JP2021089185A - Concrete structure diagnostic system, concrete structure diagnostic method, and program - Google Patents

Abstract

To provide a concrete structure diagnostic system capable of easily diagnosing a state of a concrete structure even if a temperature change of concrete is small.SOLUTION: A concrete structure diagnostic system 1 includes: an imaging section 10 for capturing a thermal image indicating a temperature distribution of a predetermined area to be inspected on a surface of concrete; and a diagnostic device 20 that generates a first thermal image by removing noise from the thermal image, generates a second thermal image by averaging a range in which an uneven temperature gradient generated in the first thermal image, generates a temperature change image by emphasizing a portion in which a temperature change is locally suddenly generated in the second thermal image as an edge, generates an edge image by converting a temperature change amount in the temperature change image into an absolute value, generates a third thermal image by adding data obtained by multiplying data of the edge image by a predetermined coefficient to data of the second thermal image, and extracts a temperature varying region on the basis of a predetermined threshold value in the third thermal image.SELECTED DRAWING: Figure 1

Description

The present invention relates to a concrete structure diagnostic system for diagnosing the state of concrete, a concrete structure diagnostic method and a program.

In a concrete structure, there is a risk that deformed parts such as so-called floating or peeling may occur, such as a part of a concrete block separating on the surface layer of concrete due to the cause of corrosion and expansion of the internal reinforcing bars due to aging. is there. Therefore, it is important to take measures such as inspecting the surface of the concrete structure to detect and repair the deformed part at an early stage.

In order to diagnose a concrete structure, a hitting sound inspection is generally performed in which the surface of the concrete structure is hit with a hammer or the like and a deformed portion generated on the concrete surface layer due to a difference in sound is detected. In the tapping sound inspection, it is often necessary to build a scaffold on the surface of the structure and work at a high place for inspection, which is troublesome and requires a long period of time because the inspection range becomes wider according to the size of the structure. .. Further, in the tapping sound test, the operator diagnoses the difference in tapping sound, so that the diagnosis result may vary.

Patent Document 1 describes a method of diagnosing the state of a concrete structure based on an captured thermal image. According to the information processing apparatus described in Patent Document 1, data processing of thermal image data taken by an infrared camera is performed to extract a temperature change portion.

Japanese Unexamined Patent Publication No. 2012-98170

A concrete structure such as a tunnel has a problem that the internal temperature is stable, it is difficult to capture the temperature change generated on the surface of the concrete by imaging a thermal image, and it is difficult to extract a deformed portion of the concrete.

An object of the present invention is to provide a method for diagnosing a concrete structure and a system for diagnosing a concrete structure, which can easily diagnose the state of the concrete structure even if the temperature change of the concrete is small.

In order to achieve the above object, the present invention is a concrete structure diagnosis system for diagnosing a deformed portion of a concrete structure, and images a thermal image showing a temperature distribution of a predetermined area of a concrete surface to be inspected. The imaging unit and the first thermal image in which noise is removed from the thermal image are generated, and a second thermal image is generated by averaging the range in which the temperature gradient generated in the first thermal image is inhomogeneous. A temperature change image is generated in which a portion of a two-heat image in which a sudden temperature change occurs locally is emphasized as an edge, and the amount of temperature change in the temperature change image is made absolute to generate an edge image, and the second heat is generated. A third thermal image is generated by adding data obtained by multiplying the image data by a predetermined coefficient to the edge image data, and the temperature is changed based on a predetermined threshold in the third thermal image. It is a concrete structure diagnostic system characterized by including a diagnostic device for extracting an area.

According to the present invention, by performing a process of multiplying a deformed portion in a thermal image by a predetermined coefficient and emphasizing it, an image obtained by extracting the deformed portion of the concrete structure can be displayed even if the temperature change of the concrete is small.

Further, in the present invention, the diagnostic apparatus may be configured to generate a diagnostic image from which the temperature change region is extracted.

According to the present invention, it is possible to visualize a deformed portion that cannot be discriminated by a visible image by performing a process of highlighting the deformed portion.

Further, the present invention is configured such that when the temperature of the external environment of the concrete rises, the diagnostic apparatus extracts a region equal to or higher than a predetermined threshold value in the third thermal image as a temperature changing region. It may have been done.

According to the present invention, a region predicted to be a deformed portion in the third thermal image under environmental conditions in which the outside air temperature rises can be extracted from the third thermal image and displayed.

Further, the present invention is configured such that when the temperature of the external environment of the concrete is lowered, the diagnostic apparatus extracts a region equal to or less than a predetermined threshold value in the third thermal image as a temperature changing region. You may be.

According to the present invention, a region predicted to be a deformed portion in the third thermal image can be extracted from the third thermal image and displayed under environmental conditions in which the outside air temperature decreases.

Further, in the present invention, the diagnostic apparatus extracts a temperature change region based on the temperature parameter value in the third thermal image, and calculates an area ratio which is a ratio of the predetermined area to the extracted area of the temperature change region. In the relationship between the area ratio and the temperature parameter value, the temperature value at the bending point of the area ratio is set as a threshold value, and the temperature change region extracted from the third thermal image based on the threshold value is defined as the deformation. It may be configured to be determined as a part.

According to the present invention, by setting a threshold value for accurately extracting the temperature change region in the thermal image, it is possible to display an image in which the deformed portion of the concrete structure is extracted even if the temperature change of the concrete is small.

Further, the present invention is determined by the fact that the diagnostic apparatus detects the deformed portion in a predetermined ratio or more based on a plurality of the deformed portions as a sample of the deformed portion in the third thermal image. It may be configured to extract the region which is equal to or higher than the threshold value of the temperature as the deformed portion.

According to the present invention, the temperature of concrete is set by setting a threshold value for accurately extracting a temperature change region in a thermal image so that a deformed portion is detected at a predetermined ratio or more by using a visible image and a thermal image. Even if there is little change, it is possible to display an image extracted from the deformed part of the concrete structure.

The present invention also presents a dispersion of a plurality of the thresholds determined by the diagnostic apparatus by changing the position and time of the diagnosis target in the concrete structure and the difference between the average temperature value of the healthy region determined by the thresholds. Is calculated, a calibration curve for calibrating the threshold is calculated based on the relationship appearing in the scatter plot based on the variance and the difference, the third thermal image of another diagnostic target is generated, and the third thermal image is generated. It may be configured to calculate the temperature variance of the provisional region excluding the region having a peculiar temperature from the above and set the temperature threshold in the other diagnostic object using the calibration curve.

According to the present invention, by calculating a calibration curve for calibrating the threshold value using the past sample data extracted based on the threshold value, the accuracy is higher regardless of the conditions such as the location condition of the tunnel and the distance from the wellhead. A high threshold can be calculated.

In the present invention, the imaging unit captures a visible image of a predetermined area of the concrete surface to be inspected, and the diagnostic apparatus uses the deformed portion and the third thermal image determined based on the third thermal image. Machine learning is performed by comparing with the visible image in which the deformed portion corresponding to the above is marked, the threshold value is calculated based on the result of the machine learning, and the new thermal image captured by the imaging unit is used. A new third thermal image may be generated, and the deformed portion may be extracted from the new third thermal image based on the calculated threshold value.

According to the present invention, the threshold value based on the machine learning result can be determined according to the imaging environment, and the deformed portion can be accurately extracted from the third thermal image.

In order to achieve the above object, the present invention is a concrete structure diagnosis method for diagnosing a deformed portion of a concrete structure, in which a computer shows a thermal image showing a temperature distribution of a predetermined area to be inspected on the surface of the concrete. A step of generating a first thermal image in which noise is removed from the heat image, a step of generating a second thermal image obtained by averaging a range in which the temperature gradient generated in the first thermal image is inhomogeneous, and a step of generating the second thermal image. A step of generating a temperature change image in which a portion where a sudden temperature change occurs locally is emphasized as an edge, a step of absoluteizing the amount of temperature change in the temperature change image and generating an edge image, and the second step. Based on the step of generating a third thermal image by adding data obtained by multiplying the thermal image data by a predetermined coefficient to the edge image data and a predetermined threshold value in the third thermal image. It is a concrete structure diagnosis method including a step of extracting a temperature change region.

In order to achieve the above object, the present invention is a program applied to a concrete structure diagnosis system for diagnosing a deformed part of a concrete structure, and a computer is used to heat a predetermined area of a concrete surface to be inspected. A first thermal image in which noise is removed from the thermal image showing the distribution is generated, and a second thermal image in which the range in which the temperature gradient generated in the first thermal image is inhomogeneous is averaged is generated to generate the second heat. A temperature change image is generated in which a portion of the image in which a sudden temperature change occurs locally is emphasized as an edge, and the amount of temperature change in the temperature change image is made absolute to generate an edge image. The data obtained by multiplying the data of the edge image by a predetermined coefficient is added to the data to generate a third thermal image, and the temperature change region is set based on the predetermined threshold value in the third thermal image. It is a program characterized by being extracted.

According to the present invention, it is possible to provide a concrete structure diagnosis method and a concrete structure diagnosis system that can easily diagnose the state of a concrete structure even if the temperature change of the concrete is small.

It is a figure which shows the structure of the concrete structure diagnostic system which concerns on embodiment of this invention. It is a figure which shows the visible image of the surface of the concrete structure which a deformed part occurred. It is a figure which shows the thermal image and the binarized image of the deformed part. It is a figure which shows the 1st thermal image and the binarized image which removed the noise component. It is a figure which shows the 2nd thermal image and the binarized image which averaged the region where the temperature gradient is inhomogeneous. It is a figure which shows the 3rd thermal image and the binarized image which emphasized the deformed part. It is a figure which shows the diagnostic image which combined the image which extracted the temperature change region with the visible image. It is a flowchart which shows the flow of the process executed in the concrete structure diagnosis system. It is a figure which shows the relationship between the threshold value and the extraction area. It is a figure which shows the threshold value and the extraction rate of a floating part. It is a figure which shows the calibration curve of the threshold value generated based on the cumulative data.

Hereinafter, embodiments of the concrete structure diagnosis method and the concrete structure diagnosis system according to the present invention will be described with reference to the drawings.

As shown in FIG. 1, the concrete structure diagnostic system 1 carries an imaging unit 10 that captures an image in an imaging region, a diagnostic device 20 that processes an image captured by the imaging unit 10, and a moving image pickup unit 10. A gantry 30 is provided. The imaging unit 10 and the diagnostic device 20 are connected, for example, by wire or wirelessly.

The imaging unit 10 includes, for example, an infrared camera 11 and a visible image camera 12. The infrared camera 11 captures a thermal image showing the temperature distribution of the concrete surface of the structure. As the infrared camera 11, a general infrared thermography device can be used. The visible image camera 12 captures a visible image of the concrete surface. As the visible image camera 12, a general digital camera can be used.

The imaging unit 10 is installed on, for example, a mobile pedestal 30. The moving pedestal 30 moves while being separated from the surface of the concrete structure to be imaged by a predetermined distance. The moving gantry 30 allows the imaging unit 10 to capture a continuous image to be imaged. The imaging unit 10 also includes a range finder 16 that measures the moving distance of the moving pedestal 30. As the range finder 16, for example, an encoder, a laser range finder, or the like is used.

The mobile gantry 30 is a self-propellable mobile trolley. The moving pedestal 30 moves the imaging unit 10 by a predetermined distance in the traveling direction of the tunnel. At this time, the imaging unit 10 continuously images the captured image data of the inner wall surface (concrete surface) of the tunnel. The moving pedestal 30 may be one that automatically travels or one that is moved by an operator.

In the mobile gantry 30, the infrared camera 11 and the visible image camera 12 are installed so as to face in the same direction so as to image substantially the same overlapping region on the inner wall surface of the tunnel. In the mobile gantry 30, the infrared camera 11 and the visible image camera 12 are movable so as to be able to capture continuous images in a strip shape along the inner wall of the arch-shaped tunnel.

After imaging one strip-shaped region, the moving pedestal 30 is moved by a predetermined distance in the tunnel traveling direction, so that the imaging unit 10 can image an image of the inner wall of the tunnel. The plurality of images captured by the imaging unit 10 are image-processed by the diagnostic apparatus 20, the overlapping portion is removed, and a continuous image of the entire inner wall of the tunnel is generated.

The image pickup unit 10 outputs the captured image data and the distance data of the range finder 16 to the diagnostic apparatus 20. The image data captured by the imaging unit 10 may be provided to the diagnostic apparatus 20 via a storage medium capable of recording data such as a memory card, a hard disk drive, or a CD-ROM or DVD-ROM.

The diagnostic device 20 is realized by, for example, a terminal device such as a personal computer (PC), a tablet-type terminal device, or a smartphone. The diagnostic device 20 is connected to the imaging unit 10 in a communicable manner by wire or wirelessly. The diagnostic device 20 may be mounted on the mobile gantry 30 or may be installed at a place different from the mobile gantry 30. The diagnostic device 20 may be a server device connected to the imaging unit 10 by a network. The diagnostic device 20 includes an acquisition unit 22 that acquires an image from the image pickup unit 10, a calculation unit 24 that calculates a diagnosis result based on the image, a display unit 26 that displays the diagnosis result, and a storage unit 28 that stores various information. And.

The acquisition unit 22 is an interface for inputting captured image data and distance data output from the imaging unit 10. The acquisition unit 22 is a wired or wireless device. The acquisition unit 22 may be an interface for reading data from a storage medium such as a USB, a card reader, a CD drive, or a DVD drive. The acquisition unit 22 stores the acquired image data in the storage unit 28.

The arithmetic unit 24 is realized by, for example, a hardware processor such as a CPU (Central Processing Unit) executing a program (software). Some or all of these components are hardware such as LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), GPU (Graphics Processing Unit), etc. It may be realized by (including circuits), or it may be realized by the cooperation of software and hardware. The program may be stored in advance in a storage device such as an HDD (Hard Disk Drive) or a flash memory, or is stored in a removable storage medium such as a DVD or a CD-ROM, and the storage medium is stored in the drive device. It may be installed by being attached. Further, this program may be distributed to a computer via a communication line, and the computer receiving the distribution may execute the program.

The calculation unit 24 reads the image data from the storage unit 28. The calculation unit 24 performs image processing of the thermal image based on the distance data and the image data. The calculation unit 24 combines the image-processed thermal image and the visible image to generate a diagnostic image in which the floating or peeling region generated on the concrete surface is extracted. The detailed contents of the processing of the calculation unit 24 will be described later. The calculation unit 24 controls the display unit 26 to display the display contents such as the diagnostic image generated.

The display unit 26 displays the image generated by the calculation unit 24. The display unit 26 is realized by, for example, a display device such as a liquid crystal display, an organic EL display, or an LED display.

The storage unit 28 stores image data and data necessary for the calculation of the calculation unit 24. The storage unit 28 is a storage device such as an HDD, a flash memory, a RAM (Random Access Memory), or a ROM (Read Only Memory).

Next, the process executed by the diagnostic apparatus 20 will be described.

When floating or peeling occurs on the surface layer of the concrete to be diagnosed, the concrete lump of the floating or peeled part separates from the concrete structure. This concrete block has the property of being easier to heat and cool than other parts. Therefore, when a temperature change occurs in the atmosphere in contact with the concrete structure, the temperature change of the concrete block occurs at a higher gradient than the other parts. Therefore, when an infrared image of the surface of concrete is imaged in an environment where the temperature changes in the outside air, the region of the concrete block in which floating or peeling occurs appears as a temperature changing region having a temperature different from that of the surrounding concrete.

As shown in FIG. 2, in the visible image P in which a part of the surface of the concrete in which the known deformed portion such as a float is generated is captured, the deformed portion cannot be visually recognized.

As shown in FIG. 3, the thermal image H shows the temperature distribution on the surface of the concrete to be imaged. The thermal image H is generated so that the color (luminance) of the pixels changes according to the temperature. In the figure, for reference to the thermal image H, a binarized image Q displaying a portion higher than the threshold value described later is shown side by side.

It is difficult to directly extract a deformed portion such as floating or peeling based on the thermal image H itself. In the environment where the temperature changes as described above, the deformed portion can be extracted by the image processing of the thermal image H. However, as described above, the temperature inside the tunnel structure is stable as compared with the external environment, and it is difficult to extract the deformed portion by the image processing of the thermal image H.

The process executed in the diagnostic apparatus 20 is to highlight and extract the deformed portion based on the thermal image in the environment where the temperature change is stable. In the following description, a state in which the temperature of the temperature-changing region of the floating or peeling portion is higher than that of the surroundings in the thermal image H is illustrated.

In addition to the temperature change region, the thermal image H includes a region where the temperature is high even though it is not a noise component or a floating or peeling portion. The calculation unit 24 removes local noise in the thermal image H by filtering processing based on the data of the thermal image H. The arithmetic unit 24 removes the noise of the thermal image H by, for example, image processing using the median filter processing. The median filter sets the center pixel as the pixel of interest in the area of 9 pixels of 3 × 3, calculates the median of the 8 pixels around each pixel of interest, and calculates each pixel of interest. Is replaced with the median value of the surrounding eight pixels.

As shown in FIG. 4, the calculation unit 24 performs the above processing on all the pixels, removes a local noise component from the thermal image H, and generates a first thermal image H1. The arithmetic unit 24 may remove noise in the thermal image H by using a moving average filter process, a Gaussian filter process, or the like in addition to the median filter process.

On the concrete surface, there is a temperature difference between the upper part and the lower part of the tunnel, before and after the unevenness of the surface shape, and between the part hidden by the accessories and the part not. Gradient (unevenness) occurs. When such a temperature gradient exists in the first thermal image H1 after the noise processing, it becomes difficult to extract the floating or peeled portion.

The calculation unit 24 performs a filter process for averaging a range in which the temperature gradient in the first thermal image H1 is inhomogeneous. The calculation unit 24 calculates, for example, the average temperature TA, which is the average of the temperature values of the entire pixels of the first thermal image H1. The calculation unit 24 calculates the average temperature TB of the peripheral region of a certain pixel of interest with respect to the overall average temperature TA of the first thermal image H1. The peripheral area is a rectangular image range in which one side is set with a predetermined dimension. The calculation unit 24 calculates, for example, the temperature correction value TC obtained by correcting the temperature value T of the pixel of interest of the first thermal image H1 based on the following mathematical formula (1).
TC = T + TA-TB (1)
By performing the above processing on all the pixels of the first thermal image H1, the temperature gradient generated on a macro scale can be reduced.

As shown in FIG. 5, the calculation unit 24 performs the above-mentioned filter processing on all the pixels to generate the second thermal image H2. In the second thermal image H2, the non-uniformity of the temperature gradient occurring over a wide area on the concrete surface can be suppressed.

In the second thermal image H2, the calculation unit 24 calculates the temperature gradient of a certain pixel of interest by performing four-direction differentiation in the vertical, horizontal, and diagonal directions with respect to the temperature difference between the eight adjacent pixels. The calculation unit 24 superimposes each of the calculated temperature gradients in each direction. The calculation unit 24 performs the above processing on all the pixels to generate a temperature change image in which the temperature change is emphasized. By the above processing, a temperature change image in which a portion (edge) where a temperature change occurs locally and rapidly is emphasized can be obtained. Since the temperature change image in which the edge is emphasized is an image displaying the temperature change, for example, when the temperature rises from left to right, it shows a positive value, and when it falls, it shows a negative value.

In order to emphasize the edge in the temperature change image, a process of removing the sign of the amount of temperature change and converting it into an absolute value is performed by paying attention only to the height of the absolute value regardless of whether the temperature change is positive or negative. The calculation unit 24 generates an edge image by converting the data of the temperature change image into an absolute value. The calculation unit 24 generates emphasis data obtained by multiplying all pixels (temperature change amount) by a predetermined coefficient α in the edge image. Here, the value of α represents the intensity of edge enhancement, and an appropriate value is selected in advance so that a desired image can be obtained.

The calculation unit 24 calculates a combined value of the pixel of interest (temperature data) of the second thermal image H2 and the pixel in the emphasis data at the same position as the pixel of interest. That is, the amount of temperature change of α times the same position as the pixel of interest is added to the pixel of interest in the second thermal image H2, and the calculation process of “temperature value + amount of temperature change of α times” is performed to calculate the combined value. To.

As shown in FIG. 6, the calculation unit 24 performs the above processing on all the pixels, and makes corrections based on the original temperature value so as to emphasize the portion where the local temperature change occurs. The performed third thermal image H3 is generated.

The calculation unit 24 compares the value of the pixel with the threshold value by using the threshold value corresponding to the condition described later of the third thermal image H3, and determines that the pixel above the threshold value is the temperature change region. The threshold value is determined in advance according to the condition in which the temperature change region is most easily extracted in the third thermal image H3, and is stored in the storage unit 28. In the present embodiment, the threshold value on the high temperature side where the temperature change region is high temperature is set to, for example, 26.16 degrees.

As shown in FIG. 7, the calculation unit 24 determines that the region that is equal to or greater than the threshold value and is the temperature change region based on the third thermal image H3, and extracts the temperature change region from the diagnostic image S. Generate.

Generally, when the thermal image H is processed in an environment where the temperature rises such as in the daytime, the temperature of the floating portion is higher than that in the healthy region, so pixels above the threshold value on the high temperature side are extracted as the temperature changing region. On the contrary, when the thermal image H is processed in an environment where the temperature drops such as at night, the temperature of the floating portion is lower than that in the healthy region, so pixels below the threshold value on the low temperature side are extracted as the temperature changing region. ..

Next, the flow of processing executed in the concrete structure diagnosis system 1 will be described.

As shown in FIG. 8, the thermal image H is captured by the infrared camera 11 (step S10). A visible image is captured by the visible image camera 12 (step S12). The calculation unit 24 generates a first thermal image in which noise is removed from the thermal image H (step S14). The calculation unit 24 generates a second thermal image obtained by averaging a range in which the temperature gradient generated in the first thermal image is inhomogeneous (step S16). The calculation unit 24 generates a temperature change image in which a portion of the second thermal image in which a sudden temperature change occurs locally is emphasized as an edge (step S18).

The calculation unit 24 makes an absolute value of the amount of temperature change in the temperature change image and generates an edge image (step S20). The calculation unit 24 generates a third thermal image by adding data obtained by multiplying the edge image data by a predetermined coefficient with respect to the data of the second thermal image (step S22). The calculation unit 24 extracts the temperature change region based on a predetermined threshold value in the third thermal image (step S24). The calculation unit 24 generates a diagnostic image by synthesizing the image in which the temperature change region is extracted with the visible image (step S26). Each of the above steps does not necessarily have to be processed in the above order, and may be performed before or after.

Next, a method of determining the threshold value used for the image processing executed by the calculation unit 24 will be described.

The calculation unit 24 sets a temperature parameter value with the threshold value as a variable in order to determine the threshold value, and changes from the third thermal image showing the temperature distribution of a predetermined area of the concrete surface to be inspected based on the temperature parameter value. Extract the warm region. The calculation unit 24 calculates the area ratio, which is the ratio of the predetermined area to the extracted area of the temperature change region.

As shown in FIG. 9, in the figure showing the relationship between the temperature parameter and the area ratio, the calculation unit 24 calculates the second derivative based on the temperature parameter of the area ratio. The calculation unit 24 calculates the maximum value (inflection point) of the second derivative value and sets the temperature parameter value at the inflection point as a threshold value. The calculation unit 24 determines the temperature-changing region extracted from the third thermal image based on the threshold value set as the deformed portion.

The threshold is also set by other methods. The imaging unit 10 captures a sample image of a plurality of visible images and a plurality of thermal images of the deformed portion marked on the floating portion of the concrete surface which is known in advance. The calculation unit 24 generates a plurality of third thermal images of the sample. The calculation unit 24 extracts a region equal to or larger than the threshold value based on the third thermal image.

As shown in FIG. 10, the calculation unit 24 sets the temperature at which the deformed portion of the concrete surface becomes a predetermined ratio (for example, 80%) or more based on the third thermal image while changing the threshold value as the temperature parameter value. Determined as a threshold. The threshold value in the field environment can be determined based on the above threshold value setting method.

Since the temperature of the healthy area changes depending on the location conditions of the tunnel and the distance from the wellhead, it is necessary to set the threshold value in consideration of these effects. Therefore, the threshold value in the on-site environment is set based on the above-mentioned threshold value setting method.

As shown in FIG. 11, a scatter plot showing the relationship between the variance of the temperature value in the healthy region and the set threshold value is created by using the above two methods. A plurality of threshold values are obtained based on a plurality of samples in which the position and time of the diagnosis target are changed. The scatter plot shows, for example, the relationship between the variance and the difference between the corresponding threshold and the average temperature. The scatter plot shows the data group A in the tunnel A and the data group B in the tunnel B.

The calculation unit 24 calculates the variance between the difference between the plurality of threshold values determined by changing the position and time of the diagnosis target in the concrete structure and the average temperature value of the healthy region. The calculation unit 24 calculates a calibration curve Z that calibrates the threshold value based on the relationship that appears in the scatter plot as the relationship between the variance and the difference. According to the calibration curve Z, the threshold value can be set according to the difference in conditions such as the location condition of the tunnel and the distance from the wellhead.

The calculation unit 24 generates a third thermal image of the surface of concrete to be newly diagnosed. Based on the third thermal image of the surface of the concrete to be diagnosed, the calculation unit 24 extracts a provisional region which is a provisional sound region excluding a peculiar temperature region. The calculation unit 24 calculates the variance of the temperature in the provisional region, and applies the calculated variance to the calibration curve Z to calculate the threshold value. The calculation unit 24 sets the calculation result to the threshold value of another diagnosis target. By the above processing, the calculation unit 24 can accurately extract the deformed portion from the third thermal image by correcting the threshold value applied to the third thermal image based on the newly captured thermal image using the calibration curve Z. ..

Machine learning may be used to extract the deformed portion. For example, when a large amount of visible images of a sample marking a deformed portion and a large amount of a third thermal image corresponding to the visible image of the sample are accumulated, the calculation unit 24 is based on the sample image and the third thermal image. Then, deep machine learning using deep learning or the like using a neural network may be performed to determine the deformed portion in the third thermal image.

The calculation unit 24 reads the third thermal image and extracts the deformed portion in the third thermal image based on a preset threshold value. The calculation unit 24 compares the extracted deformed portion with the marking portion in the visible image of the sample, and corrects the weighting for the temperature parameter according to the imaging environment applied to the threshold value based on the error. The calculation unit 24 repeats the same processing for a large amount of the third thermal image and the visible image of the corresponding sample, and generates a model including the temperature parameter corresponding to the imaging environment applied to the threshold value.

The calculation unit 24 generates a third thermal image from the newly captured thermal image, determines a temperature parameter based on the generated threshold model, calculates a threshold, and uses the third thermal image based on the threshold. Determine the deformed part. The calculation unit 24 may not only determine the deformed portion from the third thermal image by learning, but may also directly determine the deformed portion from the thermal image by machine learning. By the above processing, the calculation unit 24 can determine the threshold value according to the imaging environment based on the learning result of machine learning, and can accurately extract the deformed portion from the generated third thermal image.

As described above, according to the concrete structure diagnosis system, it is possible to easily extract the deformed portion such as floating or peeling generated on the surface of the concrete structure. According to the concrete structure diagnosis system, even in an environment where there is little temperature change such as in a tunnel, the deformed part can be easily discriminated by image processing that emphasizes the deformed part. According to the concrete structure diagnostic system, the boundary of the temperature change region on the thermal image is detected based on the set threshold value, regardless of the operator's visual inspection or tapping sound inspection, so that the floating or peeling region can be made highly accurate. Can be judged. According to the concrete structure diagnosis system, the inspection can be labor-saving, and the floating or peeling area can be determined regardless of the conditions such as the location condition of the tunnel and the distance from the wellhead.

Although one embodiment of the present invention has been described above, the present invention is not limited to the above-mentioned one embodiment, and can be appropriately modified without departing from the spirit of the present invention. For example, in the above embodiment, it is illustrated that a floating or peeling region generated on the surface of concrete in a tunnel is extracted, but the present invention can be applied not only to a tunnel structure but also to other concrete structures. Good.

1 Concrete structure diagnostic system 10 Imaging unit 11 Infrared camera 12 Visible image camera 16 Distance meter 20 Diagnostic device 22 Acquisition unit 24 Calculation unit 26 Display unit 28 Storage unit 30 Mobile mount A, B Data group H Thermal image H1 First thermal image H2 2nd thermal image H3 3rd thermal image P Visible image Q Binarized image S Diagnostic image Z Calibration curve

Claims (10)

A concrete structure diagnosis system that diagnoses deformed parts of concrete structures.
An imaging unit that captures a thermal image showing the temperature distribution of a predetermined area of the concrete surface to be inspected.
A first thermal image in which noise is removed from the thermal image is generated, a second thermal image is generated by averaging a range in which the temperature gradient generated in the first thermal image is inhomogeneous, and in the second thermal image. A temperature change image in which a portion where a sudden temperature change occurs locally is emphasized as an edge is generated, the amount of temperature change in the temperature change image is made absolute, and an edge image is generated, and the data of the second thermal image is used. On the other hand, the data obtained by multiplying the data of the edge image by a predetermined coefficient is added to generate a third thermal image, and the temperature change region is extracted based on the predetermined threshold value in the third thermal image. It is characterized by being provided with a diagnostic device.
Concrete structure diagnostic system. The diagnostic device generates a diagnostic image from which the temperature change region is extracted.
The concrete structure diagnostic system according to claim 1. When the temperature of the external environment of the concrete is rising, the diagnostic apparatus extracts a region equal to or higher than a predetermined threshold value in the third thermal image as a temperature changing region.
The concrete structure diagnostic system according to claim 1 or 2. When the temperature of the external environment of the concrete is lowered, the diagnostic apparatus extracts a region equal to or less than a predetermined threshold value in the third thermal image as a temperature changing region.
The concrete structure diagnostic system according to claim 1 or 2. The diagnostic apparatus extracts a temperature-changing region based on the temperature parameter value in the third thermal image, calculates an area ratio which is a ratio of the predetermined area to the extracted area of the temperature-changing region, and obtains the area ratio and the area ratio. The temperature value at the turning point of the area ratio is set as a threshold in relation to the temperature parameter value, and the temperature changing region extracted from the third thermal image based on the threshold is determined as the deformed portion.
The concrete structure diagnostic system according to any one of claims 1 to 4. The diagnostic apparatus is at least a temperature threshold determined by detecting the deformed portion in a predetermined ratio or more based on a plurality of the deformed portions as samples in the third thermal image. The region is extracted as the deformed portion,
The concrete structure diagnostic system according to any one of claims 1 to 4. The diagnostic device calculates the variance between a plurality of the thresholds determined by changing the position and time of the diagnosis target in the concrete structure and the difference between the average temperature value of the healthy region determined by the thresholds. A calibration curve that calibrates the threshold is calculated based on the relationship that appears in the scatter plot based on the variance and the difference, the third thermal image of another diagnostic target is generated, and a unique temperature is generated from the third thermal image. The temperature variance of the provisional region excluding the region having the above is calculated, and the temperature threshold in the other diagnostic object is set using the calibration curve.
The concrete structure diagnostic system according to claim 5 or 6. The imaging unit captures a visible image of a predetermined area of the concrete surface to be inspected.
The diagnostic apparatus performs machine learning to compare the deformed portion determined based on the third thermal image with the visible image in which the deformed portion corresponding to the third thermal image is marked, and the machine performs the machine learning. The threshold value is calculated based on the learning result, a new third thermal image is generated based on the new thermal image captured by the imaging unit, and the new third thermal image is generated based on the calculated threshold value. Extracting the deformed portion from
The concrete structure diagnostic system according to any one of claims 1 to 4. A computer is a method for diagnosing deformed parts of concrete structures.
A process of generating a first thermal image in which noise is removed from a thermal image showing the temperature distribution of a predetermined area of a concrete surface to be inspected, and
A step of generating a second thermal image in which a range in which the temperature gradient generated in the first thermal image is inhomogeneous is averaged, and a step of generating the second thermal image.
A step of generating a temperature change image in which a portion of the second thermal image in which a sudden temperature change occurs locally is emphasized as an edge, and a step of generating the temperature change image.
A process of generating an edge image by making the amount of temperature change in the temperature change image absolute and
A step of generating a third thermal image by adding data obtained by multiplying the data of the edge image by a predetermined coefficient with respect to the data of the second thermal image.
It is characterized by comprising a step of extracting a temperature change region based on a predetermined threshold value in the third thermal image.
Concrete structure diagnostic method. A program applied to a concrete structure diagnostic system that diagnoses deformed parts of a concrete structure, and is applied to a computer.
A first thermal image in which noise is removed from a thermal image showing the temperature distribution of a predetermined area of the concrete surface to be inspected is generated.
A second thermal image is generated by averaging the range in which the temperature gradient generated in the first thermal image is inhomogeneous.
In the second thermal image, a temperature change image is generated in which a portion where a temperature change occurs locally and rapidly is emphasized as an edge.
The amount of temperature change in the temperature change image is made absolute and an edge image is generated.
Data obtained by multiplying the data of the edge image by a predetermined coefficient is added to the data of the second thermal image to generate a third thermal image.
The third thermal image is characterized in that a temperature change region is extracted based on a predetermined threshold value.
program.

Images