602. The force deviation value refers to the deviation between the pressure detection information and the baseline pressure value. The force deviation value is calculated by calculating the difference between the pressure detection information and the baseline pressure value, and referring to step S803, the drone is controlled to move the detection device to the recessed position by an offset distance.

[0124] Step S806: Determine the extension length according to the force deviation value, control the knocking device preset on the detection device to extend, and re-acquire the sound detection information.

[0125] The extension length refers to the length that the knocking device needs to be extended. The knocking device is controlled by the extension length to be inflated and extended, and the knocking is performed again to obtain sound detection information.

[0126] Based on the same inventive concept, an embodiment of the present invention provides a drone AI detection system for hollowing of buildings and facades, comprising: An acquisition module is used to obtain building range, image detection information, inspection image information, scanning information, convex scanning information, solar radiation information, current temperature, thermal image information, detection emission time, detection pressure value, sound detection information, and pressure detection information; A memory for storing a drone AI detection method for hollowing of buildings and facades; The processor is configured to load, execute, and implement the program stored in the memory.

[0127] Based on the same inventive concept, an embodiment of the present invention provides a terminal including a memory and a processor, wherein the memory stores a drone AI detection method for hollowing of buildings and facades that can be loaded and executed by the processor.

[0128] Those skilled in the art will clearly understand that for the sake of convenience and brevity, the division of the above-mentioned functional modules is only used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working processes of the above-mentioned systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0129] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A drone AI detection method for hollowing of buildings and facades, characterized by: include: Get the building range of the house building; Determine the flight altitude based on the building range and the preset shooting range; Control the preset UAV to fly from the preset reference docking position to the building area and obtain image detection information based on the flight altitude; Determine the specifications of the building to be inspected based on the image detection information and the preset building features; Determine the inspection route based on the specifications of the building to be inspected, control the preset drone to fly along the inspection route, and obtain inspection image information; A building model is formed based on the inspection image information, the crack location is determined based on the building model and preset crack characteristics, and preset abnormality information is output based on the crack location.

2. The drone AI detection method for hollowing of buildings and facades according to claim 1 is characterized in that: The methods for controlling the preset drone to fly along the inspection path include: Determine a vertical scanning position according to the inspection building specifications, and obtain scanning information according to the vertical scanning position; Determine the position of the protrusion based on the scanning information; Determine the bulge scanning path according to the bulge position and the inspection path, and obtain the bulge scanning information; Update the building model based on the convex scan information; Determine the marked raised position according to the updated building model and the preset raised features; The preset prompt information is output according to the raised position of the mark.

3. The drone AI detection method for hollowing of buildings and facades according to claim 2 is characterized in that: Also includes: When the preset raised features do not appear in the updated building model, obtaining solar radiation information and current temperature at the location of the building; Determine the detection range and exposure time based on solar radiation information and the updated building model; Acquire thermal image information according to the detection range; Determine the base thermal color based on solar radiation information, radiation time, preset detection time point and current temperature; The abnormal color position is determined based on the thermal image information and the reference thermal color, and a preset prompt message is output.

4. The drone AI detection method for hollowing of buildings and facades according to claim 3 is characterized in that: Methods after determining the location of the abnormal color include: Determine the irradiation temperature based on the current temperature, irradiation time and solar irradiation information; Determine the heat consumption time based on the current air temperature, irradiation temperature and preset heat conductivity coefficient; Determine the heat dissipation range based on the detection range and solar radiation information; Obtain the detection emission time based on the heat emission range; When the detection emission time is consistent with the heat consumption time, the thermal image information is updated according to the heat emission range; The abnormal color position is updated according to the updated thermal image information and the reference thermal color.

5. The drone AI detection method for hollowing of buildings and facades according to claim 3 is characterized in that: The method after determining the detection range also includes: Determine the remaining detection range based on the detection range and the updated building model; Determine the marked crack position based on the remaining detection range and the crack position; Determine crack prediction parameters based on the marked crack positions and image detection information; Determine the baseline air pressure value and blowing position based on the crack prediction parameters; Controlling the air pressure detection device preset on the drone to blow air at the blowing position at the reference air pressure value and obtain the detection air pressure value; When the detected air pressure value is less than the reference air pressure value, the marked crack position of the detected air pressure value less than the reference air pressure value is defined as the target crack position, and preset prompt information is output according to the target crack position.

6. The drone AI detection method for hollowing of buildings and facades according to claim 5 is characterized in that: Also includes: When the remaining detection range does not include the crack position, the knocking path and the initial detection position are determined according to the remaining detection range and the updated building model; According to the knocking path and the preset reference suction force, the preset detection device is controlled to be adsorbed at the initial detection position and moved; Controlling the air pressure extension device between the tires preset on the detection device to perform interval tapping according to the preset inflation pressure and obtaining sound detection information; Updating sound detection information according to preset noise characteristics; When the updated sound detection information is consistent with the preset reference hollow drum sound, the hollow drum position is determined according to the tapping path, the updated sound detection information and the reference hollow drum sound, and the preset prompt information is output according to the hollow drum position.

7. The drone AI detection method for hollowing of buildings and facades according to claim 6 is characterized in that: Methods for determining inflation pressure include: Determine the specifications of the knocked wall according to the knocking path and the preset manufacturing specifications; Determine the knocking force according to the knocking wall specifications and the knocking specifications of the knocking device preset on the detection device; Determine the extension distance based on the force of the tap; Determine the inflation pressure based on the extended distance; The interval detection time is determined according to the knocking force and the preset moving speed, and the air pressure extension device is controlled to inflate and deflate according to the interval detection time and the inflation pressure.

8. The drone AI detection method for hollowing of buildings and facades according to claim 7 is characterized in that: The method after determining the location of the hollow drum also includes: Determine the tapping interval based on the interval detection time and movement speed; Determine the location of the sound source based on the knock path and knock interval; The sound source position where no sound detection information appears is defined as a concave position; Determine an offset distance based on the depression position and a preset tire position, and control the drone to move the detection device by the offset distance; Obtain pressure detection information on the gas extension device; Calculating a force deviation value based on the pressure detection information and a preset baseline pressure value, and controlling the drone to move the detection device to a recessed position by an offset distance; The extension length is determined according to the force deviation value, and the knocking device preset on the detection device is controlled to extend, and the sound detection information is re-acquired.

9. A drone AI detection system for hollowing of buildings and facades, characterized by: include: An acquisition module is used to obtain building range, image detection information, inspection image information, scanning information, convex scanning information, solar radiation information, current temperature, thermal image information, detection emission time, detection pressure value, sound detection information, and pressure detection information; A memory for storing the drone AI detection method for hollowing of buildings and facades according to any one of claims 1 to 8; The processor is configured to load, execute, and implement the program stored in the memory.

10. A terminal, characterized in that: The invention comprises a memory and a processor, wherein the memory stores a method for detecting hollow walls of buildings and facades by a drone that can be loaded and executed by the processor as described in any one of claims 1 to 8.

Images