JP2016099296A - Flaw detection system for detecting internal defect of flaw detection object and flaw detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a defect even when a flaw detection object moves, in a flaw detection system that detects a defect, such as a flow or peeling, existing inside a flaw detection object by heating the flaw detection object and measuring a surface temperature.SOLUTION: A flaw detection system heats the surface of a flaw detection object S2; acquires thermal images in time series S3 by imaging the heated flaw detection object moving within a visual field; acquires an offset indicating a relative movement amount of the flaw detection object S4; acquires an overlapped image by overlapping the thermal images in a state of moving the flaw detection object in a movement direction by a movement amount defined based on the offset S5; displays the overlapped image S6; and detects a defect existing inside the flaw detection object on the basis of the overlapped image S8-S10.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a flaw detection technique for detecting defects such as scratches and peeling occurring inside a flaw detection object by measuring the temperature of the flaw detection object.

  An infrared thermography method is one of the methods for detecting defects generated inside the inspection object by nondestructive inspection. In this infrared thermography method, the surface of a test object is heated by a heat source, and the subsequent surface temperature is measured by an infrared camera or the like. When there is a defect inside, the flow of heat propagating to the inside changes, so that a local temperature change occurs on the surface. By detecting this temperature change, it is possible to detect a defect occurring inside the flaw detection object.

  For example, in the flaw detection method described in Patent Document 1, a defect of a flaw detection object (inspection object) is detected and displayed by the following procedure. First, the surface of the flaw detection target is pulse-heated, and the surface temperatures of a plurality of portions of the flaw detection target after heating are detected at a set sampling frequency for a set time. As a result, data indicating the relationship between the elapsed time after heating and the surface temperature is acquired for each cell (for each pixel). Although it is possible to detect and inspect defects from changes in the thermal image of the surface, the acquired data is further subjected to Fourier transform to convert it into data indicating the relationship between frequency and phase. A phase value at the set inspection frequency is obtained for the converted data. A phase image is generated from the phase value for each cell and displayed on the display unit. In this phase image, the phase value changes depending on the presence / absence of defects inside the flaw detection object, so that defects inside the flaw detection object can be detected and displayed based on the phase image. And it can be made into an image with fewer deep defects and less noise than a thermal image.

JP 2011-247718 A

  The flaw detection method described in Patent Document 1 targets a stationary flaw detection object. For this reason, when applied to a moving flaw detection object, there is a problem in that the phase image is disturbed and a defect cannot be detected. Similarly, when an imaging unit such as an infrared camera moves, there is a problem that a defect cannot be detected because the phase image is disturbed.

  The present invention has been made in view of such circumstances, and an object of the present invention is to detect an internal defect of a flaw detection object even when the flaw detection object relatively moves within the field of view of the imaging unit. There is.

  In order to achieve the above-described object, the flaw detection system according to the present invention captures a heating unit that heats the surface of the flaw detection target and the flaw detection target after heating that moves relative to the visual field at predetermined time intervals. A thermal image acquisition unit that acquires, in a time series, a thermal image in which cells indicating values corresponding to the amount of heat are arranged in a matrix, and a plurality of the thermal images, and the flaw detection object at the thermal image acquisition interval A superimposed image acquisition unit that obtains a superimposed image by superimposing in a state of being moved in the relative movement direction by a movement amount determined based on an offset indicating a relative movement amount, and a display for displaying the superimposed image Part.

  According to the present invention, a plurality of thermal images are moved and overlapped by a movement amount determined based on an offset indicating a relative movement amount of the flaw detection object. Thereby, it is possible to obtain a superimposed image in which flaw detection objects are superimposed. And since this superimposed image is used, even if it is a flaw detection object which moves relatively, the defect which exists in the surface or inside can be detected. In the present invention, the offset may be acquired automatically or may be set manually.

  In the flaw detection system described above, frequency-phase data indicating the relationship between frequency and phase is obtained by performing Fourier transform on the time-temperature data indicating the relationship between time and temperature in the superimposed image, and obtained from the frequency-phase data. It is preferable to have a flaw detection unit that detects a defect existing inside the flaw detection object based on the obtained phase image. In this flaw detection system, a defect existing inside the flaw detection object is detected based on the phase image obtained from the frequency-phase data, so that a defect at a deep location from the surface can also be detected.

  In the above-described flaw detection system, it is preferable that the flaw detection unit detects a defect present inside the flaw detection target based on a phase image based on an equiphase image at a set frequency. In this flaw detection system, defects at different depths from the surface can be detected by setting the frequency. This frequency can be arbitrarily determined. For example, it can be set to a frequency determined from a thermal image capturing time and a capturing time interval.

  Further, the flaw detection method according to the present invention has a heating step for heating the surface of the flaw detection target and a value corresponding to the amount of heat by photographing the flaw detection target after heating that moves relative to the visual field at predetermined time intervals. A thermal image acquisition step of acquiring, in time series, thermal images in which cells indicating the above are arranged in a matrix, and a plurality of the thermal images, an offset indicating a relative movement amount of the flaw detection object at the thermal image acquisition interval A superimposed image acquisition step of acquiring a superimposed image by superimposing in a state of being moved in the relative movement direction by a movement amount determined based on the display, and a display step of displaying the superimposed image. Features.

  According to the present invention, it is possible to detect an internal defect of a flaw detection object even when the flaw detection object relatively moves within the field of view of the imaging unit.

(A) is a schematic diagram showing an example of a flaw detection system. (B) is a figure explaining the memory | storage area | region provided in a memory | storage part. It is a top view explaining the visual field of an infrared camera, and the moving direction of a test object. (A) is a figure explaining the relationship between a thermal image (frame) and a cell (pixel). (B) is a figure which illustrates typically the movement of the flaw detection target object in a cell. It is a flowchart explaining the flaw detection process with respect to the flaw detection target to move. It is a figure explaining the heat processing of a test target. It is a figure explaining the acquisition process of a thermal image. It is a figure explaining the acquired thermal image (frame). It is a figure explaining the movement amount at the time of superimposition. It is a figure explaining the example of a superposition in case offset is 0.21 cell. It is a figure explaining the example of a superposition in case offset is 0.23 cell. It is a figure explaining the 20th thermal image, (a) is a comparative example, (b) is the example which performed the movement process. It is a figure explaining the 200th thermal image, (a) is a comparative example, (b) is the example which performed the movement process. It is a figure explaining the 400th thermal image, (a) is a comparative example, (b) is the example which performed the movement process. It is a figure explaining the 600th thermal image, (a) is a comparative example, (b) is the example which performed the movement process. It is a figure explaining a phase image, (a) is a comparative example, (b) is the example which performed the superimposition process. It is a figure which shows the internal defect of the test target in the phase image which performed the superimposition process.

  Embodiments of the present invention will be described below with reference to the drawings. A flaw detection system 10 shown in FIG. 1A includes an infrared camera 11, a xenon lamp 12, and a control device 13.

  The infrared camera 11 is a device for photographing a subject, detecting infrared energy emitted from the surface of the subject, and visualizing it as a thermal image. The thermal image data output from the infrared camera 11 is input to the control device 13. For this reason, the infrared camera 11 corresponds to a thermal image acquisition unit together with the control unit 16 of the control device 13.

  In the illustrated infrared camera 11, infrared energy within the field of view is collected by an optical system and input to the infrared detection element. The infrared detection element converts the input infrared energy into an electrical signal corresponding to the intensity and outputs the electrical signal. That is, the infrared detection element divides the field of view into a matrix (matrix), and outputs an electrical signal for each division unit (cell) as thermal image data. Here, the higher the temperature of the subject, the greater the infrared radiation emitted from the surface of the subject, and the greater the value (for example, voltage value) of the electrical signal output from the infrared detection element.

  As shown in the plan view of FIG. 2, the angle of view is θ, the distance to the subject is R, the field of view is X, the moving speed of the flaw detection object Z is V, and the number of cells (number of pixels) in the horizontal width direction in the field of view is N. Consider the case where the imaging frequency of the infrared camera 11 is F. In this case, the visual field X can be expressed by 2 × R × tan (θ / 2), and the visual field Δ per cell can be expressed by X / N. Further, the moving pitch p of the flaw detection object Z is a moving amount in one photographing interval and can be expressed by V / F. In addition, in this embodiment, the offset S is also calculated. Here, the offset S is a numerical value indicating the amount of movement of the flaw detection object Z in the thermal image acquisition interval in units of cell size, and is expressed by p / Δ (= V / F / Δ). .

  As shown in FIG. 3A, in the infrared camera 11, a thermal image FL in which the cells C are arranged in a matrix and the color corresponding to the temperature (= heat amount) is set for each cell C is obtained. Images can be taken at the imaging frequency F. In the infrared camera 11 of the present embodiment, for example, thermal images FL of 320 horizontal cells × 240 vertical cells (aspect ratio 4: 3) are captured at an interval of 50 images per second (imaging frequency 50 Hz; 0.02 second interval). You can get it. The number of cells C constituting the thermal image FL and the imaging frequency F of the infrared camera 11 can be set as appropriate. Further, as shown in FIG. 3B, when a certain point P of the flaw detection object Z moves at a speed V, the point P is the nth, n + 1th, n + 2th,... In the image FL, it moves by the moving pitch p.

  As shown in FIG. 1A, the xenon lamp 12 corresponds to a heating unit, and heats the surface of the flaw detection object Z by light emission. In this embodiment, as shown by the symbol P0 in FIG. 5, the flaw detection target Z is heated by the xenon lamp 12 at a heating position set upstream of the visual field X of the infrared camera 11.

  Since the light emitted by the xenon lamp 12 is for heating the surface of the flaw detection object Z, a light source other than the xenon lamp 12 may be used as the heating unit. The light emission mode of the xenon lamp 12 and other light sources may be flash light such as a flash, or may be continuous light that is continuously irradiated onto the flaw detection target Z. Further, the light emission control of the xenon lamp 12 may be performed by the control device 13 or manually.

  Returning to FIG. 1A, the control device 13 controls the operation of the infrared camera 11 and the xenon lamp 12, and processes and displays the data of the thermal image FL from the infrared camera 11, for example, a personal computer. Consists of. The control device 13 includes a control unit 16 having a CPU 14 and a storage unit 15, an input unit 17 configured with a mouse, a touch panel, a keyboard, and the like, and a display unit 18 configured with various displays.

  The CPU 14 reads computer programs stored in the storage unit 15 (the program storage area 15a and the flaw detection program storage area 15b), and operates according to these computer programs and operation signals from the input unit 17. By the operation of the CPU 14, the control unit 16 controls the operation of the infrared camera 11 and the xenon lamp 12, performs a calculation based on the data of the thermal image FL, and displays the calculation result on the display unit 18. The storage unit 15 is a part that stores a computer program and various data, and a storage element capable of writing and reading such as a hard disk drive, a solid state drive, a random access memory, and a flash memory is used.

  As shown in FIG. 1B, a partial area of the storage unit 15 includes a program storage area 15a, a flaw detection program storage area 15b, a parameter storage area 15c, a thermal image storage area 15d, a superimposed image storage area 15e, a time- It is used as a temperature data storage area 15f, a frequency-phase data storage area 15g, and a phase image storage area 15h.

  The program storage area 15a stores a computer program for causing the set of the infrared camera 11, the xenon lamp 12, and the control device 13 to function as the flaw detection system 10. The flaw detection program storage area 15b stores a computer program for causing the control unit 16 to perform automatic inspection and determination by referring to determination conditions based on a thermal image and a phase image described later.

  Various parameters such as determination conditions required in the flaw detection process are stored in the parameter storage area 15c. For example, the distance R to the subject, the visual field X, the moving speed V of the flaw detection object Z, the number N of cells in the width direction, the imaging frequency F, the visual field Δ per cell C, the moving pitch p of the flaw detection object Z, and the offset S Is memorized. Further, the resolution (number of cells C) at the time of photographing with the infrared camera 11, the operation content of the xenon lamp 12 (light emission time and intensity, heating condition of the flaw detection object Z) and the like are also stored in the parameter storage area 15c.

  In the thermal image storage area 15d, a time-series thermal image FL output from the infrared camera 11 is stored. In the superimposed image storage area 15e, a superimposed image obtained by overlaying the time-series thermal images FL while being shifted is stored. The time-temperature data storage area 15f stores data indicating the relationship between time and temperature for each cell C acquired for the superimposed image. The frequency-phase data storage area 15g stores frequency-phase data indicating the relationship between frequency and phase, which is obtained by Fourier transforming time-temperature data. In the phase image storage area 15h, a phase image generated based on the frequency-phase data (a phase image based on an equiphase image at a set frequency) is stored.

  Next, flaw detection processing in the flaw detection system 10 having the above-described configuration will be described with reference to the flowchart of FIG.

  In the flaw detection system 10, parameter setting processing is first performed (S1). In this setting process, the operator sets the resolution at the time of photographing with the infrared camera 11, the photographing frequency F, the operation content of the xenon lamp 12, and the like. For example, a screen prompting input of these parameters is displayed on the display unit 18, and necessary parameters are input by operating the input unit 17. The set parameters are stored in the parameter storage area 15c of the storage unit 15.

  If the necessary parameters are set, the flaw detection object Z is heated (S2). As shown in FIG. 5, the xenon lamp 12 emits light in this heat treatment. By being irradiated with light from the xenon lamp 12, the flaw detection object Z moving at a predetermined speed V is heated at the heating position P0. In the present embodiment, the xenon lamp 12 emits light according to the operation content set in the setting process of step S1. Moreover, since the heated flaw detection object Z continues to move at the predetermined speed V, it enters the field of view X of the infrared camera 11. In addition, regarding this heat processing, you may perform by the operator operating the on / off switch of the xenon lamp 12 manually.

  Next, acquisition and storage processing of the thermal image FL is performed (S3). As shown in FIG. 6, when the thermal image FL is acquired, photographing with the infrared camera 11 is performed. As a result, the thermal image FL described in FIG. 3A, that is, the thermal image FL arranged in a matrix obtained by the output of the cell C indicating a value corresponding to the amount of heat is as shown in FIG. Acquired in time series. In these thermal images FL, the flaw detection target Z moves in the visual field X in the movement direction. Further, the thermal image FL output from the infrared camera 11 is stored in the thermal image storage area 15 d of the storage unit 15. Note that the shooting end timing may be ended by an operation of the input unit 17 by the operator, or the shooting may be ended on the condition that a predetermined shooting time has elapsed.

  If the thermal image FL is stored, the offset S is acquired and stored (S4). That is, the control unit 16 acquires the moving speed V of the flaw detection target Z from the time-series thermal image FL and divides the flaw detection target Z by the imaging frequency F, thereby acquiring the movement pitch p of the flaw detection target Z. Further, the control unit 16 acquires the offset S by dividing the acquired movement pitch p by the visual field Δ per cell C. Thus, the control unit 16 that performs the offset S acquisition process corresponds to an offset acquisition unit.

  The moving speed V of the flaw detection object Z can be acquired by various methods. For example, it can also be obtained by causing the control unit 16 to refer to a plurality of thermal images FL having different imaging timings and calculating the movement amount of a feature point such as an edge of the flaw detection target Z. When the moving speed V of the flaw detection object Z is known in advance, the moving speed V may be input by the operator in the parameter setting process (S1). The acquired offset S is stored in the parameter storage area 15 c of the storage unit 15.

  Next, a superimposed image acquisition and storage process is performed (S5). This process is performed by the control unit 16. Therefore, the control unit 16 corresponds to a superimposed image acquisition unit that acquires a superimposed image.

  Acquisition of a superimposed image is generated by moving a plurality of thermal images FL having different shooting timings in units of cells. As shown in FIG. 8, in a certain thermal image FL, an imaginary point at a position P1, which is an intermediate point (Δ / 2) in the X direction in the cell C with M = 1, is M = 5 in the thermal image FL after n frames. A case where the cell C has moved to the position P2 in the cell C will be described as an example. This virtual point moves in the X direction by a pitch p for each frame of the thermal image FL. And after n frames, since it has moved to the position P2, the total movement amount of the virtual points is p × (n−1). Here, the size of the cell in the X direction is Δ. Therefore, the calculation of p × (n−1) /Δ+0.5 is performed while increasing the value of n by 1, 2, 3,... The superimposed image is acquired by moving the FL to the opposite side of the X direction by the integer cells. Note that the offset S may be used and the thermal image FL may be moved to the opposite side of the X direction by an integer cell each time S × (n−1) +0.5 exceeds an integer.

  This overlay process will be described with a specific example. FIG. 9 is a diagram illustrating a first specific example. In the first specific example, the imaging frequency F is 50 Hz (0.02 seconds at the imaging interval 1 / F of the thermal image FL), and the offset S is 0.21 cell. In the first specific example, the numerical value indicated by the symbol ΣS is a calculation result of S × (n−1).

  In the first specific example, for the first and second thermal images FL, the calculation result of ΣS + 0.5, that is, the calculation result of S × (n−1) +0.5 is less than 1, so that it does not move. Overlapping. For the third to seventh thermal images FL, since the calculation result of ΣS + 0.5 is 1 or more and less than 2, the images are moved and overlapped by one cell in the direction opposite to the moving direction of the flaw detection object Z. Similarly, for the eighth to eleventh thermal images FL, since the calculation result of ΣS + 0.5 is 2 or more and less than 3, they are moved and overlapped by two cells in the direction opposite to the moving direction of the flaw detection object Z. The same applies to the other thermal images FL. The 12th to 16th thermal images FL are moved by 3 cells, and the 17th to 21st thermal images FL are moved by 4 cells and overlapped.

  FIG. 10 is a diagram illustrating a second specific example. In the second specific example, the imaging frequency F is 50 Hz (1 / F is 0.02 seconds), and the offset S is 0.23 cells. In the second specific example, the first and second thermal images FL are superimposed without moving. Further, the third to sixth thermal images FL are overlapped by moving one cell in the moving direction of the flaw detection object Z. Similarly, the 7th to 10th thermal images FL are moved by 3 cells, the 11th to 15th thermal images FL are moved by 3 cells, and the 16th to 19th thermal images FL are moved by 4 cells. Let them overlap.

  In both the first specific example and the second specific example, analysis based on the superimposed image was possible. Experimentally, at least in the range where the offset S is 0.21 cell to 0.25 cell, analysis based on the superimposed image was possible. Thus, the reason why there is a width with respect to the offset S is considered to be because the influence of heat at a certain point extends to the periphery. Thus, since the value of the usable offset S has a width, even if the distance to the infrared camera 11 indicated by the symbol R in FIG. Analysis based on the superimposed image is possible. This is because the offset S and the pitch p slightly vary in the X direction.

  FIGS. 11 to 14 are diagrams for explaining the moving process of the thermal image FL. FIG. 11 shows the 20th thermal image FL out of 1000 shots, and FIG. 12 shows the 200th thermal image FL. ing. Similarly, FIG. 13 shows the 400th thermal image FL, and FIG. 14 shows the 600th thermal image FL. In these drawings, (a) is a comparative example in which the movement process is not performed, and (b) is an example (example) in which the movement process is performed.

  As is apparent from FIGS. 11A to 14A, in the comparative example in which the movement process is not performed, it can be seen that the flaw detection object Z is gradually moving to the right side of the figure. In contrast, in the example in which the movement process is performed as shown in FIGS. 11B to 14B, it can be seen that the flaw detection object Z is fixed to the left side in the thermal image FL.

  Returning to the flowchart of FIG. 4, if the superimposed image is stored, the control unit 16 causes the display unit 18 to display the superimposed image (thermal image) (S6). Thereby, an operator of the flaw detection system 10 can visually recognize the heat storage state of the flaw detection object Z on the display unit 18.

  Next, a series of processing (flaw detection processing) for detecting defects existing inside the flaw detection object Z is performed based on the superimposed image. This flaw detection process comprises the processes of S7 to S10 described below. In these processes, the control unit 16 corresponds to a flaw detection unit that detects a defect existing inside the flaw detection target Z.

  In this flaw detection processing, first, time-temperature data storage processing is performed (S7). In this process, the control unit 16 reads the superimposed image from the superimposed image storage area 15e, and acquires the relationship between time and temperature for each of the cells C constituting the image. Then, data indicating the relationship between time and temperature for all the cells C is stored in the time-temperature data storage area 15f of the storage unit 15 as time-temperature data.

  Next, Fourier transform is performed on the time-temperature data (S8). By this Fourier transform, the time-temperature data is converted into frequency-phase data indicating the relationship between frequency and phase. This Fourier transform is performed by the control unit 16. Therefore, the control unit 16 corresponds to a data conversion unit. Since the time-temperature data in this embodiment is discrete, discrete Fourier transform is performed.

The discrete Fourier transform is performed according to the following equations (1) and (2).

Here, F _n represents a conversion result (frequency component, complex intensity) at a frequency n times the value obtained by dividing the sampling frequency f _s by 2 times the number of acquired data DN. Re _n represents the real part of the conversion result, and Im _n represents the imaginary part of the conversion result. T is the temperature of the kth sampling. A _n represents the strength of the conversion result. In discrete Fourier transform, the frequency range after transformation is determined by the sampling frequency and the number of acquired data. The minimum value f _min and the maximum value f _max of the frequency after conversion are expressed by the following equations (3) and (4) using the sampling frequency f _s and the acquired data number DN.

The phase-frequency data obtained by the above Fourier transform, and the data of the minimum frequency f _min and the maximum value f _max after the conversion are stored in the frequency-phase data storage area 15 g of the storage unit 15. Thereafter, a phase image is generated and stored (S9). In the phase image generation processing, the inspection frequency is set to a discrete arbitrary frequency that is not less than the aforementioned minimum value f _{min and not} more than the maximum value f _max , and the phase value of the determined inspection frequency is used. A phase image is generated.

In the phase image generation process, the number of data DN calculated at a frequency of 1 / T between f _min = 1 / T, 2 / T, 3 / T... F _s / 2 (= N / 2T). Then, f _max = f / 2 (= (N / 2) / T), and half phase images of the number N of cells C constituting the acquired thermal image FL are generated.

The inspection frequency can be arbitrarily determined as long as it is in the range of the minimum value f _min or more and the maximum value f _max or less. For example, it can be set to a frequency determined from a thermal image capturing time and a capturing time interval.

  The generation of the phase image is performed by the control unit 16. Therefore, the control unit 16 corresponds to a phase image generation unit. Then, the generated phase image is stored in the phase image storage area 15 h of the storage unit 15. Moreover, the internal state from which the depth from the surface differs can be confirmed by changing a test | inspection frequency. The inspection frequency can be set by the operator via the input unit 17.

  If the phase image is generated and stored, the phase image and the inspection result are displayed (S10). In this display process, the phase image stored in the phase image storage area 15 h is read out by the control unit 16, converted into a video signal, and displayed on the display unit 18. Here, FIG. 15 is a diagram illustrating a phase image displayed on the display unit 18. And (a) is a phase image of the comparative example which has not performed the superimposition process, (b) is a phase image of the example which performed the superimposition process. As shown in FIG. 15A, it can be seen that in the comparative example, the phase image has a pattern flowing in the width direction. On the other hand, as shown in FIG. 15B, it can be seen that the flaw detection target Z can be clearly visually recognized by performing the overlapping process. Furthermore, as shown in an enlarged view in FIG. 16, it can be seen that the internal defect of the flaw detection object Z can be clearly seen in the phase image by performing the overlay process. The display unit 18 also displays the inspection result.

  As can be seen from the above description, in the flaw detection system 10 of the present embodiment, the heating step (S2) for heating the surface of the flaw detection target Z and the heated flaw detection target Z moving within the field of view are photographed. A thermal image FL acquisition step (S3) for acquiring, in time series, a thermal image FL composed of cells C indicating values corresponding to the heat amount, and an offset acquisition step (S4) for acquiring an offset indicating the relative movement amount of the flaw detection object Z ) And a plurality of thermal images FL in a state where they are moved in the width direction of the thermal image FL (the moving direction of the flaw detection object Z) by the amount of movement determined based on the offset, It is possible to realize a flaw detection method including a superposed image obtaining step (S5) to be obtained and a display step (S6) for displaying the superposed image.

  In this flaw detection method, a superimposed image in which the flaw detection target Z is superimposed at the same position can be acquired. Therefore, even when the flaw detection target Z moves within the field of view, a thermal image of the flaw detection target Z is displayed on the display unit 18. And the defect existing inside the flaw detection object Z can be recognized. This flaw detection method is suitable for individual inspection of moving objects of mass-produced products, and can capture and inspect a thermal image in the time required to move within the camera field of view. Moreover, it is useful when detecting a big defect rapidly. In particular, for a large and large-area flaw detection target such as an aircraft, application to inspection while moving a large defect is useful.

  In the flaw detection system 10, the frequency-phase data indicating the relationship between the frequency and the phase is acquired by performing Fourier transform on the time-temperature data indicating the relationship between the time and the temperature in the superimposed image. And the defect which exists in the inside of a flaw detection target is detected based on the phase image (Phase image by the equiphase image in the set frequency) produced | generated based on frequency-phase data. In this flaw detection system, since defects existing inside the flaw detection object are detected based on the phase image obtained from the frequency-phase data, defects at a deep location from the surface can also be detected (S7 to S10).

  In the flaw detection system 10, the control unit 18 (flaw detection unit) detects a defect present inside the flaw detection target based on the phase image based on the equiphase image at the set frequency. In this flaw detection system 10, it is possible to detect a defect at a location having a different depth from the surface by setting the frequency. This frequency can be arbitrarily determined. For example, it can be set to a frequency determined from a thermal image capturing time and a capturing time interval.

  In addition, in the flaw detection system 10, the inspection and determination are automatically performed with reference to the determination condition from the superimposed image (thermal image) and the phase image, but the configuration is not limited thereto. For example, the determination based on the superimposed image may be left to an operator of the flaw detection system 10 or the like.

  Moreover, in the flaw detection system 10, since the offset S is acquired based on the moving speed of the flaw detection target Z within the visual field, the number of cells C corresponding to the visual field, and the imaging frequency F, the acquisition of the offset S can be automated. . As can be understood from the fact that there is a range in the numerical value of the offset S, a certain degree of error is allowed in the moving speed of the flaw detection object Z in the visual field. Experimentally, an error of about 10% to 15% is allowed, so this moving speed may be acquired by image processing. Further, the moving speed of the flaw detection target Z may be separately detected.

  The above description of the embodiment is for facilitating the understanding of the present invention, and does not limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and the present invention includes equivalents thereof. For example, you may comprise as follows.

  In the above-described embodiment, the case where the infrared camera 11 is fixed and the flaw detection object Z moves in the width direction within the field of view X of the infrared camera 11 has been described as an example. However, the present invention is limited to this configuration. is not. The infrared camera 11 may be moved with respect to the fixed flaw detection target Z.

  In the above-described embodiment, the case where the flaw detection object Z moves in the X direction (the horizontal direction of the thermal image FL) has been described. However, the case where the flaw detection object Z moves in the Y direction (the vertical direction of the thermal image FL). Can be applied similarly. In addition, the present invention can be similarly applied to the case where the flaw detection object Z moves in each of the X direction and the Y direction (the oblique direction of the thermal image FL).

  In the above-described embodiment, time-temperature data and frequency-phase data are acquired for one cell C. However, the present invention is not limited to this configuration. For example, the frequency-phase data may be acquired with an average value of a plurality of cells C.

  In the above-described embodiment, the phase image is acquired based on the frequency-phase data, and the scratch existing in the flaw detection object Z is searched for by the phase image. However, the present invention is not limited to this configuration. For example, a defect existing inside the flaw detection object Z may be detected directly from the time-temperature data in the superimposed image.

  In the above-described embodiment, the automatically acquired offset S is used. However, the offset S obtained separately may be input in the parameter setting process (S1).

DESCRIPTION OF SYMBOLS 10 ... Flaw detection system, 11 ... Infrared camera, 12 ... Xenon lamp, 13 ... Control apparatus, 14 ... CPU, 15 ... Storage part, 15a ... Program storage area, 15c ... Parameter storage area, 15d ... Thermal image storage area, 15e ... Superimposed image storage area, 15f ... Time-temperature data storage area, 15g ... Frequency data-phase storage area, 15h ... Phase image storage area, 16 ... Control section, 17 ... Input section, 18 ... Display section, θ ... Infrared camera Angle of view, X: field of view of the infrared camera in the X direction, F: imaging frequency of the infrared camera, N: number of cells in the horizontal direction (number of pixels) in the field of view, Δ: field of view per cell, R: from infrared camera to subject , FL ... thermal image, C ... cell constituting thermal image, Z ... flaw detection object, V ... flaw detection object movement speed, p ... flaw detection object movement pitch, P ... flaw detection pair Certain point of things, S ... offset

Claims (4)

A heating unit for heating the surface of the test object;
Thermal image acquisition that acquires thermal images in which cells indicating values corresponding to the amount of heat are arranged in a matrix form in time series by photographing the flaw detection target after heating that moves relative to the visual field at predetermined time intervals And
A plurality of the thermal images are overlaid in a state in which the thermal images are moved in the relative movement direction by a movement amount determined based on an offset indicating a relative movement amount of the flaw detection object in the thermal image acquisition interval. A superimposed image acquisition unit for acquiring images;
A flaw detection system comprising: a display unit that displays the superimposed image.   Based on the phase image obtained from the frequency-phase data, the time-temperature data indicating the relationship between time and temperature in the superimposed image is subjected to Fourier transform to obtain frequency-phase data indicating the relationship between frequency and phase. The flaw detection system according to claim 1, further comprising a flaw detection unit that detects a defect existing inside the flaw detection object.   The flaw detection system according to claim 2, wherein the flaw detection unit detects a defect existing inside the flaw detection target based on a phase image based on an equiphase image at a set frequency. A heating step for heating the surface of the test object;
Thermal image acquisition that acquires thermal images in which cells indicating values corresponding to the amount of heat are arranged in a matrix form in time series by photographing the flaw detection target after heating that moves relative to the visual field at predetermined time intervals Steps,
A plurality of the thermal images are overlaid in a state in which the thermal images are moved in the relative movement direction by a movement amount determined based on an offset indicating a relative movement amount of the flaw detection object in the thermal image acquisition interval. A superimposed image acquisition step of acquiring an image;
And a display step of displaying the superimposed image.