Moving body, processing device, system, method, and program for detecting uplift of mortar or tile of outer wall

Abstract

To provide a moving body, a processing device, a system, a method, and a program for detecting uplift of mortar or tiles on an outer wall of a building.SOLUTION: A moving body for detecting uplift of mortar or tiles on an outer wall of a building according to the present invention includes an irradiation unit that irradiates the outer wall with a millimeter wave, a reception unit that receives the millimeter wave reflected from the outer wall, and a recording unit that records an intensity of a signal of the received millimeter wave.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a mobile object, a processing device, a system, a method, and a program for detecting loose mortar or loose tiles on exterior walls. In particular, the present invention aims to detect loose mortar and tile on the exterior of a building, where the exterior is finished by applying mortar to the foundation wall, or by applying mortar to the foundation wall and then attaching tiles, in a non-contact and non-destructive manner by irradiating and receiving millimeter waves. [Background technology]

[0002] Inspecting exterior wall tiles for peeling or loosening, as well as loose mortar, is extremely important for ensuring the safety of buildings, and in Japan, regular inspections are required by law. The details of these inspections are explained below.

[0003] (Regular reporting system) Based on Ministry of Land, Infrastructure, Transport and Tourism Notification No. 282, the Building Standards Act Enforcement Regulations have been revised to require periodic inspections of exterior wall tiles for peeling, loosening, and other deterioration. This notification came into effect in April 2008, and requires the following inspections for specific buildings: (1) Survey frequency: - Surveys by percussion within reach should be conducted approximately once every six months to three years. - For parts that may pose a risk to pedestrians and others if they fall, a full inspection including percussion should be carried out once every 10 years. (2) Survey Methodology: - Test hammers should be used as the basis for sounding, and infrared inspections or visual inspections using binoculars should be carried out as necessary. - The technical advice issued in May 2018 also allows for confirmation by tensile adhesion tests in the case of organic adhesive installation methods. (3) Survey items: - Check the deterioration and damage of tiles, stonework, mortar, etc.

[0004] (Specified building periodic report) Specified buildings are required to submit periodic reports every three years. These periodic reports include an investigation into the deterioration of the exterior finishing materials on the building's exterior walls, and the building owner or manager must report the results of the investigation to the specified administrative agency, conducted by a qualified inspector.

[0005] (Importance of research) These regular exterior wall inspections are essential for early detection of peeling or loose tiles and preventing falling accidents. Strict management is required to ensure safety, especially in buildings facing public areas.

[0006] Inspecting exterior building materials for signs of peeling and other damage is extremely important for maintaining the safety and longevity of buildings. If peeled building materials fall, they can pose a danger to pedestrians and vehicles. Furthermore, as peeling progresses, the structural integrity of the building decreases, increasing the risk of collapse. Furthermore, if water seeps in through peeling or cracks, the interior of the building becomes more susceptible to moisture and mold, and the resulting deterioration of waterproofing performance can cause damage to internal assets and equipment.

[0007] Damage to the exterior of a building reduces its value, and poor appearance reduces the satisfaction of residents and users. On the other hand, early repairs can reduce the cost of major repairs and renovations, and regular inspections can detect problems early, reducing long-term maintenance costs.

[0008] (Traditional method) Conventional methods include the following.

[0009] - Visual inspection: The surface of the exterior wall is visually checked to check for cracks, peeling, discoloration, and other abnormalities. Binoculars and drones are used to inspect high places and hard-to-reach areas in detail, but loose tiles (peeling from the mortar) are difficult to see with the naked eye.

[0010] - Percussion inspection: Lightly tap the exterior wall with a hammer or special tool to hear any differences in sound. If there is peeling, a hollow sound will be heard. However, if the area is high and not adjacent to a walkway, scaffolding must be erected, which is costly and time-consuming. In addition, working at a high altitude can be dangerous.

[0011] - Palpation inspection: Check the exterior wall by touching it with your hands to identify any peeling or loosening areas. This has the same issues as the percussion inspection.

[0012] - Infrared thermography: An infrared camera is used to check the temperature distribution of exterior walls. Temperature anomalies can be detected if there is peeling or internal abnormalities. Mounted on a drone, it eliminates the need for scaffolding, and measurement takes a short time, is safe, and is inexpensive, making it popular. However, the infrared method (thermography), which has traditionally been used as a non-destructive inspection method for buildings, etc., is significantly affected by the season and weather, with fluctuations in solar radiation and temperature particularly affecting the observation results.

[0013] For example, measurements are highly accurate on days with large diurnal temperature fluctuations or during the daytime temperature rise, but accuracy declines at night or when temperature fluctuations are small. Furthermore, measuring north-facing surfaces is difficult and may be less accurate than measuring south-, east-, or west-facing surfaces. For north-facing surfaces, a method utilizing heat flow from the inside to the outside during winter nights is recommended. Furthermore, resolution decreases as the distance increases; with a typical camera, tile lift can only be detected within a distance of approximately 30 m. Different color tones also affect infrared absorption rates, which can affect measurement results. Tiles with particularly high reflective strength are susceptible to external noise, making measurement difficult if there is an obstacle between the wall and the camera. Results may also vary depending on the performance of the equipment and the image processing method.

[0014] Therefore, there is a need for a system and method for detecting floating wall materials that is less affected by weather, temperature, sunlight conditions, etc., can be mounted on a drone, and does not require the construction of scaffolding.

[0015] As for the trend in exterior wall deterioration diagnosis technology using infrared methods, with current technology, when the temperature of a building's exterior wall changes due to changes in sunlight or outside temperature, the rate of temperature change differs between the floating parts (where there is an air layer) and the parts that are in close contact with the substrate. Generally, the temperature changes faster in floating parts. This technology uses thermography to measure the temperature when this difference becomes noticeable, and identifies the floating parts. This can be said to be a passive infrared method, where inspections are dependent on environmental changes.

[0016] For example, Patent Document 1 discloses "detecting areas of mortar that have lifted by measuring temperature distribution." In other words, Patent Document 1 discloses the "infrared method," which has traditionally been used as a non-destructive inspection method for buildings and the like. When the temperature of a building's exterior wall changes due to changes in sunlight or outside air temperature, the rate of temperature change differs between the lifted areas (where there is an air layer) and the areas that are in close contact with the base. Generally, the temperature changes more quickly in the lifted areas. This technology uses thermography to measure the temperature when this difference becomes noticeable, and identifies the lifted areas. This can be said to be a passive infrared method in which inspection is dependent on environmental changes.

[0017] However, Patent Document 1 detects the difference in temperature change over time between the floating and non-floating parts, and has the problem that the measurement is affected by the time of day, climate, and sunlight conditions.

[0018] Furthermore, for example, Patent Document 2 discloses a method for detecting lifted portions of an exterior wall by effectively utilizing the temperature difference that occurs between lifted portions and normal portions of the exterior, and measuring the temperature distribution on the wall surface with an infrared temperature measuring device that can obtain thermal images or thermal photographs. In other words, Patent Document 2 also discloses the "infrared method," which has traditionally been used as a non-destructive inspection method for buildings and the like.

[0019] However, Patent Document 2 also detects the difference in temperature change over time between the floating portion and the non-floating portion, and does not solve the problem of Patent Document 1. [Prior art documents] [Patent documents]

[0020] [Patent Document 1] Japanese Patent Application Publication No. 57-189007 [Patent Document 2] Japanese Patent Application Publication No. 59-106669 Summary of the Invention [Problem to be solved by the invention]

[0021] It provides an active inspection method that is not dependent on the environment. [Means for solving the problem]

[0022] We irradiate the exterior wall with millimeter waves and measure the intensity of the reflection to actively detect any loosening. Millimeter waves pass through exterior tiles and mortar. Normally, there is little reflection from the boundary between the closely attached foundation wall and mortar, or between the mortar and tiles, and most of the reflection comes from the outermost surface that is in contact with the air. On the other hand, if the foundation wall and mortar are peeling away, or if the tiles are loose from the mortar, reflection increases when surfaces that are not normally in contact with the air (foundation wall, mortar, backside of tile) come into contact with the air. What should normally be reflected only from the outermost surface of the exterior wall, if there is space (air) inside, reflection from the surface in contact with the air increases. This makes it possible to determine whether there is any loosening inside.

[0023] In order to solve the above problem, the present invention provides a mobile body for detecting loose mortar or tiles on the exterior wall of a building, the mobile body comprising an irradiation unit that irradiates the exterior wall with millimeter waves, a receiving unit that receives millimeter waves reflected from the exterior wall, and a recording unit that records the intensity of the received millimeter wave signal.

[0024] The moving body according to an aspect of the present invention further includes an imaging unit that acquires an image of the exterior wall.

[0025] The moving body according to an aspect of the present invention further includes a distance measurement unit that measures the distance from the moving body to the tile.

[0026] In a mobile body according to an embodiment of the present invention, the frequency of the millimeter wave may be, for example, 24 GHz.

[0027] In a mobile body according to an embodiment of the present invention, the millimeter wave may be an electromagnetic wave that transmits 5% or more through, for example, a 10 mm thick tile or mortar.

[0028] The mobile body according to an embodiment of the present invention may be a drone, a mobile body having wheels or caterpillar tracks, or a gondola suspended by a wire.

[0029] The present invention also provides a processing device for detecting loose mortar or tiles on the exterior walls of a building, which includes an acquisition unit that acquires the intensity of a millimeter wave signal reflected from the exterior wall, and a looseness detection unit that detects loose tiles based on the intensity, the intensity being observed by a moving object.

[0030] In a processing device according to one aspect of the present invention, the exterior wall consists of a foundation wall, mortar, and tiles, and the lift detection unit determines that an air layer exists between the tile and mortar, or between the mortar and the foundation wall, when the intensity exceeds a predetermined threshold.

[0031] In a processing device according to one aspect of the present invention, the threshold value may be, for example, one of (a) a value input by a user, (b) a value determined based on measurements of tiles on an exterior wall that are not floating, or (c) a value that divides multiple measured intensities into normal values ​​and abnormal values ​​that are greater than the normal values.

[0032] The processing device according to an aspect of the present invention further includes a correction unit that measures the distance between the moving object and the tile and corrects the intensity.

[0033] A processing device according to an aspect of the present invention further includes a display unit that displays an image in which intensity values ​​are expressed as a heat map superimposed on an image of the tile acquired by the moving body.

[0034] The present invention also provides a system for detecting lifting of mortar or tiles on the exterior wall of a building, the system comprising the above-mentioned mobile body and the above-mentioned processing device.

[0035] The present invention also provides a method for detecting loose mortar or tiles on the exterior wall of a building, the method including the steps of: a mobile body irradiating millimeter waves onto the exterior wall and receiving the millimeter waves reflected from the exterior wall; and a processing device acquiring the intensity of the millimeter wave signal reflected from the exterior wall and received by the mobile body, and the processing device detecting loose tiles based on the acquired intensity.

[0036] The present invention also provides a program for detecting loose mortar or tiles on the exterior walls of a building, which program causes a computer to execute the steps of reading the intensity of a millimeter wave signal reflected from the exterior wall from a recording medium, and detecting loose tiles based on the read intensity.

[0037] In another aspect of the present invention, there is provided a system for detecting loose mortar or tiles on the exterior walls of a building, the system comprising: a data collection unit that acquires the intensity distribution of millimeter waves reflected from the exterior wall and image data of the exterior wall; a data integration unit that integrates the intensity distribution of millimeter waves reflected from the exterior wall with image data of the exterior wall acquired by a color camera; a pre-processing unit that removes noise from the image data; an AI analysis unit that detects loose mortar or tiles on the exterior wall using a machine learning model; and a result display unit that displays the locations where loose mortar or tiles exist.

[0038] The system for detecting loose mortar or tiles on the exterior wall of a building according to another aspect of the present invention may further comprise a report generation unit that automatically generates a report on the location where loose mortar or tiles exist.

[0039] In another aspect of the present invention, there is provided a method for detecting loose mortar or tiles on the exterior walls of a building, the method including the steps of collecting data to generate a machine learning model, collecting data using millimeter waves and a color camera, integrating the data, pre-processing, removing noise and normalizing reflection intensity, an AI analysis step that detects loose mortar or tiles on the exterior walls using the machine learning model, segmenting abnormal areas using a SAM (Segment Anything Model) model and extracting peculiar parts using an anomaly detection algorithm, displaying the results, visualizing the abnormal areas using a user interface, and creating a report.

[0040] In the present invention, "floating" refers to something that was attached to a main body becoming detached from the main body. Specifically, in the present invention, "floating" refers to a state in which something that was attached to a wall becomes detached from the wall, creating a gap between the wall and the object. As a more specific example, in the present invention, "floating" also refers to a state in which a tile that was attached to the surface of a wall peels off, creating an air gap between the wall and the tile. [Effects of the Invention]

[0041] According to the present invention, it is possible to detect loose mortar or tiles on the exterior walls of a building using millimeter waves, and it is possible to detect loose mortar or tiles on the exterior walls of a building with high accuracy, regardless of environmental changes. Other objects, features and advantages of the present invention will become apparent from the following description of the preferred embodiments of the present invention taken in conjunction with the accompanying drawings. [Brief explanation of the drawings]

[0042] [Figure 1] FIG. 1 is a conceptual diagram showing how a moving body detects lifting of mortar or tiles on the exterior wall of a building in the lifting detection system according to the present invention. [Figure 2] FIG. 2 is a diagram showing the entire float detection system according to the present invention. [Figure 3] FIG. 3 is a diagram showing how millimeter waves are irradiated onto an exterior wall and reflected from the exterior wall in the float detection system according to the present invention. [Figure 4] FIG. 4 is a diagram showing the relationship between the moving body and the outer wall according to the present invention. [Figure 5] FIG. 5 is a diagram showing the relationship between the movement of a moving object and the images acquired according to the present invention. [Figure 6] FIG. 6 is a diagram showing a moving object equipped with a plurality of irradiating units according to another embodiment of the present invention. [Figure 7] FIG. 7 is a diagram showing an example in which a millimeter wave imaging radar is mounted on a moving object according to the present invention. [Figure 8] FIG. 8 is a flow chart of a method for detecting floatation according to the present invention. [Figure 9] FIG. 9 is a diagram showing the entirety of a floating detection system according to a second embodiment of the present invention. [Figure 10] FIG. 10 is a flowchart of a method for detecting floating according to a second embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION [Example]

[0043] FIG. 1 is a conceptual diagram showing how a moving body detects lifting of mortar or tiles on the exterior wall of a building in the lifting detection system according to the present invention. The exterior wall of a building that is the subject of the present invention typically consists of a foundation wall 10, mortar 11, and tiles 12, as shown in FIG. 1(a). The tiles 12 are decorative materials for decorating the surface of the foundation wall 10, and mortar 11, which is an adhesive layer for adhering the tiles 12 to the foundation wall 10, is provided between the foundation wall 10 and the tiles 12. In another example, as shown in FIG. 1(b), the exterior wall may not have tiles 12, and the outermost surface of the exterior wall may be mortar 11. Note that although the following description will be given assuming an exterior wall of a building, the present invention may also be applied to an interior wall existing inside the building. For example, the present invention can also be applied to the interior walls of an open-ceiling space within a building or the interior walls of a room with a sufficient ceiling height.

[0044] During construction of an exterior wall, the foundation wall 10 and the mortar 11, and the mortar 11 and the tile 12 are bonded together without any gaps. However, due to aging, damage, or the like, gaps, i.e., air spaces, may form between the foundation wall 10 and the mortar 11, or between the mortar 11 and the tile 12. Such gaps, i.e., air spaces, between the foundation wall 10 and the mortar 11, or between the mortar 11 and the tile 12, are referred to as "floating." In the example of FIG. 1(a), some of the tiles 12 are "floating." In the example of FIG. 1(b), some of the mortar 11 is "floating." The present invention aims to detect "floating" of the mortar 11 or the tile 12 with high accuracy. That is, the present invention detects gaps, i.e., air spaces, between the foundation wall 10 and the mortar 11, or between the mortar 11 and the tile 12.

[0045] The foundation wall 10 is assumed to be a concrete wall, but may be made of other materials, such as wood or gypsum board, as long as the detection method of the present invention is applicable. The mortar 11 may be made of other adhesive materials. The material of the tiles 12 is not particularly limited, and may be any decorative material used in construction, such as brick, block, ceramic, or other materials. The tiles 12 are typically separated into individual tiles before construction, but may also be made of multiple tiles joined together.

[0046] Furthermore, in the present invention, a mobile object 20 capable of emitting and receiving millimeter waves is moved along an exterior wall, millimeter waves are emitted from the mobile object 20 toward the exterior wall, and the mobile object 20 receives and records the intensity of the millimeter wave signal reflected from the exterior wall. The lifting of the mortar 11 or tiles 12 on the exterior wall is detected based on the difference in the intensity of the millimeter wave signal reflected from the exterior wall.

[0047] FIG. 2 is a diagram showing the entire float detection system according to the present invention. The lift detection system 1 according to the present invention comprises a mobile body 20 and a processing device 30. The mobile body 20 moves along the exterior wall and acquires and records data for detecting lift in the mortar 11 or tiles 12 on the exterior wall. The processing device 30 acquires the data recorded in the mobile body 20, detects lift, and displays the detection results.

[0048] A mobile body 20 according to the present invention is a mobile body for detecting loosening of mortar 11 or tiles 12 on the exterior wall of a building. The mobile body 20 includes an irradiation unit 21 that irradiates the exterior wall with millimeter waves, a receiving unit 22 that receives millimeter waves reflected from the exterior wall, and a recording unit 23 that records the intensity of the received millimeter wave signal.

[0049] The mobile body 20 may be a drone, a mobile body with wheels or caterpillar tracks, a gondola suspended by a wire, or the like. The means of movement of the mobile body 20 is not critical as long as it is capable of moving along the exterior wall. The mobile body 20 may be, for example, a drone that is operated by an operator using a controller, or may be capable of moving autonomously, such as a self-propelled robot. In the present invention, a mobile body that can fly along the exterior wall, such as a drone, is considered as a preferred example.

[0050] The moving body 20 may further include an imaging unit 24 that acquires an image of the exterior wall. The imaging unit 24 is a camera that can capture an image of the exterior wall, and can acquire RGB color image data or black and white image data. The imaging unit 24 may also be a thermoscope. The imaging unit 24 is provided to obtain sufficient resolution to accurately determine which part of the exterior wall has a lift.

[0051] The moving body 20 may further include a distance measurement unit 25 that measures the distance from the moving body 20 to the tile 12. The distance measurement unit 25 measures the distance from the moving body 20 to the tile 12 and is used when it is necessary to correct the data according to the distance.

[0052] In the moving object 20, the frequency of the millimeter wave is preferably 24 GHz, but millimeter waves of other frequencies may also be used. For example, millimeter waves of 60 GHz or 77 GHz may be used as long as they can penetrate tiles or mortar with a thickness of 10 mm.

[0053] In the mobile body 20, the millimeter waves are preferably electromagnetic waves that can transmit at least 5% of tiles or mortar that is 10 mm thick, but are not limited to this. If an air layer exists between the tile and the mortar, the millimeter waves may be those that can pass partly through the tile and the air layer, or if an air layer exists between the foundation wall and the mortar, the millimeter waves may be those that can pass at least partly through the mortar and the air layer.

[0054] The processing device 30 according to the present invention is a processing device for detecting loose mortar or tiles on the exterior wall of a building. The processing device 30 includes an acquisition unit 31 that acquires the intensity of a millimeter wave signal reflected from the exterior wall, and a loose tile detection unit 32 that detects loose tiles based on the intensity, the intensity being observed by the mobile object 20.

[0055] In the processing device 30, the exterior wall consists of a foundation wall 10, mortar 11, and tiles 12, and the lift detection unit 32 determines that an air layer exists between the tile 12 and mortar 11, or between the mortar 11 and the foundation wall 10, when the strength exceeds a predetermined threshold.

[0056] In the processing device 30, the threshold value may be, for example, one of the following: (a) a value input by the user, (b) a value determined based on measurements of tiles on the exterior wall that are not floating, or (c) a value that divides the multiple measured intensities into normal values ​​and abnormal values ​​that are greater than the normal values.

[0057] The processing device 30 further includes a correction unit 33 that measures the distance between the moving object 20 and the tile 12 and corrects the intensity. The correction unit 33 may correct the intensity according to the distance between the moving object 20 and the tile 12.

[0058] The processing device 30 further includes a display unit 34 that displays an image in which the intensity values ​​are expressed as a heat map superimposed on the image of the tile 12 acquired by the moving body 20 .

[0059] As described above, the lift detection system 1 according to the present invention is a lift detection system 1 for mortar 11 or tiles 12 on the exterior wall of a building, and includes the above-mentioned moving body 20 and the above-mentioned processing device 30.

[0060] FIG. 3 is a diagram showing how millimeter waves are irradiated onto an exterior wall and reflected from the exterior wall in the float detection system according to the present invention. The present invention irradiates millimeter waves onto an exterior wall and determines whether or not an air gap exists between the foundation wall 10 and the mortar 11, or between the mortar 11 and the tile 12, based on the difference in the intensity of the millimeter waves reflected from the exterior wall.

[0061] Fig. 3(a) shows an example in which there is no air gap between the foundation wall 10 and the mortar 11, and between the mortar 11 and the tile 12. In the example in Fig. 3(a), when millimeter waves are irradiated onto the exterior wall, the millimeter waves are reflected back only from the surface of the tile 12, which is the outermost surface.

[0062] Fig. 3(b) shows an example in which there are no tiles 12, only foundation walls 10 and mortar 11, and no air gap exists between the foundation walls 10 and mortar 11. In the example of Fig. 3(b), when millimeter waves are irradiated onto the exterior wall, the millimeter waves are reflected and returned only from the surface of the mortar 11, which is the outermost surface.

[0063] Figure 3(c) shows an example in which there is no air gap between the foundation wall 10 and the mortar 11, but there is an air gap between the mortar 11 and the tile 12. In the example of Figure 3(c), when millimeter waves are irradiated onto the exterior wall, some of the millimeter waves pass through the tile 12 and the air gap, and some are reflected back from the surface of the tile 12, which is the outermost surface, and some are reflected back from the back surface of the tile 12 and the surface of the mortar 11.

[0064] Figure 3(d) shows an example in which an air layer exists between the foundation wall 10 and the mortar 11, but no air layer exists between the mortar 11 and the tile 12. In the example of Figure 3(d), when millimeter waves are irradiated onto the exterior wall, some of the millimeter waves pass through the tile 12, mortar 11, and the air layer, and some are reflected back from the surface of the tile 12, which is the outermost surface, and some are reflected back from the back surface of the mortar 11 and the surface of the foundation wall 10.

[0065] 3(e) shows an example in which there is no tile 12, and only a foundation wall 10 and mortar 11, and an air layer exists between the foundation wall 10 and mortar 11. In the example of FIG. 3(e), when millimeter waves are irradiated onto the exterior wall, some of the millimeter waves pass through the mortar 11 and the air layer, and some are reflected back from the surface of the mortar 11, which is the outermost surface, and some are reflected back from the back surface of the mortar 11 and the surface of the foundation wall 10.

[0066] As shown in (a) to (e) in Figure 3, when an air gap exists (c) to (e), the millimeter waves are reflected back from surfaces other than the outermost surface, so the total strength of the millimeter wave signals reflected back from the outer wall is greater than in (a) and (b), when an air gap does not exist. The presence or absence of an air gap can be determined from this difference in strength.

[0067] FIG. 4 is a diagram showing the relationship between the moving body and the outer wall according to the present invention. The mobile object 20 moves along the exterior wall and irradiates the exterior wall with millimeter waves. While FIG. 4 shows an example in which the mobile object 20 is a mobile object capable of flight, such as a drone, the mobile object may be a mobile object using other means of movement. For example, in the case of a mobile object capable of flight, such as a drone, the mobile object may fly through the air parallel to the exterior wall. When the mobile object 20 irradiates millimeter waves toward the exterior wall, only the portions where the tiles or mortar are loose may appear different from the other portions. For example, the brightness of the portions where the loose tiles are loose may be made brighter than the other portions.

[0068] FIG. 5 is a diagram showing the relationship between the movement of a moving object and the images acquired according to the present invention. Since the area that can be measured by one irradiation of the mobile object 20 is narrow, the mobile object 20 measures the intensity of the millimeter wave signal reflected from the exterior wall while moving along the exterior wall, thereby covering a wide area of ​​the exterior wall. For example, as shown in the example in Fig. 5, the mobile object 20 may irradiate millimeter waves while moving vertically from bottom to top, move parallel to the next row, and then irradiate millimeter waves while moving vertically from top to bottom. The method of scanning the exterior wall is not limited to this example, and any method may be used as long as it covers the entire surface of the exterior wall.

[0069] In FIG. 5, I indicates a region of interest in which lifting is to be detected. W indicates the width of the image, and H indicates the height of the image. OH indicates the horizontal overlap of the image, and OV indicates the vertical overlap of the image. The movement density and sampling density of the moving object 20 are, for example, per tile, but synthetic aperture processing may be used to make the horizontal movement density and vertical sampling density coarser than per tile.

[0070] FIG. 6 is a diagram showing a moving object equipped with a plurality of irradiating units according to another embodiment of the present invention. In the example of Fig. 6, multiple millimeter-wave radars, each having at least the functions of an irradiation unit 21 and a receiving unit 22, are linked together and mounted on a mobile object 20. Fig. 6 is a conceptual example, and the method of linking multiple millimeter-wave radars is not limited to the example of Fig. 6, and other linking methods are also possible. In this way, by providing multiple millimeter-wave radars in the mobile object 20, the travel time of the mobile object 20 during detection work can be shortened. For example, if the mobile object 20 is a drone, the flight time of the drone can be shortened, reducing battery consumption, etc.

[0071] FIG. 7 is a diagram showing an example in which a millimeter wave imaging radar is mounted on a moving object according to the present invention. The horizontal angle of the radar corresponding to one grid unit of the millimeter-wave imaging radar is, for example, about 1°. The vertical field of view angle of the radar corresponding to one grid unit of the millimeter-wave imaging radar is, for example, about 20 to 30°, and the angular resolution is about 1 to 2°.

[0072] FIG. 8 is a flow chart of a method for detecting floatation according to the present invention. The method for detecting loose mortar 11 or tiles 12 on the exterior wall of a building according to the present invention includes a step (S801) in which a mobile body 20 irradiates millimeter waves onto the exterior wall and receives the millimeter waves reflected from the exterior wall, and a step (S802) in which a processing device 30 acquires the intensity of the millimeter wave signal reflected from the exterior wall and received by the mobile body 20, and detects loose tiles based on the acquired intensity.

[0073] Furthermore, the program for detecting loosening of mortar 11 or tiles 12 on the exterior wall of a building according to the present invention is a program that causes a computer to execute the steps of reading the intensity of a millimeter wave signal reflected from the exterior wall from a recording medium, and detecting loosening of tiles 12 based on the read intensity. The program may be built into the processing device 30. Alternatively, the program may be stored in a device other than the processing device 30. Alternatively, the program may be recorded on a recording medium such as a memory, a CD-ROM, a DVD-ROM, a USB, or an SD card. Alternatively, the program may be stored on the cloud, or may be executable via wired or wireless communication. [Example]

[0074] In Example 2, AI (Artificial Intelligence) analysis is performed to detect loose mortar or tiles on an exterior wall. That is, in Example 2, loose mortar or tiles on an exterior wall are detected using a machine learning model.

[0075] Reflection intensity varies depending on the millimeter-wave radar's irradiation angle, distance, surface roughness, etc., and a simple threshold may not be able to identify loose parts. Therefore, using AI is extremely effective for identifying peeling tiles or mortar when conditions are not uniform across the entire area. In particular, the ability to distinguish parts that have a different atmosphere from their surroundings is important for detecting abnormal parts. SAM (Segment Anything Model) is particularly effective for distinguishing parts that have a different atmosphere from their surroundings. Using SAM (Segment Anything Model) in exterior wall inspection can improve the accuracy of loose part detection. SAM is a model that can segment any object or area within an image, making it effective for anomaly detection and exterior wall inspection under different conditions.

[0076] Below is the architecture of an exterior wall inspection system that integrates a thermoscope, color camera, temperature distribution, and AI-based extraction of anomalous areas.

[0077] FIG. 9 is a diagram showing the entirety of a floating detection system according to a second embodiment of the present invention. A system 90 for detecting loose mortar or tiles on the exterior walls of a building according to Example 2 includes a data collection unit 91 that acquires the intensity distribution of millimeter waves reflected from the exterior wall and image data of the exterior wall, a data integration unit 92 that integrates the intensity distribution of millimeter waves reflected from the exterior wall with image data of the exterior wall acquired by a color camera, a preprocessing unit 93 that removes noise from the image data, an AI analysis unit 94 that detects loose mortar or tiles on the exterior wall using a machine learning model, and a result display unit 95 that displays the locations of loose mortar or tiles on the exterior walls of a building. The system 90 for detecting loose mortar or tiles on the exterior walls of a building may further include a report generation unit 96.

[0078] The data collection unit 91 acquires the intensity distribution of millimeter waves reflected from the exterior wall and an image of the exterior wall. The data collection unit 91 may be equipped with a means for transmitting and receiving millimeter waves and a color camera. The means for transmitting and receiving millimeter waves acquires the intensity distribution of millimeter waves reflected from the exterior wall. The color camera captures an image of the exterior wall and acquires a visible image of the exterior wall.

[0079] The data integrator 92 integrates the intensity distribution of the millimeter waves reflected from the exterior wall with the image data of the exterior wall acquired by the color camera. The data integrator 92 integrates the intensity distribution of the millimeter waves reflected from the exterior wall with the image data of the exterior wall acquired by the color camera and maps them into a common coordinate system.

[0080] The pre-processing unit 93 removes noise from the image data and normalizes the data. The pre-processing unit 93 may include a noise removal unit and a temperature data normalization unit. The noise removal unit removes noise from the image data. The temperature data normalization unit normalizes the temperature data.

[0081] The AI ​​analysis unit 94 detects the lifting of the exterior wall using a machine-learned model. It may have a SAM model and an anomaly detection algorithm. The SAM model performs segmentation of abnormal areas, i.e., areas where lifting exists. The anomaly detection algorithm extracts peculiar parts.

[0082] The result display unit 95 visualizes the detection results of the float and displays abnormal areas, i.e., areas where the float exists. The result display unit 95 may highlight areas where the float exists. The result display unit 95 may also have a user interface (for example, a GUI (Graphical User Interface)) for displaying the results.

[0083] The report generating unit 96 automatically generates a detailed report on the abnormal location, that is, the location where the lift exists.

[0084] The learning method of the machine learning model used in the AI ​​analysis unit 94 will be described below. The reflection intensity varies depending on the material, the millimeter-wave radar's irradiation angle, distance, surface roughness, etc., and conditions are not uniform across the entire wall. A simple threshold value may not be able to identify the loose parts. In this situation, to find peeling tiles or mortar, it is necessary to identify the peeling parts among various noises. In this case, a simple threshold value will not determine the loose parts. Here, it is effective to use AI to find specific changes, parts that have a different atmosphere from the surroundings. We will verify this idea and show the AI ​​used here. A simple flowchart is also shown.

[0085] (Verification method and construction of the AI ​​model to be used) 1. Data Collection Reflection intensity distribution by millimeter-wave radar: Data is collected to create a reflection intensity distribution by millimeter-wave radar. A large number of images showing the millimeter-wave reflection intensity distribution of the exterior wall are collected. It is important to collect data taken under a variety of conditions, taking into account different materials, surface roughness, thickness, etc. Annotation: Prepare a dataset in which experts have accurately annotated peeling and other anomalies. 2. Data Preprocessing Noise Reduction: Remove noise from collected data to improve image quality. Normalization: Normalize data under different temperature conditions to align them to a certain standard. 3. AI model selection and training Anomaly detection algorithm: We propose the following AI model as an algorithm for anomaly detection. Convolutional Neural Network (CNN): Due to its high accuracy in image recognition, it is effective in detecting peeling areas on exterior walls. Autoencoder: Used to learn normal states and detect abnormal parts as specific changes. Generative Adversarial Networks (GANs): They have the ability to generate and detect anomalies even from noisy data. 4. Model training and evaluation Model training: The AI ​​model is trained using the collected dataset, including data under different conditions, to improve generalization ability. Evaluate and adjust: Use validation data to evaluate model performance and adjust the model if necessary. 5. Implementation of an anomaly detection system System construction: The trained model is used to build a real exterior wall inspection system. Field testing: Test the system on real buildings to verify accuracy.

[0086] The overall flow from generating the machine learning model used in the float detection method using AI analysis in Example 2 to building, verifying, and operating the float detection system is mainly as follows: (1) data collection, (2) data preprocessing, (3) AI model selection, (4) machine learning model training, (5) machine learning model evaluation and adjustment, (6) anomaly detection system construction, (7) field testing, and (8) start of operation.

[0087] FIG. 10 is a flowchart of a method for detecting floating according to a second embodiment of the present invention. The method for detecting a float according to the second embodiment of the present invention includes a step of collecting data for generating a machine learning model (S1001), a step of collecting data using millimeter waves and a color camera (S1002), a step of integrating the data (S1003), a step of integrating the data using an image integration system (S1004), a step of performing preprocessing (S1005), a step of performing noise removal and normalizing reflection intensity (S1006), an AI analysis step of performing AI analysis (S1007), a step of segmenting abnormal areas using a SAM model and extracting peculiar parts using an anomaly detection algorithm (S1008), a step of displaying the results (S1009), a step of visualizing the abnormal areas using a user interface (S1010), a step of creating a report (S1011), and a step of creating a detailed report using an automatic report generation system (S1012).

[0088] By utilizing AI, it is possible to detect specific changes with high accuracy even in the uneven distribution of reflectance characteristics of exterior walls. In particular, by applying models such as CNN, autoencoder, and GAN, peeling and abnormal areas on exterior walls can be effectively detected. Here, even higher accuracy can be achieved by using SAM (Segment Anything Model). High accuracy can be achieved by using SAM for exterior wall inspection. SAM is a model that can segment any object or area within an image, and is effective for anomaly detection and exterior wall inspection under different conditions.

[0089] (How to use SAM) 1. Data Collection and Annotation Collection of millimeter wave reflection intensity distribution: Taking into account different conditions, millimeter wave reflection intensity distributions of various exterior walls are collected. Annotation: Experts accurately annotate peeling and abnormal areas. SAM uses this annotation data for learning. 2. Data Preprocessing Noise removal and normalization: The data is prepared in a format suitable for input to SAM. Noise is removed and the reflection intensity distribution is normalized. 3. SAM Training Segmentation training: SAM is trained using the collected annotation data, including data under different conditions, to identify abnormalities. 4. Development of an anomaly detection system Model application: Integrate the trained SAM into an exterior wall inspection system. Real-time analysis: Analyzes infrared images in real time and segments anomalies. 5. Evaluate and tune the model Validation and Evaluation: Conduct field tests to evaluate the accuracy of the model. Adjust the model as needed to improve its accuracy.

[0090] The overall process from generating the machine learning model (SAM) used in the exterior wall inspection system using SAM to building, verifying, and operating the lift detection system is as follows: (1) data collection, (2) data preprocessing, (3) SAM training, (4) model application, (5) real-time analysis, (6) identification of abnormalities, (7) field testing, (8) model evaluation and adjustment, and (9) start of operation.

[0091] SAM enables highly accurate segmentation of abnormal areas during exterior wall inspection. Even in environments with uneven reflectance characteristics, SAM is able to easily capture specific changes, making it a highly effective tool for anomaly detection. In this way, peeling tiles and mortar on exterior walls can be identified efficiently and accurately.

[0092] (SAM and anomaly detection algorithms, differences and advantages of SAM and CNN DL for extracting raised areas) 1. SAM (Segment Anything Model) and Anomaly Detection Algorithm (1)SAM(Segment Anything Model) - Uses: SAM is a model for segmenting arbitrary objects or regions in an image. - Advantages: The advantages of SAM are as follows: Flexibility: Any object or region can be segmented. High accuracy: High ability to segment new anomalies based on existing data. Pre-trained: Pre-trained on large datasets, allowing for quick application to specific tasks. (2) Anomaly detection algorithm - Use: Anomaly detection algorithms are algorithms for detecting unusual patterns or outliers. - Advantages: The advantages of the anomaly detection algorithm are as follows: Specificity: Can be designed to specialize in specific anomaly detection tasks. Real-time detection: Simple algorithms allow real-time processing.

[0093] 2. Extraction of raised areas using SAM and CNN (Convolutional Neural Network) (1)SAM - Use: The purpose and use of SAM here is the segmentation of specific objects or regions. - Advantages: The advantages of SAM here are: Versatile: Any part of an image can be segmented with high accuracy. Consistency: Once learned, it performs consistently across different conditions. (3)CNN(Convolutional Neural Network) - Applications: CNN is a deep learning model specialized for image recognition and anomaly detection. - Advantages: The advantages of CNN are as follows: High accuracy: High accuracy in detecting anomalies, especially when trained on large datasets. Adaptability: Adaptable to different types of image data. In particular, the following improvements can be achieved by using SAM: (i) High-precision segmentation of abnormalities - Detailed boundary detection: SAM can segment abnormalities in the image at a very fine level, allowing clear identification of the boundaries of peeling or lifting areas. - Effective even in complex backgrounds: SAM can accurately segment target areas even when there is background noise caused by lighting conditions or wind flow. (ii) Supports various anomaly detection - Responds to various anomalies: SAM can respond to different types of anomalies (cracks, peeling, discoloration due to moisture, etc.) with high accuracy. - Use of pre-trained models: SAM is pre-trained on large datasets, allowing it to quickly adapt to new environments and conditions. (iii) Consistent Performance - Consistency across multiple conditions: Consistent, highly accurate anomaly detection is possible when inspecting exterior walls, which are affected by different sunlight conditions and wind flow. (iv) Rapid Adaptation - Rapid Adaptation to New Data: Leveraging pre-learned knowledge, it can quickly adapt to new datasets and different environments.

[0094] SAM enables highly accurate segmentation of abnormal areas in exterior wall inspections, improving the accuracy and efficiency of anomaly detection. It is particularly effective in detecting peeling or loose tile or mortar on exterior walls, as it exhibits consistent performance even under complex backgrounds and different environmental conditions and is capable of rapid adaptation.

[0095] One of the advantages of SAM (Segment Anything Model) is that it is trained on a large dataset in advance, so the cost of collecting additional training data for new tasks or environments is low. The advantages of SAM are explained in detail below.

[0096] (Advantages of SAM and reduced training data collection costs) 1. Use of pre-trained models SAM is already pre-trained on a large amount of data to adapt to a wide variety of images and environments, significantly reducing the need for additional data collection and training when adapting to new applications. 2. High versatility SAM's ability to consistently segment a variety of objects and regions with high accuracy allows it to perform well across different conditions and backgrounds, making it easy to adapt the model to specific exterior wall inspections. 3. Reducing the burden of data annotation By utilizing existing pre-trained models, additional data annotation work is minimized, saving the cost and time of detailed annotation work by experts. 4. Rapid Adaptation and Deployment Because SAM has already learned knowledge, it can quickly adapt to new tasks, facilitating system development and deployment.

[0097] (Advantages of SAM when building a system) 1. Rapid prototyping: - Using pre-trained SAMs, you can quickly build and test prototypes, ensuring highly accurate results from the earliest stages of your system. 2. Cost-effective data collection: Even when new data collection is required, SAM's powerful versatility allows highly accurate results to be obtained with a small amount of data, significantly reducing data collection costs. 3. Scalability: - High scalability, as once the system is built, it can be easily applied to other projects and different environmental conditions. 4. Improved user experience: - High-precision segmentation enables users to quickly and accurately identify anomalies, improving the efficiency and accuracy of inspection work.

[0098] In particular, using SAM significantly reduces data collection costs when building an exterior wall inspection system, while enabling rapid and highly accurate anomaly detection. The powerful versatility and high performance of the pre-trained model makes it easy to adapt to new environments and tasks, enabling cost-effective system development. For research on the use and benefits of SAM (Segment Anything Model), see 1. Benefits of SAM and reduction of training data collection costs SAM has been trained on a large-scale dataset (SA-1B) in advance, and is highly versatile across a wide range of images and environments. This significantly reduces the need for additional data collection and training when applying it to new tasks. Specifically, it has the following advantages: - High versatility and flexibility: SAM can consistently and accurately segment a variety of objects and regions, and performs well across different conditions and backgrounds. - Rapid adaptation: The system can be quickly adapted to new tasks, facilitating development and deployment. - Low-cost data collection: Using pre-trained models minimizes the need for additional data annotation, saving the cost and time of detailed annotation by experts.

[0099] 2. Specific application examples of SAM SAM is also highly effective in exterior wall inspections under different conditions. For example, it enables highly accurate segmentation of abnormal areas and detailed boundary detection, which is extremely useful for clearly identifying the boundaries of peeling or loose areas. Furthermore, its consistent performance and rapid adaptability, even in complex backgrounds and under different environmental conditions, make it extremely effective in detecting peeling or loose areas of exterior wall tiles and mortar.

[0100] According to the wall material lift detection system 1 of the present invention described above, it is possible to detect lifting of mortar or tiles on the exterior walls of a building using millimeter waves, and it is possible to detect lifting of mortar or tiles on the exterior walls of a building with high accuracy, regardless of environmental changes. Although the above description has been given with reference to the preferred embodiment, it will be apparent to those skilled in the art that the present invention is not limited thereto, and that various changes and modifications can be made within the scope of the principles of the present invention and the appended claims. [Explanation of symbols]

[0101] 1. Float detection system 10 Foundation wall 11 Mortar 12 tiles 20 Mobile 21 Irradiation unit 22 Receiving unit 23 Recording Section 24 Imaging unit 25 Distance measurement unit 30 Processing equipment 31 Acquisition Department 32 Float detection unit 33 Correction unit 34 Display section

Claims

1. A mobile object for detecting loose mortar or tiles on the exterior wall of a building, an irradiation unit that irradiates the outer wall with millimeter waves; a receiving unit that receives millimeter waves reflected from the outer wall; A recording unit that records the strength of the received millimeter wave signal; A moving body comprising:

2. The moving body according to claim 1 , further comprising an imaging unit that captures an image of the exterior wall.

3. The moving body according to claim 1 , further comprising a distance measurement unit that measures the distance from the moving body to the tile.

4. The mobile body according to claim 1 , wherein the millimeter wave has a frequency of 24 GHz.

5. 2. The mobile body according to claim 1, wherein the millimeter waves are electromagnetic waves that transmit 5% or more of a tile or mortar having a thickness of 10 mm.

6. The moving body of claim 1 , wherein the moving body is a drone, a moving body having wheels or caterpillar tracks, or a gondola suspended by a wire.

7. A processing device for detecting loose mortar or tiles on the exterior wall of a building, an acquisition unit that acquires the intensity of a millimeter wave signal reflected from the outer wall; a lifting detection unit that detects lifting of the tile based on the strength; Equipped with A processing device, characterized in that the intensity is observed by a moving object.

8. The exterior wall is made of a foundation wall, mortar, and tiles, 8. The processing device according to claim 7, wherein the lift detection unit determines that an air layer exists between the tile and the mortar, or between the mortar and the foundation wall, when the intensity exceeds a predetermined threshold value.

9. The threshold value is (a) a value input by the user; (b) A value determined based on measurements of tiles that are not floating on the exterior wall, or (c) a value that divides the measured intensities into normal values ​​and abnormal values ​​that are greater than the normal values; 9. The processing device of claim 8, wherein the processing device is any one of the following:

10. The processing device according to claim 7 , further comprising a correction unit that measures a distance between the moving object and the tile and corrects the intensity.

11. The processing device according to claim 7 , further comprising a display unit that displays an image in which the intensity values ​​are expressed as a heat map, superimposed on the image of the tile acquired by the moving body.

12. A system for detecting loose mortar or tile on the exterior wall of a building, comprising: A moving body according to any one of claims 1 to 6; The processing device according to any one of claims 7 to 11. A float detection system comprising:

13. A method for detecting loose mortar or tile on the exterior wall of a building, comprising: a step of a moving body irradiating millimeter waves onto the outer wall and receiving millimeter waves reflected from the outer wall; a step of detecting, by a processing device, the strength of a millimeter wave signal reflected from the exterior wall and received by the mobile body, and detecting, by the processing device, the loosening of the tile based on the acquired strength; A method comprising:

14. A program for detecting loose mortar or tiles on the exterior wall of a building, reading from a recording medium the intensity of the millimeter wave signal reflected from the outer wall; detecting the tile lift based on the read intensity; A program characterized by causing a computer to execute the above.

15. A system for detecting loose mortar or tile on the exterior wall of a building, comprising: a data collection unit that acquires an intensity distribution of millimeter waves reflected from the exterior wall and image data of the exterior wall; a data integration unit that integrates the intensity distribution of the millimeter waves reflected from the exterior wall and the image data of the exterior wall acquired by a color camera; a preprocessing unit that removes noise from the image data; an AI analysis unit that detects the lifting of the exterior wall using a machine-learned model; A result display section that shows where the float exists. A system comprising:

16. The system of claim 15 further comprising a report generator that automatically generates a report regarding the location of the lift.

17. 1. A method for detecting loose mortar or tile on an exterior wall of a building, comprising: collecting data for generating a machine learning model; collecting data with a mmWave and a color camera; Integrating the data; A pre-processing step; performing noise removal and reflection intensity normalization; An AI analysis step of detecting lifting of the exterior wall using a machine learning model; Segmenting the abnormality area using a Segment Anything Model (SAM) model and extracting the anomalous part using an anomaly detection algorithm; displaying the results; visualizing the anomaly location on a user interface; Steps to create a report A method comprising:

Images