Flaw detection method

Abstract

To provide a flaw detection method that detects flaws accurately even when an inspected object is out of lens focus of an infrared camera.SOLUTION: In a flaw detection method detecting a flaw width and flaw depth of a flaw occurring in the surface from a thermal image obtained by photographing the surface of an inspected object with an infrared camera, the flaw detection method is configured to: cause an infrared camera 2 to face a surface 13 of a reference body 1 having reference flaws 11 and 12 with a known width and depth preliminarily formed within a prescribed distance range including a focus position Pf of the camera; record a width g(x) of a flaw image among thermal images photographed for each prescribed moving distance in a facing direction within the distance range, and a temperature difference f(x) between the flaw image and a background image to prepare a flaw width calibration function and a temperature difference calibration function respectively; and obtain an accurate flaw width and flaw depth of the flaw by calibrating the flaw width and flaw depth of the flaw image among the thermal images in accordance with the focus position Pf and a distance of the surface of the inspected object for the flaw of the surface of the inspected object, using the flaw width calibration function and temperature difference calibration function.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a flaw detection method, and particularly to a method for detecting flaws generated on the surface of an object to be inspected from a thermal image of the surface of the object obtained by an infrared camera.

Background Art

[0002] When observing the thermal image of the surface of the object to be inspected, it is possible to detect flaws generated on the surface because the temperature changes abruptly in the image area corresponding to the flaws. Therefore, for example, in Patent Document 1, a tracking mirror is provided between the infrared camera and the object to be inspected, and the tracking mirror is appropriately vibrated in accordance with the moving speed of the object to be inspected, thereby preventing image blur caused by the movement of the object to be inspected.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, when the object to be inspected moves across in front of the infrared camera, image blur due to movement becomes a problem as described above. Separately from this, when the relative distance between the object to be inspected and the infrared camera varies, in an infrared camera with a fixed focal length, the object to be inspected moves out of its focal distance and becomes in a so-called out-of-focus state, and there is a problem that the width and depth of flaws cannot be accurately detected.

[0005] Therefore, the present invention solves such problems, and an object thereof is to provide a flaw detection method capable of accurately detecting flaws even when the object to be inspected moves out of the focal distance of the infrared camera.

Means for Solving the Problems

[0006] In order to achieve the above object, the present invention provides a flaw detection method for detecting the flaw width and flaw depth of flaws occurring on the surface of a test object from a thermal image obtained by photographing the surface of the test object with an infrared camera. The method includes: facing an infrared camera (2) to the surface (13) of a reference body (1) on which reference flaws (11, 12) having known widths and depths are previously formed within a predetermined distance range including its focus position (Pf), recording the width (g(x)) of a flaw image and the temperature difference (f(x)) between the flaw image and a background image in thermal images taken at every predetermined moving distance in the facing direction within the distance range, creating a flaw width calibration function and a temperature difference calibration function respectively, and calibrating the flaw width and flaw depth of the flaw on the surface of the test object in the thermal image according to the distance between the focus position (Pf) and the surface of the test object by using the flaw width calibration function and the temperature difference calibration function, thereby obtaining the accurate flaw width and flaw depth of the flaw.

[0007] Here, the temperature difference calibration function can be a Gaussian function.

[0008] According to the present invention, since a flaw width calibration function and a temperature difference calibration function are created and the flaw width and flaw depth of a flaw image in a thermal image are calibrated according to the distance between the focus position of the infrared camera and the surface of the test object by using these functions, even if the test object is out of the focus position of the infrared camera, the width and depth of the flaw on the surface of the test object can be accurately detected.

[0009] The signs in the above brackets are shown for reference in correspondence with the specific means described in the embodiments described later.

Advantages of the Invention

[0010] According to the flaw detection method of the present invention, flaws can be accurately detected even when the test object is out of the lens focus of the infrared camera.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Embodiment for Carrying Out the Invention

[0012] Note that the embodiments described below are merely examples, and various design improvements made by those skilled in the art without departing from the gist of the present invention are also included in the scope of the present invention.

[0013] The procedure for obtaining the flaw width calibration table and the temperature difference calibration table in the method of the present invention is as follows. As shown in FIG. 1, for example, as a reference body, linear grooves with a width of a and depths of b and c respectively are formed in parallel on the plate surface 13 of a stainless steel metal plate 1 to serve as reference flaws 11 and 12. Here, an example of a is 0.5 mm, and examples of b and c are 1 mm and 3 mm respectively. An infrared camera 2 is opposed to the metal plate 1 as shown in FIG. 1, and the surface of the metal plate 1 is photographed to obtain a thermal image of the surface.

[0014] Actually, as shown in FIG. 2, the upright metal plate 1 is installed so as to be linearly movable in a perspective manner with respect to the infrared camera 2 within a predetermined distance including the focus position Pf of the infrared camera 2 that coincides with the surface 13 of the metal plate 1, and rotatable around the center line 14 of the plate surface 13. Then, the movement amount x and the rotation amount y of the metal plate 1 from the focus position Pf are measured by a line laser distance meter 3 placed near the infrared camera 2. The thermal image data of the infrared camera 2 and the distance (changing according to the movement amount x) data of the line laser distance meter 3 are sent to a data processing device (not shown). Here, the rotation amount y of the metal plate 1 is converted into the movement amount x of the metal plate 1.

[0015] In the data processing device, a flaw width calibration table and a temperature difference calibration table are created from the thermal image data and the movement amount data as described below.

[0016] Figure 3 shows how the flaw width g(x) (in pixels) of the flaw images of the reference flaws 11 and 12 in the thermal image changes with the change in the movement amount x. In the figure, line A shows the change in the flaw width g(x) of the flaw image of the reference flaw 11, and line B shows the change in the flaw width g(x) of the flaw image of the reference flaw 12. According to this figure, even when the flaw depth changes, the degree of change in the flaw width g(x) is not very different, and the change in the flaw width g(x) generally follows well the flaw width calibration function shown in the following formula (1). In formula (1), D is the effective diameter of the lens of the infrared camera, Pw is the number of elements of the camera light-receiving sensor, Bw is the sensor size of the camera light-receiving sensor, f is the aperture value, F is the focus distance (for example, 300 mm), and WD is (F - |x|).

[0017] JPEG2025104455000002.jpg23152

[0018] Based on such a flaw width calibration function, a flaw width calibration table can be created with the case where the movement amount x is zero (i.e., when the metal plate 1 is at the focus position Pf) as the reference.

[0019] Figure 4 shows how the temperature difference f(x) between the flaw image of the reference flaw and the background image in the thermal image changes with the change in the movement amount x. In the figure, line C shows the change in the temperature difference f(x) of the flaw image of the reference flaw 12, and line D shows the change in the temperature difference f(x) of the flaw image of the reference flaw 11. According to this figure, as a whole, when the movement amount x increases, the temperature difference f(x) between the flaw image and the background image gradually decreases (i.e., the temperature of the flaw image decreases). However, because the influence of cavity radiation is large when the flaw depth is deep, especially near the focus position Pf (i.e., when the movement amount x is 0), the temperature difference f(x) from the background image appears large according to the flaw depth. The change in this temperature difference f(x) generally follows well the Gaussian function shown in the following formula (2). In the formula, a is a constant, d is the flaw depth, and σ is the standard deviation.

[0020] JPEG2025104455000003.jpg23137

[0021] Based on such a Gaussian function, a flaw depth calibration table can be created with the case where the movement amount x is zero as the reference.

[0022] As described above, prepare a flaw width calibration table and a flaw depth calibration table. When a flaw image of a flaw on the surface 13 of the object to be inspected 1 is obtained, measure the amount of movement x from the focus position Pf of the surface 13 of the object to be inspected at that time, and calibrate the flaw width and flaw depth of the flaw image with each of the above tables, so that the accurate flaw width and flaw depth of the surface flaw can be detected without being affected by the defocusing caused by the movement.

[0023] Note that, without preparing a calibration table as in the above embodiment, accurate values may be calculated by directly substituting the flaw width and flaw depth during movement into the above formulas (1) and (2). Also, the object to be inspected is not limited to metal.

[0024] Alternatively, without using the above formulas (1) and (2), the data of the flaw width and temperature difference obtained when moving a reference object having a plurality of reference flaws formed thereon may be directly tabulated, and the accurate flaw width and flaw depth of the surface flaw may be obtained by complementary calculation or the like.

Explanation of Reference Numerals

[0025] 1... reference object, 11, 12... reference flaws, 13... surface, 2... infrared camera, 3... line laser distance meter, f(x)... temperature difference, g(x)... width of flaw image, Pf... focus position.

Claims

1. In a flaw detection method for detecting the width and depth of a flaw generated on the surface of a test object from a thermal image obtained by photographing the surface of the test object with an infrared camera, an infrared camera is opposed to the surface of a reference body in which a reference flaw having a known width and depth is previously formed within a predetermined distance range including its focus position, and the width of the flaw image in the thermal image photographed for each predetermined movement distance in the facing direction within the distance range, and the temperature difference between the flaw image and the background image are recorded to create a flaw width correction function and a temperature difference correction function, respectively. For the flaw on the surface of the test object, the width and depth of the flaw image in the thermal image are corrected according to the distance between the focus position and the surface of the test object using the flaw width correction function and the temperature difference correction function, thereby obtaining the accurate width and depth of the flaw. A flaw detection method characterized by this.

2. The flaw detection method according to claim 1, wherein the temperature difference correction function is a Gaussian function.

Images