Method for characterizing the optical focus of an X-ray source.

Abstract

The invention relates to a method for characterizing the optical focus of a source (11) of an X-ray beam (13) directed towards a detector (12), comprising the following steps: - Evaluation of the blur in a radiograph of a standard object (14); and - Determination of a dimension of the optical focus; the evaluation of the radiograph blur comprising: - Determination of a gray-level profile of a segment of interest in the radiograph, - Determination, from the second derivative of said gray-level profile, of points of interest in the gray-level profile; and - Determination of the difference between the points of intersection between a line passing through said points of interest and, respectively, an upper and lower level of the gray-level profile. Figure for the abstract: Fig. 3

Description

Title of the invention: Method for characterizing the optical focus of an X-ray source. Technical field

[0001] The present invention relates to the field of radiology, and more specifically, to X-ray tomography.

[0002] The present invention is advantageously of interest in the field of non-destructive testing of industrial parts, and in particular parts of the aeronautical industry. STATE OF THE ART

[0003] Non-destructive testing (NDT) of a part refers to a set of methods which make it possible to characterize the state of integrity of this part without degrading or altering it.

[0004] It is used in many areas of industry such as space, automotive, nuclear, etc. In particular, the NDT of aeronautical parts is an essential element of the operational safety of aircraft, the objective of which is to avoid any defect which could cause a failure in flight.

[0005] Among the NDT methods, X-ray radiography stands out for its ability to visualize the interior of the part in a minimally intrusive manner and to resolve details down to micrometer size. A standardized X-ray radiograph is interpreted as an image of the attenuation of X-rays as they pass through the part.

[0006] Tomography consists of the acquisition of one or more thousand radiographs of a part which may have a structure of great geometric complexity in order to calculate a complete three-dimensional image of the part.

[0007] To perform an X-ray radiography of a part, an X-ray radiography system is required. Such a system comprises an X-ray generator emitting X-ray radiation and an X-ray sensitive detector capable of capturing the radiation emitted by the generator. The detector converts the captured physical signal into an electrical signal allowing an image or radiograph to be obtained.

[0008] The detector may be a planar detector, i.e. consisting of a surface matrix of photodiodes, or a linear detector, i.e. consisting of a plurality of photodiodes aligned with each other. A planar detector is capable of capturing the radiation transmitted by the generator and partially absorbed by the part, as well as the radiation diffused by the part being tested (which is arranged between the detector and the generator). Conversely, the linear detector only captures the radiation transmitted by the generator and partially absorbed by the part, the radiation diffused by the part being controlled being filtered. Unlike the planar detector which, being a surface matrix of photodiodes, can acquire a large amount of data in a short time, the linear detector only acquires data line by line, which is more time-consuming.

[0009] On the other hand, in the high energy range, that is to say when the emitted X-rays have an energy greater than 1 MeV, the generator comprises a linear particle accelerator which itself comprises a target.

[0010] The linear particle accelerator is configured to send electrons onto the target so that a collision between said target and said electrons causes an emission of X-rays. A linear particle accelerator is characterized in particular by an optical focus or focal spot corresponding to the imprint of the electron beam on the target, which is directly linked to the spatial resolution of the acquisition. It is therefore essential to characterize this optical focus, in particular by precisely determining its vertical and horizontal dimensions.

[0011] The IEC62976 standard proposes a method for measuring the dimensions of an optical focus of a linear particle accelerator, using an X-ray taken from a standard part. The standard part is a copper block which is characterized by a left upper surface of the hyperbolic paraboloid type.

[0012] The method for measuring the dimensions of the optical focus of a linear particle accelerator according to the aforementioned standard comprises the following steps:

[0013] First, an X-ray of the standard part is taken.

[0014] In a second step, a geometric magnification of the radiograph is determined by making the ratio between a dimension of the standard part taken on the radiograph and the actual dimension of the standard part.

[0015] In a third step, a blur of the radiograph is determined. More precisely, this blur is determined from a gray level profile of a segment of interest of the radiograph which images a transition between the interior of the standard part and the exterior of the standard part. More precisely, the segment of interest extends between a point of the radiograph imaging the exterior of the standard part and a point of the radiograph imaging the interior of the standard part.

[0016] With reference to [Fig. 1], the gray level profile therefore comprises a portion of high levels H, a portion of low levels B, and a transition zone T connecting the portions of high levels H and low levels B. By determining the distance between the positions of two points of the gray level profile corresponding on the one hand to an upper limit L1 equal to 84% of the difference between the maximum and minimum values ​​of the gray level profile, and on the other hand to a lower limit L2 equal to 16% of the difference between the maximum and minimum values ​​of the gray level profile of gray, then multiplying said distance by a predetermined coefficient of 1.47, a blur value is determined. Since blur is directly linked to the dimensions of the optical focus of the linear particle accelerator, said dimensions of the optical focus are determined, in particular using the geometric magnification determined beforehand. The values ​​of 84%, 16% and 1.47 are predetermined and defined in the IEC62976 standard.

[0017] However, this standard and this method were established at a time when the detectors used were overwhelmingly linear detectors. This method therefore does not take into account the impact of the radiation scattered by the standard part, which distorts the measurement as shown in Figures 1 and 2. More precisely, [Fig. 1] illustrates a gray level profile of the segment of interest taken on a radiograph established using a linear detector, while [Fig. 2] illustrates a gray level profile of the segment of interest imaging the same portion of the standard part taken on a radiograph established using a plane detector. In [Fig. 1], the high level H and low level B portions are substantially horizontal, while in [Fig.2], they are not, due to the capture of the radiation diffused by the standard part, so that the method recommended in the IEC62976 standard, and in particular the predetermined values, are not suitable for the characterization of a flat detector. EXPOSED

[0018] An aim of the present invention is to remedy the aforementioned drawbacks, by proposing a method for characterizing the optical focus of an X-ray source suitable for any type of X-ray detector.

[0019] To this end, according to a first aspect, a method is proposed for characterizing the optical focus of a source capable of emitting an X-ray beam towards a detector capable of generating an X-ray of an object placed between the source and the detector, the method comprising the following steps: - Evaluation of a blur of an X-ray of a standard object placed between the source and the detector; and - Determination of a dimension of the optical focus from the evaluation of the blur;

[0020] the evaluation of the blur of the radiography comprising: - Determining a grayscale profile of a segment of interest of the radiograph, the segment of interest successively imaging a portion of the interior of the standard object, a portion of the profile of the standard object and a portion of the exterior of the standard object; - The determination, from the second derivative of said gray level profile, of points of interest belonging to a transition zone of the gray level profile gray level between a high gray level portion and a low gray level portion of said gray level profile; and - The determination of a position difference between the positions of the points of intersection between, on the one hand, a straight line passing through said points of interest and, on the other hand, respectively an upper level and a lower level of the gray level profile.

[0021] Advantageously, the points of interest are points of the gray level profile located at the positions of the extrema of the second derivative of the gray level profile.

[0022] Advantageously, the method comprises, prior to the step of determining the points of interest, a step of smoothing by moving average of the second derivative of the gray level profile, the determination of the position of the extrema of the second derivative then being carried out on said smoothing.

[0023] Preferably, the determination of a dimension of the optical focus comprises: - determination of the geometric magnification of the radiograph; - the division of the position deviation determined by the geometric magnification of the radiograph.

[0024] Advantageously, the determination of the geometric magnification of the radiograph comprises the following phases: - determination of the size in pixels of the standard object on the radiograph; and - division of the size in pixels determined on the radiograph by the actual size of the standard object;

[0025] Advantageously, the size in pixels on the radiograph of the determined standard object and the actual size of the standard object are sizes of the width of the standard object.

[0026] Preferably, the upper level of the grayscale profile corresponds to 84% of the difference between the maximum and minimum values ​​of the grayscale profile, and the lower level of the grayscale profile corresponds to 16% of the difference between the maximum and minimum values ​​of the grayscale profile.

[0027] Advantageously, the segment of interest is perpendicular to an upper surface of the standard object.

[0028] Advantageously, the upper face is a left face of the hyperbolic paraboloid type.

[0029] The invention also relates to a computer program product for implementing the method of characterizing the optical focus of a source capable of emitting an X-ray beam as defined above. DESCRIPTION OF FIGURES

[0030] Other characteristics, aims and advantages of the invention will emerge from the following description, which is purely illustrative and non-limiting, and which must be read in conjunction with the appended drawings in which:

[0031] [Fig.l] illustrates a grayscale profile of a segment of interest taken from a radiograph established using a linear detector;

[0032] [Fig.2] illustrates a gray level profile of the segment of interest taken on a radiograph established using a plane detector;

[0033] [Fig. 3] schematically illustrates an X-ray radiography system suitable for use in implementing the method according to the invention;

[0034] [Fig.4] schematically illustrates the steps of the method according to the invention;

[0035] [Fig.5A] illustrates in perspective a standard object with a left upper surface as described in the IEC62976 standard;

[0036] [Fig.5B] illustrates in front view the standard object of [Fig.1];

[0037] [Fig.6] illustrates an X-ray of the standard object taken by the system of [Fig.3] on which two segments of interest are drawn;

[0038] [Fig.7] illustrates a first gray level profile according to one of the segments of interest of [Fig.6] with apparent indications relating to the evaluation of the blur of the radiograph;

[0039] [Fig.8] illustrates a second derivative of the first gray level profile of [Fig.7];

[0040] [Fig.9] illustrates a second gray level profile according to the other of the segments of interest of [Fig.6].

[0041] Throughout the figures, similar elements bear identical references. DETAILED DESCRIPTION

[0042] [Fig. 3] schematically illustrates an X-ray radiography system 10 suitable for use in tomography applications.

[0043] The X-ray radiography system 10 comprises an X-ray source 11 emitting X-rays in an emission direction, and a detector 12 sensitive to the X-rays capable of capturing the rays emitted by the source 11 and of generating an X-ray of an object placed between the source 11 and the detector 12 from the captured rays. The detector 12 converts the captured physical signal into an electrical signal making it possible to obtain an image or X-ray.

[0044] The X-ray source 11 is capable of emitting a beam 13 of so-called high-energy X-rays, i.e. having an energy greater than 1 MeV. Thus, the ray source 11 comprises a linear particle accelerator 11a which comprises a target 11b.

[0045] The linear particle accelerator 11a is configured to send electrons onto the target 11b such that a collision between said target 11b and said electrons causes an emission of X-ray beam 13. The X-rays of beam 13 emitted by source 11 propagate in an emission direction represented by arrows 13.

[0046] The detector 12 may be a planar detector, i.e. consisting of a surface matrix of photodiodes, a linear detector, i.e. consisting of a plurality of photodiodes aligned with each other, or even a unitary detector, consisting of a single photodiode.

[0047] The X-ray radiography system 10 is capable of enabling the implementation of a method for characterizing the optical focus (or focal spot) of the source 11, and more precisely of the linear particle accelerator 11a, the steps of which are schematically represented in [Fig. 4], and which will now be described.

[0048] In a first step E1, a standard object 14 is placed on the path of the beam 13 of X-rays emitted by the source 11, between said source 11 and the detector 12.

[0049] The standard object 14 is preferably a block of copper as described in the IEC62976 standard. Alternatively, the standard object 14 may be a block of a different suitable material, such as tungsten, or any other material sufficiently dense to provide considerable absorption of the X-rays. The standard object may also have a different shape, but must have a sharp edge perfectly aligned with the direction of emission of the X-ray beam 13.

[0050] In the following, it will be considered that the standard object 14 is a copper block as described in the IEC62976 standard, and which will now be described with reference to Figures 5A and 5B. In particular, the standard object 14 is illustrated in perspective in [Fig.5A] and in front view in [Fig.5B].

[0051] More specifically, the standard object 14 comprises a rectangular base 15, a rear face 16, a front face 17, a first lateral face 18 and a second lateral face 19. Each of the first and second lateral faces 18 and 19, the front face 17 and the rear face 16 extends perpendicularly respectively from one of the sides of the rectangular base 15. The front 17 and rear 16 faces extend perpendicularly from the longer sides of the rectangular base 15. The front face 17 comprises a first upper edge 20a opposite the rectangular base 15, and which extends between the first lateral face 18 and the second lateral face 19. Similarly, the rear face 16 comprises a second upper edge 20b opposite the rectangular base 15 and which extends between the first lateral face 18 and the second lateral face 19.The plane P perpendicular to the plane of the front face 17 comprising the first upper edge 20a forms an angle of 15° with the plane P' perpendicular to the plane of the rear face 16 and comprising the second upper edge 20b. The standard object 14 further comprises an upper face 20 opposite the rectangular base 15 and extending between the four ends of the first and second upper edges. 20a and 20b. The upper face 20 is configured so that the rays of the beam 13 passing orthogonally through the front face 17 pass through more material of the standard object 14 at the middle of the upper face 20 than between said middle and the first and second lateral faces 18 and 19. For example, the upper face 20 is a left surface of the hyperbolic paraboloid type.

[0052] The standard object 14 is positioned in the middle of a rectilinear path between one end of the source 11 by which the beam 13 is emitted and the detector 12. More precisely, the standard object 14 is oriented so that the front face 17 is opposite the source 11 and orthogonal to the direction of the rays of the beam 13 emitted by said source 11. In all cases, the standard object 14 is oriented so as to have a sharp edge which is perfectly aligned with the direction of emission of the beam 13.

[0053] In a second step E2, an X-ray of the standard object 14 is carried out.

[0054] Such an X-ray is shown in [Fig.6].

[0055] In a third step E3, a blur of the radiograph taken in step E2 is evaluated.

[0056] For this, a first gray level profile 21 of a first segment of interest SI is plotted (sub-step E31). The first gray level profile 21 illustrates more precisely the gray level of each of the points of a first segment SI plotted on the radiograph as a function of the position (in pixels) of said points in said first segment SL

[0057] The first segment SI images a transition between the interior of the standard object 14 and the exterior of the standard object 14. More precisely, the first segment SI successively images a portion of the interior of the standard object 14, a portion of the profile of the standard object 14 and a portion of the exterior of the object 14.

[0058] Preferably, the first segment S1 extends vertically on the radiograph, that is to say in a direction perpendicular to the rectangular base 15. In addition, the first segment SI preferably comprises the middle of the upper face 20 in the plane of the radiograph.

[0059] The first gray level profile 21 is illustrated in [Fig.7].

[0060] The first gray level profile 21 thus comprises a high level portion of gray H (corresponding to the portion of the exterior of the standard object 14), a portion of low gray levels B (corresponding to the portion of the interior of the standard object 14) and a transition zone T between the portion of high gray levels H and the portion of low gray levels B (corresponding to the portion of the profile of the standard object 14).

[0061] Next, a characterization of the transition zone T is carried out (sub-step E32). Indeed, the slope of the transition zone T is linked to the blur in that the steeper the slope of the transition zone T, the lower the blur. More precisely in sub-step E32, points of interest belonging to the transition zone T are determined. The first SI segment is oriented and positioned as described previously in order to limit the transition zone and facilitate its characterization.

[0062] These points of interest are determined from the second derivative 21a of the first gray level profile 21, the plot of which is shown in [Fig.8]. In particular, the positions of the extrema of the second derivative 21a of the first gray level profile 21 are determined and reported on the first gray level profile 21. The points of interest are more precisely points PI and P2 of the first gray level profile 21 located at said positions of the extrema of the second derivative 21a of the first gray level profile 21. Said points of interest PI and P2 thus make it possible to characterize a linear zone of the transition zone T, and more precisely to delimit in position the zone of steepest slope of the transition zone T.

[0063] Optionally, prior to this determination of the points of interest PI and P2 of the transition zone T, a smoothing of the second derivative 21a of the first gray level profile 21 can be carried out, for example by moving or sliding average of said second derivative 21a, in order to facilitate the location of the extrema. Alternatively, the smoothing can be carried out simply and beforehand by filtering the noise at acquisition when taking the radiograph.

[0064] When smoothing is performed on the second derivative 21a of the first gray level profile 21, the determination of the positions of the extrema is performed on said smoothing.

[0065] From the determined points of interest PI and P2, an extrapolation of the linear zone defined by the points of interest PI and P2 is carried out (sub-step E33). More precisely in sub-step E33, two intersection points II and I2 are determined.

[0066] The first point of intersection II is a point of intersection between, on the one hand, a straight line D passing through said previously determined points of interest PI and P2 and, on the other hand, an upper level NI of the first gray level profile 21. For example, the upper level N1 of the first gray level profile corresponds to 84% of the difference between the maximum and minimum values ​​of the first gray level profile 21.

[0067] Similarly, the second point of intersection 12 is a point of intersection between, on the one hand, the straight line D and, on the other hand, a lower level N2 of the first gray level profile 21. For example, the lower level N2 of the first gray level profile 21 corresponds to 16% of the difference between the maximum and minimum values ​​of the first gray level profile 21.

[0068] A position difference is then measured between the first intersection point II and the second intersection point 12 (sub-step E34). This position difference is the evaluation of the blur in pixels (the first gray level profile 21 being plotted as a function of the pixel position of the points of the first segment SI).

[0069] In a final step E4 of the method, a dimension of the optical focus of the source 11 is determined from the blur evaluated in step E3.

[0070] More precisely, in a first sub-step E41, a geometric magnification of the radiograph taken in step E1 is calculated.

[0071] For this, a portion of the standard object 14 is measured in pixels on the radiograph, then divided by its known actual size. More precisely, a second gray level profile 22 of a second segment of interest S2 is plotted.

[0072] The second segment of interest S2 is defined on the radiograph, said second segment of interest S2 imaging a portion of the object, for example and preferably the width of the standard object 14. Preferably, and as shown in [Fig. 6], the second segment of interest S2 extends perpendicular to the first lateral face 18 and to the second lateral face 19, between a point outside the standard object 14 opposite the first lateral face 18 and a point outside the standard object 14 opposite the second lateral face 19. The second segment of interest S2 therefore crosses the standard object 14 entirely in its width. In addition, the second segment of interest S2 passes through the middle of the first lateral face 18 and the middle of the second lateral face 19.The second segment of interest S2 therefore successively images a first portion of the exterior of the standard object 14, a profile portion of the first lateral face 18, a portion of the interior of the standard object 14, a profile portion of the second lateral face, then a second portion of the exterior of the standard object 14.

[0073] The second gray level profile 22 is shown in [Fig. 9], and therefore comprises a first high-level zone H1 (corresponding to the first portion of the exterior of the standard object 14), a first transition zone T1 (corresponding to the profile portion of the first lateral face 18), a low-level zone B (corresponding to the portion of the interior of the standard object 14), a second transition zone T2 (corresponding to the profile portion of the second lateral face 19), and a second high-level zone H2 (corresponding to the second portion of the exterior of the standard object 14).

[0074] To measure the width of the standard object 14, the difference between two points P3 and P4 of the second gray level profile 22 is measured. The points P3 and P4 are respectively points of the first transition zone T1 and of the second transition zone T2 both located at the same distance from the low-level zone B, so as to be taken at the same height. Preferably, the point P3 is midway on the first transition zone T1 between the first high-level zone H1 and the low-level zone B, and the point P4 is midway on the second transition zone T2 between the second high-level zone H2 and the low-level zone B.

[0075] The width of the standard object 14 in pixels is the distance between points P3 and P4 on the second grayscale profile 22.

[0076] This width of the standard object 14 in pixels is divided by the actual width in the metric system of the standard object 14 taken at mid-height of the first and second lateral faces 18 and 19. The result of this division is the geometric magnification of the radiograph.

[0077] Finally, in a second sub-step E42, a dimension of the optical focus of the source 11 is determined by means of a division of the blur evaluated in step E3 by the geometric magnification of the radiography determined in step E41. More precisely, the embodiment which has just been described, in particular with regard to the orientation of the standard object 14 (front face 16 orthogonal to the direction of the X-rays of the beam 13), allows the determination during step E42 of the dimension in a vertical direction of the optical focus. In order to determine a dimension of the optical focus in a horizontal direction, it is necessary to orient the standard object 14 so that the upper face 20 is in place of one of the lateral faces 18 or 19, then to repeat the method in a corresponding manner.By determining the blur using the second derivative 21a of the first gray level profile, the transition zone is characterized more precisely by being dependent on the specific case, whereas the method according to the IEC62976 standard only considered the transition zone with fixed limits, regardless of the detector 12. .

[0078] The method according to the invention therefore makes it possible to characterize an optical focus of an X-ray source comprising a linear particle accelerator more precisely, whatever the nature of the detector used, the latter being able in particular to be a planar detector.

Claims

Claims

1. Method for characterizing the optical focus of a source (11) capable of emitting an X-ray beam (13) towards a detector (12) capable of generating an X-ray of an object arranged between the source (11) and the detector (12), the method comprising the following steps: - Evaluation (E3) of a blur of an X-ray of a standard object (14) arranged between the source (11) and the detector (12); and - Determination (E4) of a dimension of the optical focus from the evaluation of the blur; the evaluation (E3) of the blur of the radiograph comprising: - The determination (E31) of a gray level profile (21) of a segment of interest (SI) of the radiograph, the segment of interest (SI) imaging in succession a portion of the interior of the standard object (14), a portion of the profile of the standard object (14) and a portion of the exterior of the standard object (14);- The determination (E32), from the second derivative (21a) of said gray level profile (21), of points of interest (PI, P2) belonging to a transition zone (T) of the gray level profile (21) between a portion of high gray levels (H) and a portion of low gray levels (B) of said gray level profile (21); and - The determination (E34) of a position difference between the positions of the points of intersection (II, 12) between on the one hand a straight line (D) passing through said points of interest (PI, P2) and on the other hand respectively an upper level (NI) and a lower level (N2) of the gray level profile (21).;

2. Method according to one of claims 1, in which the points of interest (PI, P2) are points of the gray level profile (21) located at the positions of the extrema of the second derivative (21a) of the gray level profile (21).

3. Method according to claim 2, comprising, prior to the step of determining (E32) the points of interest (PI, P2), a step of smoothing by moving average of the second derivative (21a) of the gray level profile (21), the determination of the position of the extrema of the second derivative (21a) then being carried out on said smoothing.

4. Method according to one of claims 1 to 3, in which the determination (E4) of a dimension of the optical focus comprises: - the determination (E41) of the geometric magnification of the radiograph; - the division (E42) of the position deviation determined by the geometric magnification of the radiograph.

5. Method according to claim 4, in which the determination (E41) of the geometric magnification of the radiograph comprises the following phases: - determination of the size in pixels of the standard object (14) on the radiograph; and - division of the size in pixels determined on the radiograph by the actual size of the standard object (14);

6. The method of claim 5, wherein the pixel size on the radiograph of the standard object (14) determined and the actual size of the standard object (14) are sizes of the width of the standard object (14).

7. A method according to one of claims 1 to 6, wherein the upper level of the grayscale profile (21) corresponds to 84% of the difference between the maximum and minimum values ​​of the grayscale profile (21), and the lower level of the grayscale profile (21) corresponds to 16% of the difference between the maximum and minimum values ​​of the grayscale profile (21).

8. Method according to one of claims 1 to 7, in which the segment of interest (SI) is perpendicular to an upper surface (20) of the standard object (14).

9. The method of claim 8, wherein the upper face (20) is a left face of the hyperbolic paraboloid type.

10. Computer program product comprising program code instructions for executing the steps of the method for characterizing the optical focus of a source (11) capable of emitting an X-ray beam (13) according to one of claims 1 to 9 when said program is executed on a computer.

Images