X-ray or gamma ray systems or devices – Electronic circuit – With display or signaling
Reexamination Certificate
1999-06-18
2001-06-19
Kim, Robert H. (Department: 2882)
X-ray or gamma ray systems or devices
Electronic circuit
With display or signaling
C378S023000
Reexamination Certificate
active
06249568
ABSTRACT:
TECHNICAL DOMAIN AND PRIOR ART
The invention described relates to the problem of radioscopy processing of moving objects.
In general, pose times for radioscopy of moving objects are short. This results in very noisy images, since only a small number of photons are detected while taking the X-ray. Two techniques are commonly applied to overcome this problem.
When the object moves slowly, the simplest technique consists of summating a small number of successive radiographs (sliding total), assuming that these images are identical. In this case a loss of geometric resolution is accepted in return for a reduction of noise on the images. This function is usually provided on most real time image processing systems. If n photons arrive per pixel on the detector for each radiograph, the signal to noise ratio of the radiograph will be {square root over (n)}. If m successive images are summated to generate an image with less noise, the signal
oise ratio of this image will be {square root over (n·m)}. However, the geometric resolution is lower. If the object moves by &Dgr;x between two successive radiographs, the order of magnitude of the system resolution will be m·&Dgr;X.
Another technique for quickly moving objects consists of summating a small number of successive radiographs by shifting them to take account of the movement of the object. This technique is commonly called TDI (Transfers Delay Integration) in CCD technology. In instruments, TDI is usually used in the sensor. In the case of CCD sensors, charges in the detector move at the same time as image points. This means that the summation can be done by analog means, limiting firstly the number of data to be transferred, and secondly the number of operations to be executed on the CPU.
This technique is limited to a small number of contributions since projections of the various structures in the object do not move at the same speed on the screen, depending on their arrangement within the depth of the object. If a large number of successive images is summated, the resulting image will represent a single plane of the object. Two configurations can be considered to illustrate this problem.
Firstly, in the case of an object that is thin compared with the source-detector distance, the position of a structure within the depth of the object will have very little influence on the position of the projection. Consequently, the TDI operation can be applied to the entire sensor, in other words to the number of images corresponding to the presence of the structure searched for on the image. The TDI is the reference method in this configuration.
In the case of an object with a significant thickness with respect to the source-detector distance, the position of the projection of a structure moves a great deal as a function of the depth component of its position. Therefore, the TDI cannot be easily applied. The TDI will tend to produce a sharp image of the plane corresponding to the offset between superposed images, and a blurred image of other parts of the object. In particular, specific processing had to be developed for this configuration, and this processing is the subject of this invention.
In the case of a quickly moving object, the object moves a long way on all images to be combined. Typically, in an application for the inspection of rocket propulsion units, the number of radiographs that may be used in a sliding total is of the order of 6. The geometric resolution of the resulting image degrades very quickly for larger numbers. In the same manner, the use of TDI is limited to about 20 radiographs in which only the central detector area is used.
DESCRIPTION OF THE INVENTION
One purpose of the invention is to supply a radiographic image of a thick moving object which would combine a large number of successive images of this object during its displacement.
The proposed method consists of improving the signal to noise ratio in the radiograph, concomitant with a slight deterioration in the geometric resolution of the radiograph. Depending on the processing used, an image is generated which will be as similar as possible to the radiograph that would be obtained with the source at infinity. This is the radiograph that would be obtained if each ray arrived at the detector along a plane perpendicular to the detector called the projection plane. The proposed method uses the information contained in the detector area corresponding to the projection of the projection plane during its displacement.
The proposed method provides a means of combining a large number of radiographic images in order to limit noise on images while limiting the loss of geometric resolution. The other advantage of this method is that it can be used on the same software architecture as tomosynthesis. Consequently, this method may be used in real time in radioscopy.
The first purpose of the invention is a radiography process for use with a moving object between a source and a detector, comprising:
determination of a plane (P), called the object projection plane, such that the radiographic image of this plane is a line, for one of the positions of the moving object,
the production of radiographic images of the object while it is moving, using the source and the detector,
the combination of the various contributions of projections of the projection plane in the series of radiographic images obtained.
The combination step may be a step in which the values corresponding to the projection of the projection plane are summated.
According to one embodiment, the object moves in translation along a direction (D) and has a thickness l along the direction of the projection plane, the number of combined pixels to generate one pixel in the resulting image being equal to:
N
p
s
=
∑
i
=
1
i
=
m
(
l
·
Δ
x
·
m
(
FG
+
l
)
×
FGd
FG
×
1
P
+
1
)
where p is the size of a pixel, and where FG and FGd represent the distance between the source and the object, and the distance between the source and the detector respectively, and where m is the number of successive positions of the object separated by a distance &Dgr;x for which an image is created.
The object may also move along a circular path, the projection plane being a radial plane from the object.
A correspondence table may be built up associating each pixel in the radiographic image with a pixel on the projection line from the projection plane.
Thus, a table Tab(k, Ie, Je)=Js can be defined which associates pixel (Ie, Je) in image N+k, where k is a positive or negative integer, with pixel Js in the projection from the projection plane.
Another purpose of the invention is a process for taking a radiograph of an object moving between a source and a detector, comprising:
determination of an area S
0
for a given position of the object, this area being projected onto the detector along a curve (C
1
), its projection taking place based on an area S
p
in its other positions,
production of radiographic images of the object as it moves, using the source and the detector,
processing of the projection along (C
1
), this processing comprising the following steps:
a) elementary breakdown of the area S
0
, firstly along lines 1 along the direction of the radiation from the source to the detector (called radiation attenuation lines) and secondly along lines in a direction different from the direction of lines 1, each point on the meshed area being identified by its coordinates (I
e
, J
e
),
b) for each position k of the object, calculation of the position (I
d
, J
d
), in S
p
, of the projection onto the detector of each point identified by its coordinates (i
e
, J
e
), the value of the measurement in (I
d
, J
d
) being f(I
d
, J
d
) on S
p
,
c) calculation of the projection of S
0
on C
1
which is a function of I
d
, for all positions of the area S
0
while the object is moving, starting from the different positions k of the area, and definition of a function:
∑
k
∑
Jd
f
k
(
I
d
,
J
d
)
=
projIe
,
d) movement of the object,
Rizo Philippe
Sauze Roland
Anderson Kill & Olick P.C.
Commissariat A l'Energie Atomique
Dunn Drew A.
Kim Robert H.
LandOfFree
Process for improving a signal/noise ratio of the image of a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Process for improving a signal/noise ratio of the image of a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for improving a signal/noise ratio of the image of a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2439035