Damage inspection method

Abstract

Provided is a damage inspection method capable of accurately detecting a damaged part. 【Solution means】 A step ST1 of acquiring a plurality of thermal images by sequentially imaging while relatively moving an infrared imaging device with respect to an object to be inspected; a step ST2 of extracting a temperature profile for each of the plurality of thermal images; a step ST3 of calculating an average temperature profile by averaging the temperature profiles; a step ST4 of calculating a temperature difference dT between adjacent evaluation points based on the average temperature profile; a step ST5 of specifying an evaluation point whose absolute value of the temperature difference dT exceeds a threshold value Ra as a damage part candidate Pa; a step ST6 of specifying an evaluation point whose absolute value of the temperature difference dT is less than or equal to a threshold value Rb1 as a healthy part; and a step ST7 of specifying a damage part candidate Pa whose absolute value of the difference between the average temperature corresponding to the damage part candidate Pa and the average value Tp of the healthy part temperatures exceeds a predetermined threshold value Rb2 as a damage part.

Description

Technical Field

[0001] The present invention relates to a damage inspection method capable of detecting damage sites such as cracks existing in an object to be inspected by using a plurality of thermal images obtained by sequentially imaging the object to be inspected while relatively moving an infrared imaging device with respect to the object to be inspected. In particular, the present invention relates to a damage inspection method that can accurately detect damage sites existing in an object to be inspected without the constraint of forcibly causing stress fluctuations in the object to be inspected.

Background Art

[0002] Conventionally, as a method for detecting damage sites such as cracks existing in an object to be inspected by using a plurality of thermal images obtained by sequentially imaging the object to be inspected while relatively moving an infrared imaging device with respect to the object to be inspected, for example, the method described in Patent Document 1 has been proposed.

[0003] The method described in Patent Document 1 is a method for detecting defects in a structure (girder) on which a moving body (trolley) that generates a moving load travels. An infrared imaging device is installed on the moving body, and the infrared imaging device images the structure in which stress fluctuations are generated due to the travel of the moving body, measures the temperature distribution fluctuation on the surface of the structure as a thermal image, and thereby detects defects existing in the structure. That is, in the method described in Patent Document 1, forcibly causing stress fluctuations in the structure that is the object to be inspected while the infrared imaging device is moving is a prerequisite for detecting defects that are damage sites. For this reason, there is a problem that the object to be inspected is limited to a cargo handling machine or the like, which is a structure on which a moving body that generates a moving load travels.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] The present invention has been made in view of the problems of the above prior art, and provides a damage inspection method that can accurately detect damage sites existing in an object to be inspected without the restriction of forcibly causing stress fluctuations in the object to be inspected.

Means for Solving the Problems

[0006] In order to solve the above problems, the present inventors have intensively studied a damage inspection method capable of detecting damage sites such as cracks existing in an object to be inspected by using a plurality of thermal images obtained by sequentially imaging the object to be inspected while relatively moving an infrared imaging device with respect to the object to be inspected. Specifically, for each of the plurality of thermal images, the present inventors extracted a temperature profile, which is the temperature distribution on an evaluation line segment in the thermal image corresponding to the same virtual line segment of the object to be inspected, and considered calculating an average temperature profile, which is the average temperature distribution on the evaluation line segment, by averaging the temperature profiles extracted for each of the plurality of thermal images. Then, based on the average temperature profile, the present inventors considered calculating the temperature difference dT between adjacent evaluation points located on the evaluation line segment. It has been found that when the object to be inspected is subjected to heat from sunlight or ambient environmental heat, the absolute value of the temperature difference dT becomes large at damage sites such as cracks due to the heat insulation effect of the voids. In addition, temperature fluctuations caused by minute unevenness and the like other than the damage sites of the object to be inspected occur in the temperature profile, which can become noise in damage site detection. However, it has been found that this noise can be reduced by using the average temperature profile obtained by averaging the temperature profiles (the absolute value of the temperature difference dT caused by minute unevenness and the like is likely to be sufficiently smaller than the absolute value of the temperature difference dT at the damage site).

[0007] The present invention has been completed based on the findings of the present inventors described above. That is, in order to solve the above problems, the present invention includes a thermal image acquisition step of acquiring a plurality of thermal images showing the temperature distribution of the inspection target by sequentially imaging the inspection target while relatively moving an infrared imaging device with respect to the inspection target; a temperature profile extraction step of extracting, for each of the plurality of thermal images, a temperature profile that is the temperature distribution on an evaluation line segment in the thermal image corresponding to the same virtual line segment of the inspection target; an average temperature profile calculation step of calculating an average temperature profile that is the average temperature distribution on the evaluation line segment by averaging the temperature profiles extracted for each of the plurality of thermal images; a temperature difference calculation step of calculating a temperature difference dT between adjacent evaluation points located on the evaluation line segment based on the average temperature profile; a damaged part candidate identification step of identifying, when the absolute value of the temperature difference dT exceeds a predetermined threshold value Ra, the evaluation point corresponding to the temperature difference dT as a damaged part candidate Pa; a healthy part identification step of identifying, when the absolute value of the temperature difference dT is less than or equal to a predetermined threshold value Rb1 smaller than the threshold value Ra, the evaluation point corresponding to the temperature difference dT as a healthy part; an average value of the average temperatures corresponding to the healthy parts in the average temperature profile is defined as a healthy part temperature average value Tp, and a damaged part identification step of identifying the damaged part candidate Pa as a damaged part when the absolute value of the difference between the average temperature corresponding to the damaged part candidate Pa in the average temperature profile and the healthy part temperature average value Tp exceeds a predetermined threshold value Rb2. A damage inspection method is provided.

[0008] In the present invention, the "evaluation point corresponding to the temperature difference dT" means an evaluation point located on either one of the predetermined sides among the adjacent evaluation points for which the temperature difference dT is calculated. According to the damage inspection method of the present invention, in the thermal image acquisition step and the temperature profile extraction step, for each of a plurality of thermal images obtained by sequentially imaging the object to be inspected while relatively moving the infrared imaging device with respect to the object to be inspected, a temperature profile on the evaluation line segment in the thermal image corresponding to the same virtual line segment of the object to be inspected is extracted. For example, when the infrared imaging device is relatively moved in one direction with respect to the object to be inspected, the evaluation line segment in the thermal image corresponding to the same virtual line segment of the object to be inspected will move in the reverse direction of the one direction in each sequentially acquired thermal image. Since the movement amount of the evaluation line segment between each thermal image is determined by the relative movement speed, frame rate, and imaging resolution of the infrared imaging device, these values can be used to calculate the position of the evaluation line segment of each thermal image corresponding to the same virtual line segment of the object to be inspected, and the temperature profile on the calculated evaluation line segment can be extracted. In the thermal image acquisition step of the damage inspection method according to the present invention, it is only necessary to relatively move the infrared imaging device with respect to the object to be inspected, and there is no constraint of forcibly causing stress fluctuations in the object to be inspected.

[0009] And according to the damage inspection method of the present invention, in the average temperature profile calculation step, by averaging the temperature profiles extracted for each of the plurality of thermal images, an average temperature profile, which is the average temperature distribution on the evaluation line segment, is calculated. By using this average temperature profile, the damage inspection method according to the present invention can reduce noise caused by minute unevenness or the like other than the damaged part of the object to be inspected, as described above.

[0010] Also, according to the damage inspection method of the present invention, in the temperature difference calculation step, based on the average temperature profile, the temperature difference dT between adjacent evaluation points located on the evaluation line segment is calculated. By the magnitude of the absolute value of this temperature difference dT, the damage inspection method according to the present invention can detect damaged parts such as cracks, as described above. However, at evaluation points where the absolute value of the temperature difference dT is large, there may be elements other than the damaged part. Therefore, in order to accurately detect the damaged part, the damage inspection method according to the present invention, instead of immediately identifying an evaluation point corresponding to the temperature difference dT as the damaged part when the absolute value of the temperature difference dT exceeds a predetermined threshold value Ra in the damaged part candidate identification step, identifies it as a damaged part candidate Pa. Next, in the healthy part identification step, when the absolute value of the temperature difference dT is less than or equal to a predetermined threshold value Rb1 smaller than the threshold value Ra, an evaluation point corresponding to the temperature difference dT is identified as a healthy part. Finally, in the damaged part identification step, the average value of the average temperatures corresponding to the healthy parts in the average temperature profile is defined as the healthy part temperature average value Tp, and only when the absolute value of the difference between the average temperature corresponding to the damaged part candidate Pa in the average temperature profile and the healthy part temperature average value Tp exceeds a predetermined threshold value Rb2, the damaged part candidate Pa is identified as the damaged part. Thereby, compared with the case of immediately identifying an evaluation point corresponding to a temperature difference dT exceeding a predetermined threshold value Ra as the damaged part, it is possible to accurately identify (detect) the damaged part. As described above, according to the damage inspection method of the present invention, there is no restriction on forcibly causing stress fluctuations in the object to be inspected, and it is possible to accurately detect the damaged part existing in the object to be inspected.

[0011] In the damage inspection method according to the present invention, when it is possible to predict in advance a part where damage is likely to exist in the object to be inspected, it is conceivable to set one virtual line segment so as to pass through this part (therefore, one evaluation line segment is also set for each thermal image). However, just setting one virtual line segment (setting one evaluation line segment) may miss the damaged part in cases where the above prediction is difficult or where damage exists in a part slightly deviated from the above predicted part. In order to reduce the risk of missing the damaged part, it is preferable to set a plurality of virtual line segments (set a plurality of evaluation line segments).

[0012] That is, in the damage inspection method according to the present invention, in the temperature profile extraction step, for each of the plurality of evaluation line segments in the thermal image corresponding to the plurality of virtual line segments of the object to be inspected, the temperature profile is extracted. In the average temperature profile calculation step, for each of the plurality of evaluation line segments, the average temperature profile is calculated. In the temperature difference calculation step, for each of the plurality of evaluation line segments, the temperature difference dT is calculated. In the damage site candidate identification step, for each of the plurality of evaluation line segments, it is determined whether there is a temperature difference dT whose absolute value exceeds the threshold value Ra. In the damage site candidate identification step, for the evaluation line segment where the temperature difference dT whose absolute value exceeds the threshold value Ra exists, after identifying the evaluation point corresponding to the temperature difference dT as the damage site candidate Pa, the healthy site identification step and the damage site identification step are executed. In the damage site candidate identification step, for the evaluation line segment where the temperature difference dT whose absolute value exceeds the threshold value Ra does not exist, after determining that there is no damage site, it is preferable not to execute the healthy site identification step and the damage site identification step.

[0013] According to the above preferred method, for each of the plurality of evaluation line segments in the thermal image corresponding to the plurality of virtual line segments of the object to be inspected, the temperature profile extraction step, the average temperature profile calculation step, the temperature difference calculation step, and the damage site candidate identification step are executed. Then, for the evaluation line segment where the damage site candidate is identified (the evaluation line segment where the temperature difference dT whose absolute value exceeds the threshold value Ra exists), as described above, the healthy site identification step and the damage site identification step are executed. On the other hand, for the evaluation line segment where the damage site candidate is not identified (the evaluation line segment where the temperature difference dT whose absolute value exceeds the threshold value Ra does not exist), since there is no meaning in executing the healthy site identification step and the damage site identification step, after determining that there is no damage site, the healthy site identification step and the damage site identification step are not executed. According to the above preferred method, since a plurality of virtual line segments (a plurality of evaluation line segments) are set, the possibility of overlooking the damage site can be reduced.

Advantages of the Invention

[0014] According to the present invention, there is no constraint of forcibly causing stress fluctuations in the object to be inspected, and damage sites existing in the object to be inspected can be accurately detected.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Mode for Carrying Out the Invention

[0016] Hereinafter, a damage inspection method according to an embodiment of the present invention will be described with appropriate reference to the accompanying drawings. FIG. 1 is a flowchart schematically showing the steps of the damage inspection method according to the present embodiment. As shown in FIG. 1, the damage inspection method according to the present embodiment includes a thermal image acquisition step ST1, a temperature profile extraction step ST2, an average temperature profile calculation step ST3, a temperature difference calculation step ST4, a damage site candidate identification step ST5, a healthy site identification step ST6, and a damage site identification step ST7. Hereinafter, each step will be described in order.

[0017] <Thermal Image Acquisition Step ST1> FIG. 2 is a diagram schematically explaining the outline of the thermal image acquisition step ST1. FIG. 2(a) is a diagram schematically showing the outline of the apparatus configuration for executing the thermal image acquisition step ST1. FIG. 2(b) is a diagram schematically showing the thermal images sequentially acquired in the thermal image acquisition step ST1. In the thermal image acquisition step ST1, as shown in FIG. 2(a), while relatively moving an infrared imaging device with respect to an object to be inspected (in the example shown in FIG. 2(a), a welded joint), the object to be inspected is sequentially imaged to acquire a plurality of thermal images showing the temperature distribution of the object to be inspected. The plurality of acquired thermal images are input to the image processing device shown in FIG. 2(a). The image processing device is composed of, for example, a computer installed with predetermined image processing software or programs and capable of executing image processing and various arithmetic processes. In the example shown in FIG. 2(a), while the object to be inspected is stationary, the infrared imaging device is placed on a mobile cart and moved in one direction (X direction) with respect to the object to be inspected. However, the present invention is not limited to this, and it is also possible to move the object to be inspected while the infrared imaging device is stationary, or to sequentially image the object to be inspected while moving both the infrared imaging device and the object to be inspected (for example, moving in the opposite direction).

[0018] The relative moving speed of the infrared imaging device is preferably up to 1000 mm / sec, although it also depends on the size of the inspection area of the object to be inspected. Further, the frame rate of the infrared imaging device is preferably 3 frames / sec or more, more preferably 10 frames / sec or more. Furthermore, the imaging resolution of the infrared imaging device is preferably 5.0 mm / pixel or less. Note that when executing the thermal image acquisition step ST1, it is not necessary to apply an artificial thermal load to the object to be inspected, and it is sufficient if heat due to sunlight or ambient environmental heat is applied. However, for example, it is also possible to apply a thermal load using a halogen lamp or the like.

[0019] Here, as shown in FIG. 2(a), consider setting two virtual line segments (line A and line B) in the object to be inspected. The line A shown in FIG. 2(a) is a virtual line segment passing through a crack which is a damaged part, and the line B is a virtual line segment not passing through the crack which is the damaged part. As shown in FIG. 2(b), consider sequentially acquiring a total of n thermal images in the thermal image acquisition step ST1. Then, as shown in FIG. 2(b), in the first thermal image, the evaluation line segments corresponding to the virtual line segments line A and line B are line A1 and line B1, respectively, and in the m(1 < m < n)-th thermal image, the evaluation line segments corresponding to the virtual line segments line A and line B are line A m , line B m respectively, and in the n-th thermal image, the evaluation line segments corresponding to the virtual line segments line A and line B are line A n , line B n respectively. Then, the evaluation line segments in each thermal image will move in the reverse direction of the one direction (X direction). And if a plurality of thermal images are continuously acquired according to the frame rate, the amount of movement of the evaluation line segments between the acquired thermal images is determined by the relative moving speed, frame rate, and imaging resolution of the infrared imaging device.

[0020] FIG. 3 shows an example of the first thermal image obtained in the thermal image acquisition step ST1. The thermal image shown in FIG. 3 was obtained under the following inspection conditions. [Inspection Conditions] (1) Object to be inspected: A lap joint obtained by arc-welding two steel plates with a thickness of 2.6 mm as shown in FIG. 2(a), where one of the steel plates has a crack as a damaged part. (2) Relative movement speed of the infrared imaging device: As shown in FIG. 2(a), the infrared imaging device was placed on a mobile cart and moved in the X direction at a relative movement speed of 12 mm / sec. (3) Frame rate of the infrared imaging device: 10 frames / sec (4) Imaging resolution of the infrared imaging device: 0.6 mm / pixel (5) Imaging time of the infrared imaging device: 1 sec (continuously imaging according to the frame rate) According to the above inspection conditions, between each of the acquired thermal images, the infrared imaging device moves 1.2 mm (12 [mm / sec] / 10 [frames / sec] = 1.2 [mm / frame]), so the movement amount of the evaluation line segment between each thermal image is 1.2 mm (= equivalent to 2 pixels).

[0021] <Temperature Profile Extraction Step ST2> In the temperature profile extraction step ST2, using the image processing device shown in FIG. 1, for each of a plurality of thermal images as shown in FIG. 2(b) and FIG. 3, a temperature profile, which is the temperature distribution on the evaluation line segment in the thermal image corresponding to the same virtual line segment of the object to be inspected, is extracted. Specifically, in this embodiment, using an image processing apparatus, for the first thermal image, evaluation line segments (line A1, line B1) corresponding to the virtual line segments (line A, line B) of the object to be inspected are set, and a temperature profile, which is the temperature distribution on this evaluation line segment, is extracted. Next, for each thermal image after the second one, first, a temperature profile on a line segment at the same position as the evaluation line segments (line A1, line B1) of the first thermal image (the line segment whose coordinates in each thermal image are the same as those of the evaluation line segments of the first thermal image) is extracted. The temperature profiles extracted in this way are not profiles corresponding to the positions of the virtual line segments (line A, line B) of the object to be inspected for each thermal image after the second one, but are temperature profiles on a line segment shifted in the relative movement direction (X direction) of the infrared imaging device. Therefore, by correcting this shift (performing position adjustment), for the thermal images after the second one as well, a temperature profile on an evaluation line segment corresponding to the virtual line segments (line A, line B) of the object to be inspected is extracted. First, the content to be executed for the evaluation line segment corresponding to line A will be described below, and then the content to be executed for the evaluation line segment corresponding to line B will be described.

[0022] FIG. 4 is a diagram showing an example of a temperature profile on an evaluation line segment corresponding to line A, which is extracted in the temperature profile extraction step ST2 under the above-described inspection conditions. FIG. 4(a) shows the temperature profile before correcting the positional deviation extracted for the 1st to 3rd thermal images, and FIG. 4(b) shows the temperature profile after correcting the positional deviation extracted for the 1st to 3rd thermal images. Since the horizontal axis in FIG. 4(a) is before correcting the positional deviation, for any temperature profile, it is the position on the evaluation line segment (line A1) of the 1st thermal image (the distance from the left end of line A1 shown in FIG. 2(b) corresponding to the left end of line A shown in FIG. 2(a)). Since the horizontal axis in FIG. 4(b) is after correcting the positional deviation, for any temperature profile, it is the position on the evaluation line segment corresponding to line A (line A1 for the 1st thermal image, line A2 for the 2nd thermal image, and line A3 for the 3rd thermal image) (the distance from the left end of each of lines A1 to A3). As can be seen by comparing FIG. 4(a) and FIG. 4(b), for the 1st thermal image, the temperature profile does not change. For the 2nd thermal image, the temperature profile is positionally adjusted 1.2 mm (equivalent to 2 pixels) to the left (X direction) before and after correction. For the 3rd thermal image, the temperature profile is positionally adjusted 2.4 mm (equivalent to 4 pixels) to the left (X direction) before and after correction. In this embodiment, for each thermal image after the 2nd one, first, the temperature profile on a line segment at the same position as the evaluation line segment of the 1st thermal image is extracted, and then the mode of correcting the positional deviation has been described. However, the present invention is not limited to this. Since the amount of movement of the evaluation line segment between each thermal image can be calculated in advance based on the relative movement speed, frame rate, and imaging resolution of the infrared imaging device, it is also possible to adopt a mode in which the positions (coordinates in each thermal image) of the evaluation line segments of each thermal image corresponding to the virtual line segment of the object to be inspected are obtained first, and the temperature profile on the obtained evaluation line segment is directly extracted.

[0023] <Average temperature profile calculation step ST3> In the average temperature profile calculation step ST3, the average temperature T on the evaluation line segment is calculated by averaging the temperature profiles extracted for each of the plurality of thermal images as shown in FIG. 4(b) using the image processing apparatus shown in FIG. 1. ave is the distribution of the average temperature profile. FIG. 5 is a diagram showing an example of the average temperature profile on the evaluation line segment corresponding to line A, which is calculated in the average temperature profile calculation step ST3 in the case of the above-described inspection conditions. In FIG. 4(b), only the temperature profiles extracted for the first to third thermal images are shown, but in the above-described inspection conditions, actually, 10 thermal images (10 [frames / sec] × 1 [sec] = 10 [frames]) are acquired in 1 second of the imaging time of the infrared imaging device. Therefore, in the case of the above-described inspection conditions, in the average temperature profile calculation step ST3, the average temperature profile is calculated by averaging the temperature profiles extracted for each of the 10 thermal images.

[0024] <Temperature difference calculation step ST4> In the temperature difference calculation step ST4, using the image processing apparatus shown in FIG. 1, the temperature difference dT between adjacent evaluation points located on the evaluation line segment is calculated based on the average temperature profile as shown in FIG. 5. The evaluation points are, for example, the pixels constituting the thermal image, and in this case, the adjacent evaluation points are adjacent pixels. In the present embodiment, the pixels are used as the evaluation points.

[0025] <Damage site candidate identification step ST5> In the damage site candidate identification step ST5, using the image processing apparatus shown in FIG. 1, when the absolute value of the temperature difference dT exceeds a predetermined threshold value Ra, the evaluation point corresponding to the temperature difference dT is identified as the damage site candidate Pa. Specifically, it is determined whether or not there is a temperature difference dT whose absolute value exceeds a predetermined threshold value Ra (step ST51 in FIG. 1). When there is a temperature difference dT whose absolute value exceeds the threshold value Ra (when “Yes” in step ST51 in FIG. 1), the evaluation point corresponding to the temperature difference dT is identified as the damage site candidate Pa (step ST52 in FIG. 1). The threshold value Ra is determined, for example, by the following formula (1). Ra [°C] = 0.1 × L [mm] ··· (1) In the above formula (1), L means the distance between adjacent evaluation points where the temperature difference dT is calculated. In the case of the above-described inspection conditions, since the imaging resolution of the infrared imaging device is 0.6 mm / pixel and the pixel is used as the evaluation point, L = 0.6 mm, and from the above formula (1), Ra = 0.06 °C.

[0026] FIG. 6 is a diagram showing an example of the results of executing the temperature difference calculation step ST4, the damage site candidate identification step ST5, and the healthy site identification step ST6 (described later) for the evaluation line segment corresponding to line A in the case of the above-described inspection conditions. In FIG. 6, the average temperature T calculated in the average temperature profile calculation step ST3 ave is plotted with "◆", and the absolute value of the temperature difference dT calculated in the temperature difference calculation step ST4 is plotted with "+", "○", and "●". The data plotted with "◆" is the same as that shown in FIG. 5. As shown in FIG. 6, since the absolute value of the temperature difference dT plotted with "+" exceeds the threshold value Ra, the six evaluation points corresponding to the temperature difference dT plotted with "+" are to be identified as damage site candidates Pa.

[0027] <Healthy Site Identification Step ST6> In the healthy site identification step ST6, when the absolute value of the temperature difference dT is equal to or less than a predetermined threshold value Rb1 smaller than the threshold value Ra, the evaluation point corresponding to the temperature difference dT is identified as a healthy site using the image processing device shown in FIG. 1. The threshold value Rb1 is determined, for example, by the following formula (2). Rb1 [°C] = 0.067 × L [mm] ··· (2) In the above formula (2), L means the distance between adjacent evaluation points where the temperature difference dT is calculated. In the case of the above-described inspection conditions, since the imaging resolution of the infrared imaging device is 0.6 mm / pixel and the pixel is used as the evaluation point, L = 0.6 mm, and from the above formula (2), Rb1 = 0.04 °C.

[0028] As shown in FIG. 6, since the absolute value of the temperature difference dT plotted with "○" is equal to or less than the threshold value Rb1, 19 evaluation points corresponding to the temperature difference dT plotted with "○" are specified as healthy sites.

[0029] <Damage site identification step ST7> In the damage site identification step ST7, using the image processing apparatus shown in FIG. 1, the average value of the average temperatures corresponding to the healthy sites in the average temperature profile is calculated as the healthy site temperature average value Tp. Then, when the absolute value of the difference between the average temperature corresponding to the damage site candidate Pa in the average temperature profile and the healthy site temperature average value Tp exceeds a predetermined threshold value Rb2, the damage site candidate Pa is identified as the damage site. The threshold value Rb2 is determined, for example, by the following formula (3). Rb1 [°C] = 0.417 × L [mm] ···(3) In the above formula (3), L means the distance between adjacent evaluation points at which the temperature difference dT is calculated. In the case of the above-described inspection conditions, the imaging resolution of the infrared imaging apparatus is 0.6 mm / pixel, and since the pixel is used as the evaluation point, L = 0.6 mm, and from the above formula (3), Rb2 = 0.25 °C.

[0030] FIG. 7 is a diagram showing an example of the result of executing the damage site identification step ST7 for the evaluation line segment corresponding to line A in the case of the above-described inspection conditions. In FIG. 7, the average temperature T calculated in the average temperature profile calculation step ST3 ave is plotted with "×" and "◆", and the plotted data is the same as that shown in FIG. 5. Among these, the data plotted with "×" is the data corresponding to the six damage site candidates Pa identified in the damage site candidate identification step ST5. As shown in FIG. 7, among the data corresponding to the six damage site candidates Pa plotted with "×", the average temperature T corresponding to the damage site candidate Pa aveand there is one whose absolute value of the difference from the average temperature value Tp of the healthy part exceeds a predetermined threshold value Rb2. Therefore, this damage site candidate Pa is to be specified as the damage site. It was found that the specified damage site coincides with the position of the crack (see Fig. 2(a)) actually existing in the object to be inspected.

[0031] In the above description, the content to be executed for the evaluation line segment corresponding to line A was explained. Next, the content to be executed for the evaluation line segment corresponding to line B will be explained.

[0032] For the evaluation line segment corresponding to line B as well, similar to the evaluation line segment corresponding to line A, the temperature profile extraction step ST2, the average temperature profile calculation step ST3, the temperature difference calculation step ST4, and the damage site candidate specification step ST5 are executed. Fig. 8 is a diagram showing an example of the temperature profile on the evaluation line segment corresponding to line B extracted in the temperature profile extraction step ST2 in the case of the above-described inspection conditions. Fig. 8(a) shows the temperature profile before correcting the displacement extracted for the 1st to 3rd thermal images, and Fig. 8(b) shows the temperature profile after correcting the displacement extracted for the 1st to 3rd thermal images. Since the horizontal axis in Fig. 8(a) is before correcting the displacement, for any temperature profile, it is the position on the evaluation line segment (line B1) of the 1st thermal image (the distance from the left end of line B1 shown in Fig. 2(b) corresponding to the left end of line B shown in Fig. 2(a)). Since the horizontal axis in Fig. 8(b) is after correcting the displacement, for any temperature profile, it is the position on the evaluation line segment corresponding to line B (line B1 for the 1st thermal image, line B2 for the 2nd thermal image, and line B3 for the 3rd thermal image) (the distance from the left end of each of lines B1 to B3). As can be seen by comparing Fig. 8(a) and Fig. 8(b), for the 1st thermal image, the temperature profile does not change. For the 2nd thermal image, the temperature profile is position-adjusted 1.2 mm (= equivalent to 2 pixels) to the left (X direction) before and after correction. For the 3rd thermal image, the temperature profile is position-adjusted 2.4 mm (= equivalent to 4 pixels) to the left (X direction) before and after correction.

[0033] FIG. 9 is a diagram showing an example of the average temperature profile on the evaluation line segment corresponding to line B, which is calculated in the average temperature profile calculation step ST3 under the above-described inspection conditions. In FIG. 8(b), only the temperature profiles extracted from the first to third thermal images are shown. However, under the above-described inspection conditions, actually, 10 thermal images (10 [frames / sec] × 1 [sec] = 10 [frames]) are acquired in 1 second of the imaging time of the infrared imaging device. Therefore, under the above-described inspection conditions, in the average temperature profile calculation step ST3, the average temperature profile is calculated by averaging the temperature profiles extracted for every 10 thermal images.

[0034] Regarding the evaluation line segment corresponding to line B, in the temperature difference calculation step ST4, using the image processing apparatus shown in FIG. 1, the temperature difference dT between adjacent evaluation points located on the evaluation line segment is calculated based on the average temperature profile as shown in FIG. 9. The evaluation points are, for example, the pixels constituting the thermal image, and in this case, the adjacent evaluation points are adjacent pixels. In the present embodiment, the pixels are used as the evaluation points.

[0035] FIG. 10 is a diagram showing an example of the results of executing the temperature difference calculation step ST4 and the damage site candidate identification step ST5 for the evaluation line segment corresponding to line B under the above-described inspection conditions. In FIG. 10, the average temperature T calculated in the average temperature profile calculation step ST3 ave is plotted with "◆", and the absolute value of the temperature difference dT calculated in the temperature difference calculation step ST4 is plotted with "○". The data plotted with "◆" is the same as that shown in FIG. 9. As shown in Fig. 10, the absolute values of the temperature differences dT plotted with "○" do not exceed the threshold value Ra. That is, for the evaluation line segment corresponding to line B, since there is no temperature difference dT whose absolute value exceeds the threshold value Ra (it is "No" in step ST51 of Fig. 1), it is determined that there is no damaged part for the evaluation line segment corresponding to line B (step ST53 of Fig. 1). And, unlike the evaluation line segment corresponding to line A, the inspection is terminated without executing the healthy part identification step ST6 and the damaged part identification step ST7. It has been found that there is actually no damage such as a crack in line B of the inspection object for which it has been determined that there is no damaged part.

[0036] As described above, according to the damage inspection method according to the present embodiment, there is no restriction of forcibly causing stress fluctuations in the inspection object, and the damaged part existing in the inspection object can be accurately detected. In the present embodiment, the case where two virtual line segments (line A, line B) are set in the inspection object has been described as an example, but the present invention is not limited to this, and it is similarly applicable to the case where only one virtual line segment or three or more virtual line segments are set. Also, in the present embodiment, the case where the inspection object is a welded joint (see Fig. 2(a)) has been described as an example, but the inspection object to which the present invention is applicable is not limited to this. The present invention is applicable to a wide range of inspection objects such as automobile structural members other than welded joints, home appliances, materials such as slabs, crane runway girders and belt conveyors in factory buildings, walls and roofs of buildings, and social infrastructure such as highways.

Explanation of Reference Numerals

[0037] ST1 ··· Thermal image acquisition step ST2 ··· Temperature profile extraction step ST3 ··· Average temperature profile calculation step ST4 ··· Temperature difference calculation step ST5 ··· Damaged part candidate identification step ST6 ··· Healthy part identification step ST7 ··· Injury site identification step

Claims

1. A thermal image acquisition step of acquiring a plurality of thermal images showing the temperature distribution of the object to be inspected by sequentially imaging the object to be inspected while relatively moving an infrared imaging device with respect to the object to be inspected; A temperature profile extraction step of extracting, for each of the plurality of thermal images, a temperature profile that is a temperature distribution on an evaluation line segment in the thermal image corresponding to the same virtual line segment of the object to be inspected; An average temperature profile calculation step of calculating an average temperature profile that is an average temperature distribution on the evaluation line segment by averaging the temperature profiles extracted for each of the plurality of thermal images; A temperature difference calculation step of calculating a temperature difference dT between adjacent evaluation points located on the evaluation line segment based on the average temperature profile; A damaged part candidate identification step of identifying, when the absolute value of the temperature difference dT exceeds a predetermined threshold Ra, the evaluation point corresponding to the temperature difference dT as a damaged part candidate Pa; A healthy part identification step of identifying, when the absolute value of the temperature difference dT is equal to or less than a predetermined threshold Rb1 smaller than the threshold Ra, the evaluation point corresponding to the temperature difference dT as a healthy part; A damaged part identification step of identifying the damaged part candidate Pa as a damaged part when the absolute value of the difference between the average temperature corresponding to the damaged part candidate Pa in the average temperature profile and the average value of the average temperatures corresponding to the healthy parts in the average temperature profile exceeds a predetermined threshold Rb2. A damage inspection method.

2. In the temperature profile extraction step, the temperature profile is extracted for each of the plurality of evaluation line segments in the thermal image corresponding to the plurality of virtual line segments of the object to be inspected, In the average temperature profile calculation step, the average temperature profile is calculated for each of the plurality of evaluation line segments, In the temperature difference calculation step, the temperature difference dT is calculated for each of the plurality of evaluation line segments, In the damaged part candidate identification step, for each of the plurality of evaluation line segments, it is determined whether or not there is a temperature difference dT whose absolute value exceeds the threshold Ra. In the damaged part candidate specifying step, for the evaluation line segment where the temperature difference dT whose absolute value exceeds the threshold value Ra exists, after specifying the evaluation point corresponding to the temperature difference dT as the damaged part candidate Pa, the healthy part specifying step and the damaged part specifying step are executed. In the damaged part candidate specifying step, for the evaluation line segment where the temperature difference dT whose absolute value does not exceed the threshold value Ra does not exist, after determining that no damaged part exists, the healthy part specifying step and the damaged part specifying step are not executed. The damage inspection method according to claim 1.

Images