JP2003000576A - Device and method for processing image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make a correcting noise invisible which occurs when performing offset- or gain correction concerning the output value saturation part of an image sensor. SOLUTION: This image processor has: a sensor (5) including a plurality of pixels for converting a subject to an electric signal; correcting means (11 and 17) for correcting dispersion of pixel characteristics; and control means (21, 22, 23 and 24) for evaluating an output signal before the correction on the prestage of the correcting means, judging the saturation of each of elements and fixing the output value after the correction into constant value in the case saturation exists.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing technique, and more particularly to an image processing technique for imaging the amount of light transmitted through a subject by irradiation of radiation. For example, used in the medical field.

[0002]

2. Description of the Related Art A large-sized image sensor using a solid-state image pickup element called a flat panel X-ray sensor in recent years when observing an object, particularly the inside of a human body, by observing a transmission distribution of radiation represented by X-ray It is becoming more and more popular to acquire the radiation distribution.

The advantage of the solid-state image pickup device is that a plurality of image receiving pixels are present on a plane and the plane energy distribution can be directly spatially sampled to be converted into a signal.

However, as a drawback, since a plurality of pixel elements for spatial sampling are basically independent elements and have different characteristics, in order to obtain an image having uniform characteristics, One of the reasons is that the characteristic variation for each pixel must be corrected.

The main characteristic variations of pixel elements which are energy conversion elements are variations in conversion efficiency (gain) and offset. When a flat panel X-ray sensor, which is a solid-state image pickup element, is used, the first is gain. Offset correction must be done.

FIG. 4 is a block diagram schematically showing the form of the above-described gain and offset correction. In the figure, symbol 1 represents an X-ray generator, and an X-ray is emitted in a direction indicated by an arrow by a control device (not shown) that generates a high voltage. Reference numeral 2 denotes a subject represented by a human body. Shows lying down. The block indicated by 5 is an X-ray image sensor used for the purpose of converting the X-ray dose intensity distribution transmitted through the subject into an electric signal, which is configured using a large-sized solid-state image sensor and is arranged in a plane matrix. A plurality of pixels are spatially sampled on a two-dimensional plane. Hereinafter, this X-ray image sensor is referred to as a flat panel sensor. This sampling pitch is 100 μm when photographing the normal internal structure of the human body.
It is set to about 200 μm. The flat panel sensor is controlled by a controller (not shown), converts a charge value proportional to the X-ray dose existing in each pixel into an electric quantity that is a voltage or a current, and sequentially scans and outputs the electric quantity.

The block indicated by 6 is an A / D converter for converting the analog electric quantity output from the flat panel sensor into a digital value. Reference numeral 7 represents a memory means for temporarily storing the A / D converted digital value as image information. The structure indicated by 8 represents a means for reading the memory 7 and branching it into two memories 9 and 10 for storage. The memory 9 is a memory that stores the image signal output from the flat panel sensor as an offset fixed pattern image without irradiating the X-ray, and 10 stores the image obtained by actually irradiating the X-ray. Memory.

As a specific imaging method, an X-ray dose measuring device for monitoring the X-ray dose transmitted through a subject (not shown) is used for controlling X-ray exposure called a photo timer, and the integrated value of the exposed X-ray dose is used. The X-ray exposure is stopped at the moment when a certain value is reached. The controller of the present apparatus scans the flat panel sensor at the same time when the X-ray exposure is stopped, captures the image information of the subject into the memory of 7, sets the branching means of 8 to the A side, and stores it in the memory of 10. Immediately after that, the flat panel sensor is driven without performing X-ray irradiation, and charges are accumulated and the offset fixing pattern is scanned and stored in the memory 7 for the same period of time as obtained by the photo timer. At this time, the branching means 8 is set to the B side and stored in the memory 9. The block indicated by 11 is a block for performing the difference, and the value of the memory 9 at the corresponding position is sequentially subtracted from the value of the memory 10 and stored in the memory block indicated by 12.

The block indicated by 13 is a lookup table (Look Up Table,) for logarithmic value conversion used for performing division.
LUT). The above-mentioned image obtained through the normal subject is stored in the memory 15 by setting the branching means 14 to the C side through the 13 LUT.

The memory 16 is a memory for storing an image when an operation called calibration is performed in this apparatus. The memory 16 acquires the image by the same operation as described above, and the branching means 14 is operated.
Although the image is stored by setting it to the D side, the difference is that only the X-ray dose distribution itself and the gain variation for each pixel are captured in the absence of the subject 2. Normally, this calibration operation is performed about once a day at the beginning of work, and this operation corrects the gain variation that the flat panel sensor has for each pixel.

That is, by subtracting the grid stripes from the original image component by the difference function of 17, an image in which the gain variation for each pixel is corrected is completed, and stored in the memory indicated by 18,
The diagnostic image processing in the latter stage, or filing, transmission, and display are used.

In an actual X-ray radiographing device, this image is converted into a diagnostic image by gradation processing, dynamic range processing, spatial frequency processing, etc., and then transferred to an external device typified by a filing apparatus. Or a hard copy is performed.

[0013]

A characteristic of X-ray image acquisition is that the X-ray dose dynamic range to be received is very wide. In order to more accurately depict the inside of the subject, the X-ray dose is adjusted so that the dynamic range of the subject's transmittance distribution matches the dynamic range of the flat panel sensor.
The dose is very strong, due to reaching the sensor directly.

In the image intensifier that has been often used in the conventional medical X-ray image acquisition, if the above-mentioned direct line is too strong, a phenomenon such as halation occurs and even an image of the subject part may be defective. JP 11-
As described in 318876, the exposure time, light amount control, gain control, etc. had to be finely adjusted while observing the output signal.

Further, in the flat panel sensor of the solid-state image pickup device, the above-mentioned phenomenon such as halation is unlikely to occur, but the output value is saturated for an excessively strong direct line input. This saturation may occur when the charge capacity of the pixel on the sensor is exceeded, when the electric system amplifier in the subsequent stage is saturated, or in the A / D conversion system that converts the electric quantity into a digital signal value. In any case, the obtained output value is fixed to a substantially constant and invariable value.

When the above-mentioned offset or gain correction is performed in this saturated state, the correction data obtained in the non-saturated state is used for each, so that the fluctuation for each pixel to be corrected is conversely superimposed on the image. The noise is generated by the correction.

FIG. 5 is a diagram for explaining this phenomenon. Figure 5
(i) shows the characteristics of the incident X-ray dose and the output voltage value of a certain two pixels A and B. FIG. 5 (ii) shows a state after A / D conversion. In this case, the output data is saturated at the A / D conversion maximum value equal to or lower than the pixel output voltage. FIG. 5 (iii) shows pixel A and pixel
The offset value of each B (output value at zero incident X-ray) is subtracted to perform offset correction.

This offset value may be due to variations in the characteristics of the electric system, but may also be due to the accumulation of leakage current of the energy conversion system in the pixel. That is, assuming that the irradiation time of X-rays is the accumulation time, the offset amount also differs due to the difference in the accumulation time, and the offset amount is
It can be considered that it always changes depending on the X-ray imaging conditions.

After that, FIG. 5 (iv) is a diagram in which the slope of each characteristic, that is, the gain is corrected. Pixel at this time
The saturation values of A and the pixel B are different from A ₀ and B _0, and a difference of δ occurs. That is, the value at which the incident X-ray is saturated differs depending on the pixel, and this is observed as noise.

The value of the incident X-ray dose that actually saturates in the figure also saturates at doses different from a and b. That is, the dynamic range varies depending on the pixel.

Further, the problem here is that the saturated pixel value is different for each pixel and for each photographing condition, and therefore, after performing the offset / gain correction, it is impossible to determine whether or not the pixel value is saturated only by the pixel value. .

Originally, since there is no image information in the direct line portion and there is almost no problem even if it is saturated, if it is saturated, it should be fixed to such a constant value. In this case, an extra pixel-by-pixel variation appears in the direct line portion as noise generated by the correction.

Since this part is irrelevant to the image part showing the structure of the subject, it may be considered that any change can be seen, but when a medical worker observes an X-ray photograph of the human body. Is disliked, even though it is an extra psychological obstacle.

In many cases, medical images are subjected to image processing to compress the dynamic range of display images. In this case, the direct line portion is also very easy to see, which may be information that is an obstacle to diagnosis.

To solve this problem, Japanese Patent Laid-Open No. 2000-244824 solves this problem by using a gain control amplifier so that the maximum value is always a signal value that can be handled at any X-ray dose. However, this method has a problem that the contrast resolution of a necessary subject portion is significantly lowered because adjustment is performed so that even a direct line portion, which originally has no information, is accurately imaged.

It is an object of the present invention to make correction noise generated when offset or gain correction is performed on an output value saturated portion of an image sensor invisible.

[0027]

According to one aspect of the present invention, a sensor including a plurality of pixels for converting a subject image into an electric signal, a correction means for correcting a variation for each pixel, and the correction means. An image characterized by having a control means which evaluates the output signal before correction in the preceding stage of the means, judges the saturation of each element, and fixes the corrected output value to a constant value when saturated. A processing device is provided.

According to another aspect of the present invention, a sensor including a plurality of pixels for converting a subject image into an electric signal, a correction means for correcting variations in each pixel, and a correction before the correction means. An image processing apparatus is provided, which comprises: a control unit that evaluates the previous output signal to determine the saturation of each element and, if saturated, does not perform correction.

According to still another aspect of the present invention, a sensor including a plurality of pixels for converting a subject image into an electric signal, a correction means for correcting variations of each pixel of the sensor, and the image sensor. An image processing apparatus is provided, which comprises: a control unit that evaluates the saturation of an output value before correction for each pixel and adds information indicating the saturated state to the information of each pixel.

According to still another aspect of the present invention, a conversion step of converting an object image into an electric signal by a plurality of pixels and an output signal before the correction are evaluated when the variation for each pixel is corrected. And a correction step of fixing the output value after correction to a constant value when the element is saturated, and providing an image processing method.

According to still another aspect of the present invention, a conversion step of converting an object image into an electric signal by a plurality of pixels and an output signal before the correction are evaluated when the variation for each pixel is corrected. Then, the image processing method is provided, which comprises a correction step of determining the saturation of each element and not performing the correction if the element is saturated.

According to still another aspect of the present invention, a conversion step of converting an object image into an electric signal by a plurality of pixels, a correction step of correcting a variation for each pixel,
An image processing method is provided, which comprises a saturation evaluation step of evaluating saturation of an output value before correction for each pixel and adding information indicating a saturated case to information of each pixel.

According to the present invention, when the image data is acquired and corrected by the sensor, the pixel data before the correction is evaluated and it is judged whether or not it is a saturated value. It can be fixed to a value and the correction noise can be made invisible.

[0034]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings along with examples. (Embodiment 1) In contrast to the above problems, in the present embodiment, when the sensor output before gain / offset correction reaches a saturation level, the above-mentioned offset /
By not performing the gain correction, it is possible to reduce the fluctuation of the direct line portion and obtain an image without a feeling of strangeness.

FIG. 6 shows a state of the operation. 6 (i) and 6 (ii) are shown in FIG.
Since it is the same as (i) and FIG. 5 (ii), the description is omitted.
Since the saturation value is fixed at the maximum A / D conversion value in Fig. 6 (ii), it does not have a different saturation value for each pixel and for each shooting condition, so it is easy to determine whether saturation has occurred at this stage. Stick to FIG. 6 (iii) shows that offset correction is not performed for the pixel value range determined to be saturated at the stage of FIG. 6 (ii). Even in this case, the saturated pixel value does not change, as shown in FIG.
Even in (iv), the gain correction is not performed for the value corresponding to the saturated pixel value, so that the saturated pixel value is fixed and the noise generated by the correction of the saturated portion is reduced.

In this example, attention is paid to the fact that the saturation occurs at the A / D converted value, but even when the saturation does not occur at the A / D converted value and the electric system including the sensor output of the preceding stage saturates, this saturated value is However, if the pixel is not corrected for offset / gain, it is possible to determine a saturated pixel within a fixed range.

Although correction is performed in another embodiment, the same effect can be obtained by discarding the correction result according to the pixel value before the offset / gain correction and fixing it to another value.

FIG. 1 is a block diagram of an X-ray imaging apparatus according to this embodiment. Most of the blocks are the same as the correction blocks described in FIG. 4 and have the same numbers, so the description of their operation is omitted.

The difference is that 21 to 24 blocks are added. Reference numeral 21 denotes a decoder that determines the range of pixel values to be saturated or a specific value from the pixel values of the image acquired from the A / D conversion, and outputs the result. Reference numeral 22 is a signal path switching means for selecting whether the pixel value is passed by the subtractor 11 with or without offset correction. The decoder of 21 outputs the E side as a SELECT signal when the pixel value is in the saturated region and passes the offset correction, and outputs the F side when the pixel value is the normal pixel value to perform the offset correction.

Reference numeral 23 is also a decoder for pixel values similar to 21, but is a decoder for discriminating the saturated pixel value when the saturated pixel value passes through the logarithmic conversion LUT of 13. Reference numeral 24 is a signal path switching means for selecting whether the pixel value is passed by the subtractor 17 with or without gain correction. If the decoder in 23 determines that the pixel value is in the saturated area, G
Side is output, and normally the H side is output, so gain correction is not performed when the saturation region is reached. In this embodiment, the decoder and the signal path switching means correspond to the control means.

Obviously, the processing system pass, value judgment, etc. in the present embodiment can be implemented by software processing, and all the contents shown in FIG. 1 can be implemented by computer programming.

Further, although the present embodiment has been described with respect to the two-dimensional image sensor, it is obvious that the one-dimensional line sensor can be similarly applied.

(Second Embodiment) FIG. 2 is a block diagram of a second embodiment of the present invention. In the figure, blocks 31 and 32, which are control means, are newly provided in comparison with FIG. The operation of blocks 31 and 32 is shown in FIG. FIG. 3I shows the bit structure of the pixel data of this embodiment. Figure 2
The A / D converter of shall convert an analog signal into a 14-bit digital value. In FIG. 3 (i), the A / D converted value itself (D00 to D13) is stored in the 14th bit from the 0th bit to the 13th bit of the pixel. Pixel bit structure is 16
Assuming that it is a bit (2 bytes), a flag called FLG is provided at the 15th bit which is the MSB. The meaning of this flag is that if it is 1, this data is saturated, and if it is 0, it is normal pixel data that is not saturated. As mentioned in the section on pointing out issues,
After the gain correction, it is difficult to determine that the pixel data is a saturated value only, so this flag is used to determine whether the data is saturated.

The operation of the block 31 in FIG. 2 is shown in a flowchart in FIG. 3 (ii). When A / D converted data is input to the block (D00 to D13), the data value is stored, and this block evaluates the value before correction. Since it is 14-bit data, the data area is 0 to 16383, but in this block, when it is a value larger than 16380, it is determined to be saturated. That is, D15 is set to 1 when saturated, and D15 is set to 0 otherwise. The operation of the 32 blocks in FIG. 2 is shown in a flowchart in FIG. 3 (iii). When the corrected data is input to the same block, D15 is judged and D15
When 1 indicates 1, that is, a saturation value, the pixel data is fixed to 16383. This value is arbitrary.

In this case, in each operation (difference, LUT) in FIG. 2, the objects to be operated are D00 to D13, and the remaining bits are copied to the memory of the next stage without passing through the operation.

With the above operation, the offset / gain correction is performed tentatively regardless of whether or not it is saturated.
When saturated at the final stage, it is fixed to a fixed value, so noise due to correction is removed. Here, although correction is performed, depending on the pixel value before the offset / gain correction,
The configuration may be such that the correction result is discarded and fixed to another value.

As described above, according to the first and second embodiments, when the X-ray image is acquired by the solid-state image sensor and the offset / gain correction is performed, the pixel data before the correction is evaluated. By judging whether it is a saturated value,
The saturated output image can be fixed to a substantially constant value despite the value change due to the correction.

This embodiment can be realized by the computer executing the program. Further, means for supplying the program to the computer, for example, a recording medium such as a CD-ROM storing the program or a transmission medium such as the Internet for transmitting the program can be applied as the embodiment of the present invention.
The above program, recording medium, and transmission medium are included in the scope of the present invention. As the recording medium, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a magnetic tape, a non-volatile memory card, a ROM or the like can be used.

It should be noted that the above-mentioned embodiments are merely examples of specific embodiments for carrying out the present invention, and the technical scope of the present invention should not be limitedly interpreted by these. Is. That is, the present invention can be implemented in various forms without departing from its technical idea or its main features.

[0050]

As described above, according to the present invention, when the image data is acquired and corrected by the sensor, the pixel data before the correction is evaluated to determine whether the value is saturated. The saturated output image can be fixed to a substantially constant value, and the correction noise can be made invisible.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

FIG. 2 is a block diagram of another embodiment of the present invention.

FIG. 3 is a diagram for explaining the operation of the block diagram of FIG.

FIG. 4 is a diagram for explaining a conventional technique.

FIG. 5 is a diagram for explaining noise generation due to offset / gain correction in a saturation region.

FIG. 6 is a diagram for explaining the principle of the embodiment of the present invention.

[Explanation of symbols]

1 X-ray generator 2 subject 5 X-ray flat panel sensor 6 A / D converter 7 memory 8 branching means 9 memory 10 memory 11 Difference block 12 memories 13 tables 14 Branching means 15 memory 16 memory 17 Difference block 18 memory 21 decoder 22 Signal path switching means 23 Decoder 24 signal path switching means 31, 32 processing blocks

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 4C093 AA30 CA06 CA21 EB17 EB30                       FA33 FA43 FA45 FC11 FC17                       FC18 FC19 FD01 FD13                 5B047 AA17 AB02 BA02 BB04 BC14                       CB22 DA06 DC09                 5B057 AA08 BA03 BA12 CA02 CA08                       CA12 CA16 CB02 CB08 CB12                       CB16 CE02 DC22                 5C077 LL02 MM03 NN02 PP12 PP14                       PP15 PQ03 SS01

Claims (12)

[Claims] 1. A sensor including a plurality of pixels for converting a subject image into an electric signal, a correction unit for correcting a variation for each pixel, and an output signal before correction of a stage before the correction unit is evaluated. An image processing device, comprising: a control unit that determines the saturation of each element and, if saturated, fixes the corrected output value to a constant value. 2. The image processing apparatus according to claim 1, wherein the subject image has a radiant energy distribution. 3. The image processing apparatus according to claim 2, wherein the radiant energy distribution is a spatial distribution of X-ray dose. 4. The variation of each pixel is a variation of an offset or gain characteristic in each pixel and an electrical circuit that handles an electrical signal output from the pixel, according to any one of claims 1 to 3. The image processing device described. 5. The control means converts the signal value into a digital value when judging the saturation, and judges that the digital value is equal to or more than a predetermined value. The image processing device according to any one of items. 6. A sensor including a plurality of pixels for converting a subject image into an electric signal, a correction means for correcting a variation for each pixel, and an output signal before the correction in the preceding stage of the correction means is evaluated. An image processing apparatus, comprising: a control unit that determines the saturation of each element and does not perform correction when the element is saturated. 7. A sensor including a plurality of pixels for converting a subject image into an electric signal, a correction unit for correcting variations in each pixel of the sensor, and an output value of each pixel of the image sensor before correction. An image processing apparatus, comprising: a control unit that evaluates saturation and adds information indicating a saturated case to information of each pixel. 8. The control means discards the correction result and fixes it at a constant value when the information added to the pixel indicates saturation after the correction by the correction means. The image processing apparatus according to claim 7. 9. The image processing according to claim 7, wherein the control means fixes the output value of the pixel to a constant value when the information added to the pixel indicates saturation. apparatus. 10. A conversion step of converting a subject image into an electric signal by a plurality of pixels, and when correcting the variation for each pixel, the output signal before the correction is evaluated to determine the saturation of each element, And a correction step of fixing the corrected output value to a constant value when the image processing method is saturated. 11. A conversion step of converting an object image into an electric signal by a plurality of pixels, and when correcting a variation for each pixel, an output signal before the correction is evaluated to judge saturation of each element. And an image correction method which does not perform correction when saturated. 12. A conversion step of converting a subject image into an electric signal by a plurality of pixels, a correction step of correcting the variation for each pixel, a saturation of an output value before correction for each pixel is evaluated, and each pixel is evaluated. And a saturation evaluation step of adding information indicating that the information is saturated, to the image processing method.