JP2001144932A - Image processor and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processor that inserts a prescribed pattern being an electronic watermark to an X-ray image without revising its important part. SOLUTION: A radiation area extract section 302 extracts an X-ray radiation area of an X-ray image, analyzing by a histogram section 303 detects a void image space in its radiation field. The prescribed pattern is inserted to the detected void image space, the X-ray image to which the prescribed pattern is inserted is subject to discrete wavelet transform, a quantization section 3 quantizes the transform coefficient and an entropy coding section 4 applies entropy coding to the output of the quantization section 3.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image processing apparatus and method for encoding an image signal.

[0002]

2. Description of the Related Art Radiation (X-ray, α-ray,
(β-rays, γ-rays, electron beams, ultraviolet rays, etc.), a part of this radiation energy is accumulated in the phosphor, and when this phosphor is irradiated with excitation light such as visible light, It is known that a phosphor emits photostimulated light in accordance with stored energy, and a phosphor exhibiting such properties is called an accumulating phosphor (stimulable phosphor). Using such a stimulable phosphor, radiation image information of a subject such as a human body is temporarily recorded on a stimulable phosphor sheet, and the stimulable phosphor sheet is scanned and irradiated with excitation light such as laser light. To emit light. The emitted light is photoelectrically read to obtain an image signal, and based on the image signal, a radiation image of a subject is output as a visible image on a recording material such as a photographic photosensitive material or a display device such as a CRT. An information recording / reproducing system has already been proposed by the present applicant (JP-A-55-12429, JP-A-56-1).
No. 1395).

In recent years, similarly to the case described above, a semiconductor sensor is used to generate X
Devices for capturing line images have been developed. These systems have the practical advantage of being able to record images over a very wide radiation exposure area as compared to conventional radiographic systems using silver halide photography. That is, an X-ray having a very wide dynamic range is read by a photoelectric conversion unit and converted into an electric signal, and the electric signal is used to display a radiation image on a recording material such as a photographic photosensitive material or a display device such as a CRT. By outputting the image as an image, it is possible to obtain a radiation image which is not affected by the fluctuation of the exposure amount of radiation.

[0004]

Since such an X-ray image contains a very large amount of information, the amount of the information is enormous when storing and transmitting the image information. There is a problem. For this reason, when storing and transmitting such image information, the redundancy of the image is removed or the deterioration of the image quality is hardly visually recognized.
High-efficiency coding is used to reduce the amount of image information by changing the content of the image.

For example, in JPEG recommended by the ISO and ITU-T as an international standard encoding method for still images, DPCM is used for lossless compression, and Discrete Cosine Transform (DCT) is used for lossy compression. ing. For details on JPEG, see Recommendation ITU-T.
Recommendation T. 81 | ISO / IEC 1091
8-1, etc., and are omitted here.

In recent years, discrete wavelet transforms (DW
Many studies have been made on compression methods using (T-transformation). The feature of the compression method using this DWT transform is that
The point is that the blocking artifacts seen in the DCT transform do not occur.

On the other hand, when an X-ray image is digitized as described above and transferred or stored in an electronic medium,
During the intermediate process, data may be accidentally or intentionally falsified, which may lead to misdiagnosis, loss of evidence capacity in medical litigation, or loss of evidence. In order to prevent such tampering of the image, it is conceivable to use a digital watermark that has been conventionally proposed. However, since such a digital watermark can be watermarked in the entire image, In the case of diagnosing a disease or the like on the basis of the above, there is a possibility that an important area (portion) may be changed, which may hinder the diagnosis. Therefore, a more appropriate falsification prevention technology is required. Was.

The present invention has been made in view of the above conventional example, and detects a blank region included in an X-ray image and inserts a predetermined pattern corresponding to a digital watermark into the region, thereby obtaining an X-ray image. It is an object of the present invention to provide an image processing apparatus and method capable of inserting a predetermined pattern into an image without changing important parts.

Another object of the present invention is to provide a method in which, when an X-ray image does not include a blank region, a predetermined pattern is set around the irradiation field region or the highest frequency component obtained by discrete wavelet transform of the X-ray image. It is an object of the present invention to provide an image processing apparatus and method capable of inserting a predetermined pattern serving as an electronic watermark without changing an important part of an X-ray image by inserting the pattern.

[0010]

In order to achieve the above object, an image processing apparatus according to the present invention has the following arrangement.
That is, an irradiation field detecting means for detecting an X-ray irradiation area of the input X-ray image, a clear area detecting means for detecting a clear area in the irradiation field detected by the irradiation field detecting means, An insertion unit for inserting a predetermined pattern into a blank region detected by the detection unit, and an encoding unit for encoding an X-ray image in which the predetermined pattern is inserted by the insertion unit.

To achieve the above object, the image processing apparatus of the present invention has the following arrangement. That is, the input X
An irradiation field detecting means for detecting an X-ray irradiation area of a line image; a clear area detecting means for detecting a clear area in the irradiation field detected by the irradiation field detecting means; And a coding unit for coding an X-ray image in which the predetermined pattern has been inserted by the insertion unit.

In order to achieve the above object, an image processing apparatus according to the present invention has the following arrangement. That is, the input X
Irradiation field detection means for detecting an X-ray irradiation area of the X-ray image, conversion means for performing discrete wavelet transform on the X-ray image,
A blank area detecting means for detecting a blank area in the irradiation field detected by the irradiation field detecting means, and a conversion coefficient converted by the converting means when the clear area is not detected by the clear area detecting means. Insertion means for inserting a predetermined pattern into the highest frequency component of the X-ray image, quantization means for quantizing the transform coefficient into which the predetermined pattern has been inserted by the insertion means, and coding of the X-ray image quantized by the quantization means. Encoding means for encoding.

[0013] To achieve the above object, the image processing method of the present invention comprises the following steps. That is, the input X
An irradiation field detecting step of detecting an X-ray irradiation area of a line image; a blank area detecting step of detecting a blank area in the irradiation field detected in the irradiation field detecting step; And a coding step of coding an X-ray image in which the predetermined pattern has been inserted in the inserting step.

In order to achieve the above object, the image processing method of the present invention comprises the following steps. That is, the input X
An irradiation field detecting step of detecting an X-ray irradiation area of the X-ray image; a blank area detecting step of detecting a blank area in the irradiation field detected in the irradiation field detecting step; When a missing area is not detected, an insertion step of inserting a predetermined pattern around the irradiation field, and an encoding step of encoding an X-ray image in which the predetermined pattern is inserted in the insertion step, I do.

[0015] To achieve the above object, the image processing method of the present invention comprises the following steps. That is, the input X
An irradiation field detection step of detecting an X-ray irradiation area of the X-ray image, and a conversion step of performing a discrete wavelet transform of the X-ray image,
A blank area detection step of detecting a blank area in the irradiation field detected in the irradiation field detection step, and a conversion coefficient converted by the conversion step when the blank area is not detected in the blank area detection step. An insertion step of inserting a predetermined pattern into the highest frequency component, a quantization step of quantizing a transform coefficient into which the predetermined pattern has been inserted in the insertion step, and coding an X-ray image quantized in the quantization step. And an encoding step of encoding.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

The feature of the present embodiment is that a transparent area is detected from an irradiation area in an X-ray image, and a predetermined pattern is written into the transparent area at a level that is difficult to visualize. A means for detecting tampering or destruction. However, depending on the X-ray photographing, there is a case where the plain region does not exist. In this case, a predetermined pattern is written with a small value in the sub-band HH that has the least effect on diagnosis, or the irradiation area is selected to be relatively large, and the enlarged part is referred to as Foreground.
A predetermined pattern is embedded in the Foreground. In addition, a part of the irradiation area excluding the blank area is defined as a region of interest, and an image corresponding to this region of interest is level-shifted and then encoded, so that the region of interest is preferentially encoded with higher image quality. In that

The details will be described below.

FIG. 1 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention.

In FIG. 1, reference numeral 1 denotes an image input unit for inputting image data, which includes, for example, a scanner for reading a document image, an image pickup device such as a digital camera, or an interface unit having an interface function with a communication line. I have. Reference numeral 2 denotes a discrete wavelet transform unit that executes a two-dimensional discrete wavelet transform on an input image. Numeral 3 denotes a quantization unit for quantizing the coefficients subjected to the discrete wavelet transform by the discrete wavelet transform unit 2. Reference numeral 4 denotes an entropy encoding unit which entropy encodes the coefficients quantized by the quantization unit 3. 5
Is a code output unit that outputs the code encoded by the entropy encoding unit 4. Reference numeral 11 denotes an area specifying unit that specifies a region of interest of the image input from the image input unit 1.

The device according to the first embodiment is the same as that shown in FIG.
The present invention can be applied to a case where a program for realizing this function is loaded into a general-purpose PC or workstation and operated, instead of a dedicated device as shown in FIG.

In the above configuration, first, the image input unit 1
, Pixel signals constituting an image to be encoded are input in raster scan order, and the output is input to the discrete wavelet transform unit 2. In the following description, the image signal input from the image input unit 1 will be described as a monochrome multi-valued image. However, if a plurality of color components such as a color image are encoded, each of the RGB color components, The luminance and chromaticity components may be compressed as the single color components.

The discrete wavelet transform unit 2 performs a two-dimensional discrete wavelet transform process on an input image signal, calculates a transform coefficient, and outputs the result.

FIGS. 2A to 2C are diagrams illustrating the basic configuration and operation of the discrete wavelet transform unit 2 according to the present embodiment.

The image signal input from the image input unit 1 is stored in the memory 201, sequentially read out by the processing unit 202, subjected to a conversion process, and written in the memory 201 again.

FIG. 2B shows the configuration of the processing in the processing section 202 in the present embodiment. In the figure,
The input image signal is separated into a signal of an even address and a signal of an odd address by a combination of a delay element 204 and a down sampler 205, and is subjected to filter processing by two filters p and u. s and d represent a low-pass coefficient and a high-pass coefficient when one-dimensional decomposition is performed on a one-dimensional image signal, respectively, and are calculated by the following equations. And

D (n) = x (2n + 1) -floor ((x (2n) + x (2n + 2)) / 2) (Equation 1) s (n) = x (2n) + floor (( d (n-1) + d (n)) / 4 (Equation 2) where x (n) is an image signal to be converted, and in the above equation, floor {X} is the maximum value not exceeding X. Represents an integer value.

With the above processing, a one-dimensional discrete wavelet transform process is performed on the image signal from the image input unit 1. The two-dimensional discrete wavelet transform uses this 1
The two-dimensional discrete wavelet transform is sequentially performed in the horizontal and vertical directions of the image, and details thereof are known, and thus description thereof is omitted here.

FIG. 2C is a diagram showing an example of the configuration of a two-level transform coefficient group obtained by the two-dimensional discrete wavelet transform process. The image signal is a sequence of coefficients HH1, HL1, and LH1 in different frequency bands. , LL. In the following description, these coefficient sequences are called subbands. Each sub-band obtained in this way is output to the subsequent quantization unit 3.

The region designating section 11 determines a region of interest (ROI) to be decoded with higher image quality compared to the surrounding portion in the image to be coded, and assigns the target image to a discrete wavelet. Upon conversion, mask information indicating which coefficients belong to the region of interest is generated.
The details of the area specifying unit 11 will be described later.

FIG. 3A is a diagram for explaining the principle of generating mask information.

As shown on the left side of FIG. 3A, when a star-shaped region is designated as the region of interest (hereinafter, designated region), the region designation section 11 converts the image including the designated region into a discrete wavelet. Calculate the portion occupied in each subband when the conversion is performed. The area indicated by the mask information is
This is a range including surrounding conversion coefficients necessary for restoring an image signal on the boundary of the designated area. In addition, FIG.
A portion 331 indicated by a cloud on the left side of (a) is a predetermined pattern corresponding to a digital watermark. It is placed in a position that does not overlap the region of interest that is important for diagnosis.
The predetermined pattern may be an arbitrary pattern such as a figure, but in the case of a medical image, it is conceivable that the patient information is changed into a bit pattern and inserted repeatedly (see FIG. 7). Also, by inserting the patient information, the correspondence between the medical image and the patient becomes clear, and even if the patient information in the image header is damaged or falsified due to accident or intention, the true patient can be specified. effective. The predetermined pattern of the digital watermark inserted as described above does not need to be reflected in the mask information.

An example of the mask information calculated in this way is shown on the right side of FIG. In this example, FIG.
The mask information when the two-level discrete wavelet transform is performed on the image on the left of is calculated as shown in the figure. In this figure, a star-shaped portion 330 is a designated area, and the bits of the mask information in this designated area are “1”, and the other bits of the mask information are “0”. Since the entire mask information has the same structure as that of the transform coefficients obtained by the two-dimensional discrete wavelet transform, the bits in the mask information are examined to determine whether the coefficient at the corresponding position belongs to the specified area. can do.
The mask information generated in this way is output to the quantization unit 3.

Further, the area designating section 11 inputs parameters for designating image quality for the designated area from an input system (not shown). This parameter may be a numerical value representing the compression ratio assigned to the designated area or a numerical value representing the image quality. In this case, the compression ratio to be assigned can be determined based on the region information of the captured image. For example, in the case of an X-ray image, the imaging part information is information indicating an imaging part and a direction, such as a chest front image and a head side image. This information is
The information may be input by an operator through an operation panel or the like of the image input unit 1 (not shown), or may be transferred from the radiation information system prior to imaging. In general, since a chest image contains soft tissue, it is desirable not to increase compression rate much, and a bone image such as a head does not deteriorate significantly even if compression rate is increased. The area designating unit 11 calculates the bit shift amount B for the coefficient in the designated area from these parameters, and outputs it to the quantization unit 3 together with the mask.

Next, the configuration of the area specifying section 11 for automatically determining the specified area (region of interest) will be described in detail.

As shown in FIG. 1, the area designation section 11 includes an image reduction section 301, an irradiation area extraction section 302, a histogram analysis section 303, a binarization processing section 304, and a morphology processing section 305. In the image reducing unit 301, (26
For an input image of 88 × 2688 pixels, (336 × 2688)
336) Output a reduced image of about pixels. In order to shorten the calculation time of the subsequent processing, the pixel value of the input image is set to 12
It is also conceivable that the lower 4 bits of the 12 bits are deleted and the image data is reduced and converted into 8-bit image data.

The irradiation area extraction unit 302 extracts how the X-ray incidence area is distributed over the entire input image. The X-ray incident area may be distributed over the entire surface of the input image, or may be irradiated with X-rays on a part (in this case, there is an aperture stop).

First, the determination of the presence or absence of the irradiation aperture will be described with reference to FIGS.

FIG. 4A is a diagram showing an example of an input image. Here, if the input image area 400 has an irradiation aperture and there is an unirradiated part of X-rays, it is considered that the part is in the peripheral area of the image. For this reason, the average value of the pixel values in the peripheral area of the input image area 400 is compared with the average pixel value of the central part of the input image. Empirically, when the peripheral average pixel value is smaller than the central average pixel value by about 5% or more, it can be determined that the image has an aperture stop. In FIG. 4A, reference numeral 401 denotes an X-ray irradiation area, and 402 denotes a region of interest.

FIG. 4B is a diagram showing an example of the peripheral area 403 and the central area 404 in the input image area 400.

When there is an aperture stop, several profiles are extracted for each of the vertical and horizontal directions of the input image area 400. Two peak points are extracted from the secondary differential values of these extracted profiles. Then, the coordinates of the peak value of the second derivative with respect to the plurality of profiles are obtained, and the average line segment is obtained to obtain the line segment of the irradiation region.

FIG. 4C is a diagram showing an example of extracting a horizontal profile 405 and a vertical profile 406 as an example of a profile position.

FIG. 5A shows an example of secondary differential peak detection, where 407 indicates a profile, and dotted line 408 indicates secondary differential. FIG. 5B shows the detection positions of each profile, and these detection positions are indicated by circles.
FIG. 5C shows the irradiation area 409 finally extracted.
Is shown.

The histogram analyzer 303 calculates the frequency of pixel values for the region extracted as the irradiation region by the irradiation region extractor 302. Here, the correspondence between the pixel value and the input amount of X-ray is such that the larger the pixel value, the larger the incident amount. Based on the analysis of the histogram, it is determined whether or not there is any omission. Here, when there is a blank, there are two peaks, and it can be determined based on the two peaks.

In general, in imaging without any omissions even though there are aperture stops at the abdomen, chest, etc., only one peak appears despite the presence of bones and soft tissues. Techniques for detecting the number of these peaks are:
Assuming that the histogram is a waveform, a low-pass filter is applied to the histogram, and then secondary differentiation processing is performed. If the value of the secondary differentiation processing exceeds a threshold set empirically, it is determined that there is a peak. In rare cases, no peak may be detected, or three or more peaks may be detected. If no peak is detected, it is determined that there is no omission. If three or more are detected, two of the larger ones are selected, and a peak having a large pixel value is determined to be omission.

FIG. 6 shows a histogram of the irradiation region (region of interest) 409 and a peak 601 of the detected blank region.
FIG. Here, when it is determined that there is a skip, the binarization processing unit 304 performs a binarization process using the peak value SP of the skip. Also, the peak value S
It passes through P and above, and determines below that as the shooting target area.

Next, since there is a possibility that an isolated point exists or a blank region remains by the binarization processing unit 304, the morphology processing unit 305 performs a filtering process. Erosion is performed for about 3 to 5 pixels in order to remove the isolated point and the remaining transparent region. Thereafter, a labeling process is performed to limit the area to one continuous area. In this state, since there is a possibility that one continuous area has a hole, the hole is filled by performing a closing process.
The output result is a region of interest obtained by removing a blank region from the irradiation region 409. In the following description, it is assumed for convenience that the region of interest is the star 330 shown in FIG.

FIG. 7A shows an example in which a bitmap pattern of a patient ID and a patient name is inserted into a portion other than the region of interest 402 of the input X-ray image. FIG. 7 (A) shows the background region of the X-ray image so as to be prominently visible, but is actually inserted into the blank portion at a value close to the blank level.

In the following description, for convenience, the region of interest 330 will be described as the star 330 shown in FIG. 3A, and the digital watermark pattern 331 will be described as a cloud.

The quantizing section 3 quantizes the inputted coefficient by a predetermined quantization step, and outputs an index for the quantized value. Here, the quantization is performed by the following equation.

Q = sign (c) floor (abs (c) / Δ) (Equation 3) sign (c) = 1; c ≧ 0 (Equation 4) sign (c) = − 1; c <0 (Equation 5) Here, c is a coefficient to be quantized. In the present embodiment, the value of Δ includes “1”. In this case, no quantization is actually performed.

Next, the quantization unit 3 changes the quantization index based on the mask and the shift amount B input from the area specifying unit 11 by the following equation.

Q ′ = q × 2 ^ B; m = 1 (Equation 6) q ′ = q; m = 0 (Equation 7) Here, m is the value of the mask at the position of the quantization index. By the above processing, only the quantization index belonging to the space area designated by the area designation unit 11 is shifted up by B bits.

FIGS. 3B and 3C are diagrams for explaining the change of the quantization index due to the shift up. In FIG. 3B, when three quantization indices exist in each of the three subbands, the value of the mask in the networked quantization index is “1”, and the shift number B is “2”. , And the post-shift quantization index is as shown in FIG.

Further, a digital watermark pattern 3 indicated by a cloud shape
31 is a light gray bit 33 in FIG.
As shown by 2, it is inserted into the least significant bit. The quantization index changed in this way is output to the entropy encoding unit 4 that follows.

If there is no transparent region inside the irradiation region, there are two possible solutions. One is FIG.
In this method, the digital watermark pattern 331 is embedded in the least significant bit of the sub-band HH1 as shown in FIG. By doing so, it is possible to insert a watermark at a level that has the least effect during decoding. In this case, the region of interest is not particularly set, that is, the entire irradiation region can be considered as the region of interest.

In the second method, as shown in FIG. 7B, a digital watermark frame (For
eground) 410 is provided (however, in this example, an image having a clear area is used).
This is the same as expanding the irradiation region (image region) 409 and setting the region of interest therein for convenience. In the figure, Fo
The density of the reground 410 is preferably selected to be black or low luminance in the case of a medical image, but in this example, it is shown as white (high luminance) for convenience.

The quantization index changed in this way is output to the entropy encoder 4 that follows.

The entropy coding unit 4 decomposes the quantization index input from the quantization unit 3 into bit planes, performs binary arithmetic coding for each bit plane, and outputs a code stream.

FIG. 9 is a diagram for explaining the operation of the entropy coding unit 4. In this example, three non-zero quantization indexes exist in a subband having a size of 4 × 4. "+13" and "-
6 ”and“ +3. ”The entropy coding unit 4 scans this area to scan the maximum value M (“ 1 ”in this example).
3 "), and the number of bits S required to represent the maximum quantization index is calculated by the following equation.

S = ceil (log 2 (abs (M))) (Equation 8) where ceil (x) represents the smallest integer value among integers equal to or greater than x.

In FIG. 9, since the maximum coefficient value is "13", the number of bits S representing this is "4", and the 16 quantization indexes in the sequence are as shown on the right side of FIG. Processing is performed in units of four bit planes. First, the entropy coding unit 4 performs binary arithmetic coding on each bit of the most significant bit plane (represented by the MSB in the figure), and outputs it as a bit stream. Next, the bit plane is lowered by one level, and similarly, each bit in the bit plane is encoded and output to the code output unit 5 until the target bit plane reaches the least significant bit plane (represented by LSB in the figure). At this time, the code of each quantization index is entropy-coded immediately after the first non-zero bit is detected in the bit plane scanning.

There are two methods of this entropy coding, a spatial scalable method and an SNR scalable method. With the spatial scalable method, it is possible to improve the image quality from low resolution images to high images when transferring and expanding, and with SNR scalable, it is possible to display while improving the image quality with the same spatial resolution become.

First, spatial scalability will be described.

FIG. 10 is a schematic diagram showing the structure of a code string generated and output as described above.

FIG. 10A shows the entire structure of the code string, where MH is the main header, and THi (i = 0 to n-
1) indicates a tile header, and BSi (i = 0 to n-1) indicates a bit stream. As shown in FIG. 2B, the main header MH has the size of the image to be encoded (the number of pixels in the horizontal and vertical directions), the size when the image is divided into a plurality of rectangular tiles, and the number of each color component. , The size of each component, and component information indicating the bit precision. In the present embodiment, since the image is not divided into tiles, the tile size and the image size take the same value, and when the target image is a monochrome multi-valued image, the number of components is “1”.

Next, the structure of the tile header TH is shown in FIG.
It is shown in (c). The tile header TH includes a tile length including a bit stream length and a header length of the tile, an encoding parameter for the tile, mask information indicating a designated area, and a bit shift number for a coefficient belonging to the area. The coding parameters include the level of the discrete wavelet transform, the type of filter, and the like.

FIG. 10D shows the configuration of a bit stream according to the present embodiment. In FIG. 10, the bit stream is grouped for each sub-band, and the sub-band (LL) having a lower resolution is sequentially set as the head. Are arranged in the order of increasing. Further, in each subband, codes are arranged in units of bit planes from the upper bit plane (bit plane S-1) to the lower bit plane (bit plane 0).

With such a code arrangement, it is possible to perform hierarchical decoding as shown in FIG. 15, which will be described later.

Next, SNR scalable will be described.

FIG. 11 is a schematic diagram for explaining the structure of a code string generated and output at the time of SNR scalability.

FIG. 11A shows the entire structure of a code string, where MH is a main header and THi (i = 0 to 0).
n-1) is a tile header, and BSi (i = 0 to n-1) is a bit stream. The main header MH includes the size of the image to be encoded (the number of pixels in the horizontal and vertical directions), the tile size when the image is divided into tiles as a plurality of rectangular areas, and each color, as shown in FIG. Component information indicating the number of components indicating the number of components, the size of each component, and bit precision is provided. In the present embodiment, since the image is not divided into tiles, the tile size and the image size take the same value. When the target image is a monochrome multi-valued image, the number of components is “1”.
It is.

Next, the structure of the tile header TH is shown in FIG.
It is shown in (c).

The tile header TH includes a tile length including a bit stream length and a header length of the tile, an encoding parameter for the tile, mask information indicating a designated area, and a bit shift for a coefficient belonging to the area. Have a number. The coding parameters include the level of the discrete wavelet transform, the type of the filter, and the like.

FIG. 11D shows the configuration of a bit stream according to the present embodiment. The bit stream is grouped in units of bit planes, and is shifted from the upper bit plane (bit plane S-1) to the lower bit plane (bit plane). 0). In each bit plane, the result of encoding the bit plane of the quantization index in each subband is sequentially arranged in subband units. In the figure, S
Indicates the number of bits required to represent the maximum quantization index. The code string generated in this way is output to the code output unit 5.

With such a code arrangement, hierarchical decoding as shown in FIG. 16 described later can be performed.

In the above-described embodiment, the compression ratio of the entire image to be encoded can be controlled by changing the quantization step Δ.

As another method, in the present embodiment, the lower bits of the bit plane to be encoded in the entropy encoding unit 4 can be restricted (discarded) in accordance with a required compression ratio. In this case, not all bit planes are coded, and the bit planes from the upper bit plane to the number of bit planes corresponding to the desired compression ratio are coded and included in the final coded sequence.

As described above, by adopting the function of limiting the lower-order bit plane, only the bits corresponding to the designated area shown in FIG. 3 are included in the code string. That is, by compressing only the designated area at a low compression rate, it is possible to encode a high-quality image.

FIG. 17 is a flowchart showing an encoding process and a digital watermark embedding process in the image encoding apparatus according to the present embodiment.

First, in step S 1, the irradiation field area of the X-ray image is detected from the X-ray image input from the image input section 1 by the irradiation area extraction section 302 in the area specification section 11. Next, the process proceeds to step S2, where the histogram unit 303
It is determined whether or not there is a blank area in the irradiation field area based on the analysis according to. If there is, the process proceeds to step S3, and an electronic watermark is embedded in the blank region. Next, in step S4, a discrete wavelet transform is performed on the image, and the converted coefficient is quantized in step S5. Then, in step S6, the quantization result is entropy-coded.

On the other hand, if it is determined in step S2 that there is no plain area in the irradiation field area, the flow advances to step S7 to determine whether or not to embed a digital watermark in a peripheral area of the irradiation field area. If so, the process proceeds to step S8 to expand the irradiation field area, embed a digital watermark in the expanded peripheral area (see FIG. 7B), and perform discrete wavelet transform on the embedded image (S4). ). Then, the process proceeds to step S5, where the transform coefficient is quantized, and the process proceeds to step S6.
To perform entropy encoding.

If it is determined in step S7 that the image is not embedded in the surrounding area, the flow advances to step S9 to perform a discrete wavelet transform on the input image, and the subband H
An electronic watermark is embedded in H1 (see FIG. 8). Then, the process proceeds to step S5, where the transform coefficient is quantized, and the process proceeds to step S5.
In step 6, entropy coding is performed. The encoded code sequence is output from the code output unit 5.

Next, a method of decoding a bit stream encoded by the above-described image encoding apparatus will be described.

FIG. 12 is a block diagram showing the configuration of an image decoding apparatus according to the present embodiment. 6 is a code input section, 7 is an entropy decoding section, 8 is an inverse quantization section, and 9 is an inverse discrete wavelet transform. The unit 10 is an image output unit.

The code input unit 6 inputs, for example, a code string coded by the above-described coding apparatus, analyzes a header included in the code string, extracts parameters necessary for the subsequent processing, and performs processing if necessary. Or sends the corresponding parameters to the subsequent processing unit. Further, the bit stream included in the input code sequence is output to the entropy decoding unit 7.

The entropy decoding unit 7 decodes and outputs the bit stream in bit plane units. FIG. 13 shows the decoding procedure at this time.

On the left side of FIG. 13, one region of a subband to be decoded is sequentially decoded in bit plane units.
This shows a flow of finally restoring the quantization index, and the bit planes are decoded in the order of the arrows in FIG. The restored quantization index is output to the inverse quantization unit 8.

The inverse quantization unit 8 restores discrete wavelet transform coefficients from the input quantization index based on the following equation.

C ′ = Δ × q / 2 ^ U; q ≠ 0 (Equation 9) c ′ = 0; q = 0 (Equation 10) U = B; m = 1 (Equation 11) U = 0; m = 0 (Equation 12) Here, q is a quantization index, Δ is a quantization step, and Δ is the same value used at the time of encoding. B is the number of bit shifts read from the tile header, and m is the value of the mask at the position of the quantization index. c ′ is a restored transform coefficient, which is a restored coefficient represented by s or d at the time of encoding. The transform coefficient c ′ is output to the subsequent inverse discrete wavelet transform unit 9.

FIG. 14 shows an inverse discrete wavelet transform unit 9
FIG. 1 is a block diagram showing the configuration and processing thereof.

In FIG. 10A, the input transform coefficients are stored in a memory 901. The processing unit 902 performs two-dimensional inverse discrete wavelet transform by performing one-dimensional inverse discrete wavelet transform, sequentially reading out transform coefficients from the memory 901 and performing processing. The two-dimensional inverse discrete wavelet transform is executed in the reverse procedure of the above-described forward discrete wavelet transform, but the details are known and will not be described.

FIG. 13B shows a processing block of the processing unit 902. The input conversion coefficient is subjected to two filter processes of u and p,
After being up-sampled by 01, they are superimposed to output an image signal x ′. These processes are performed by the following equations.

X ′ (2n) = s ′ (n) −floor ((d ′ (n−1) + d ′ ((n)) / 4) (Equation 13) x ′ (2n + 1) = d ′ (n) + floor ((x ′ (2n) + x ′ (2n + 2)) / 2) (Expression 14) Here, (Expression 1), (Expression 2), (Expression 13), and (Expression 1)
Since the forward and backward discrete wavelet transforms according to 4) satisfy the perfect reconstruction condition, the quantization step Δ is “1” in the present embodiment, and all the bit planes are decoded in the bit plane decoding. If it has been decoded, the restored image signal x 'is the signal x of the original image.
Matches.

The image is restored by the above processing and output to the image output unit 10. Here, the image output unit 10 may be an image display device such as a monitor or a storage device such as a magnetic disk.

Next, a description will be given of an image in the case of encoding by spatial scalability.

A display mode of an image when an image is restored and displayed according to the above-described procedure will be described with reference to FIG.

FIG. 10A shows an example of a code string, and the basic configuration is based on FIG. Here, the entire image is set as a tile, and therefore, only one tile header (TH0) and bit stream (BS0) are included in the code string. This bit stream (B
In S0), as shown in the figure, codes are arranged in order of increasing resolution from LL, which is a subband corresponding to the lowest resolution, and in each subband, an upper bit plane (bit plane S- The codes are arranged from 1) to the lower bit plane (bit plane 0).

The decoding device sequentially reads the bit stream and displays an image when the code corresponding to each bit plane is decoded. FIG. 4B shows each sub-band,
FIG. 9 is a diagram illustrating a size of an image displayed corresponding thereto and a change in a reproduced image accompanying decoding of a code string of each subband. In the figure, the code sequence corresponding to LL is sequentially read, and the image quality is gradually improved as the decoding process of each bit plane progresses. At this time, the portion of the star 330 that has become the designated region (region of interest) at the time of encoding is restored with higher image quality than the other portions.

This is because the quantization unit 3 shifts up the quantization index belonging to the designated area at the time of encoding, so that the quantization index is faster than the other parts in the bit plane decoding. This is because it is decoded at the time. The reason why the designated area portion is decoded with high image quality is the same for other resolutions.

Further, when all the bit planes are decoded, the designated area and the other parts are the same in terms of image quality. However, if the decoding is discontinued at an intermediate stage, the designated area is replaced with the other parts. An image restored with higher image quality than the area is obtained. If the display is terminated during the decoding, the digital watermark pattern is not decoded.

Further, when a digital watermark is inserted only into the sub-band HH1 as a modification, the digital watermark can be recognized only when the display is performed at the highest resolution. Even in this case, if the decoding process is terminated during the decoding, the digital watermark is not displayed.

Next, a description will be given of expansion in the case of encoding with SNR scalable.

A display mode of an image when an image is restored and displayed by the above-described procedure will be described with reference to FIG.

FIG. 11A shows an example of a code string. The basic configuration is based on FIG. 11, but here the entire image is set as a tile, and therefore, the code string is included in the code string. Contains only one tile header (TH0) and bit stream (BS0). This bit stream BS
0, as shown in the figure, from the most significant bit plane (bit S-1) to the least significant bit plane (bit 0).
The symbols are arranged toward.

The decoding device sequentially reads the bit stream and displays an image when the code of each bit plane is decoded. In FIG. 3B, the image quality is gradually improved from bit S-1 to bit S-2,..., Bit 0 as the decoding process of each bit plane progresses. Star-shaped part 3
30 is restored to a higher image quality than the other parts. In addition, the digital watermark 331 can be viewed only when displayed at the highest image quality.

As described above, this is because the quantization unit 3 shifts up the quantization index belonging to the specified area when performing encoding by the encoding apparatus, so that the quantization This is because the decryption index is decoded earlier with respect to the other parts.

Further, when all the bit planes have been decoded, the designated area 330 and the other parts are the same in terms of image quality. An image restored with higher image quality than the area is obtained. Similarly, the digital watermark 331 is not displayed when the decoding is discontinued on the way.

In the above-described embodiment, the amount of coded data to be received or processed is reduced by limiting (ignoring) the lower bit planes to be decoded in the entropy decoding unit 7, and consequently the compression ratio is controlled. It is possible. By doing so, it is possible to obtain a decoded image of a desired image quality only from encoded data of a necessary data amount. If the quantization step Δ at the time of encoding is “1” and all the bit planes have been decoded at the time of decoding, lossless encoding / decoding at which the restored image matches the original image is realized. You can also.

If the function of limiting the lower bit plane is used, the code string to be decoded contains only bits corresponding to the region of interest shown in FIG. 3 more than the other regions. Specifically, it is possible to achieve the same effect as decoding image data obtained by encoding only the designated area 330 as a high-quality image with a low compression ratio.

As described above, the encoding using the discrete wavelet has been generalized, but in the present embodiment,
It is characterized in that an irradiation area is detected, and a part obtained by removing a transparent area from the irradiation area is compressed as a region of interest (designated area) with high quality (or reversibly compressed).

Even if the present invention is applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), a device composed of one device (for example, a copying machine, a facsimile machine, etc.) ) May be applied.

Further, an object of the present invention is to supply a storage medium (or a recording medium) in which a program code of software for realizing the functions of the above-described embodiments is recorded to a system or an apparatus, and to provide the system or the apparatus with the recording medium. It is also achieved by a computer (or CPU or MPU) reading and executing the program code stored in the storage medium.

In this case, the program code itself read from the storage medium implements the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention. By executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also an operating system (OS) running on the computer based on the instruction of the program code. This also includes a case where some or all of the actual processing is performed and the functions of the above-described embodiments are realized by the processing.

Further, after the program code read from the storage medium is written into the memory provided in the function expansion card inserted into the computer or the function expansion unit connected to the computer, the program code is read based on the instruction of the program code. This also includes the case where the CPU provided in the function expansion card or the function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

As described above, according to the present embodiment, the watermark does not affect the diagnostic image by inserting the electronic watermark pattern into the transparent region in the irradiation region of the X-ray image.

Even if there is no blank area in the irradiation area, by inserting a watermark pattern around the irradiation field area or in a high-frequency sub-band,
Digital watermarks can be embedded.

When the embedded digital watermark pattern is embedded in a high-frequency sub-band, the decoding process is terminated and displayed, so that the embedded digital watermark pattern can be prevented from being displayed.

[0119]

As described above, according to the present invention, X
By detecting a blank area included in the line image and inserting a predetermined pattern corresponding to a digital watermark into the area,
A predetermined pattern can be inserted into an X-ray image without changing an important part of the image.

According to the present invention, when the X-ray image does not include a transparent region, a predetermined pattern is assigned to the periphery of the irradiation field region or the highest frequency component obtained by performing discrete wavelet transform on the X-ray image. The insertion has an effect that a predetermined pattern serving as a digital watermark can be inserted without changing an important part of the X-ray image.

According to the present invention, by inserting the patient information, the correspondence between the medical image and the patient becomes clear,
Even if the patient information in the image header is damaged or falsified due to an accident or intention, there is an effect that a true patient can be specified.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an image encoding device according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a configuration of a wavelet transform unit according to the present embodiment and subbands obtained by the transform.

FIG. 3 is a diagram illustrating conversion of a region of interest (designated region) in an image and bit shift of image data in the region.

FIG. 4 is a diagram illustrating an example of extracting a region of interest in an irradiation region of an X-ray image according to the present embodiment.

FIG. 5 is a diagram illustrating an example of extracting a region of interest in an irradiation region of an X-ray image according to the present embodiment.

FIG. 6 is a diagram illustrating an example of extracting a region of interest in an irradiation region of an X-ray image according to the present embodiment.

FIG. 7 is a diagram illustrating an example of embedding a digital watermark in the present embodiment.

FIG. 8 is a diagram illustrating an example of digital watermarking and embedding into subbands in the present embodiment.

FIG. 9 is a diagram illustrating an operation of an entropy encoding unit according to the present embodiment.

FIG. 10 is a schematic diagram illustrating a configuration of a code string generated and output by spatial scalability.

FIG. 11 is a schematic diagram illustrating a configuration of a code string generated and output at the time of SNR scalability.

FIG. 12 is a block diagram illustrating a configuration of an image decoding device according to the present embodiment.

FIG. 13 is a diagram illustrating bit planes and the decoding order for each bit plane by the entropy decoding unit according to the present embodiment.

FIG. 14 is a block diagram illustrating a configuration of a wavelet decoding unit according to the present embodiment.

FIG. 15 shows an example of a code string in the case of spatial scalability, each sub-band when decoding it, the size of an image displayed corresponding thereto, and the decoding of a code string of each sub-band. FIG. 9 is a diagram for explaining a change in a reproduced image accompanying the change.

FIG. 16 is a diagram illustrating an example of a code string in the case of SNR scalability and a decoding process thereof.

FIG. 17 is a flowchart illustrating a process of embedding a digital watermark in the encoding device according to the present embodiment.

Continued on the front page F-term (reference) 4C093 AA26 CA21 FD01 FD03 FD04 FD05 FD09 FF09 FF13 FF16 5B057 AA08 BA03 CA02 CA08 CA12 CA16 CB02 CB08 CB12 CB16 CE08 CG07 DA08 DB02 DB09 DC23 5C076 AA47 BA06 DA07

Claims (11)

[Claims] 1. An irradiation field detecting means for detecting an X-ray irradiation area of an input X-ray image, a transparent area detecting means for detecting a transparent area in an irradiation field detected by the irradiation field detecting means, An insertion unit for inserting a predetermined pattern into a blank region detected by a blank region detection unit, and an encoding unit for encoding an X-ray image in which the predetermined pattern is inserted by the insertion unit. Image processing apparatus. 2. An irradiation field detecting means for detecting an X-ray irradiation area of an input X-ray image, a transparent area detecting means for detecting a transparent area in the irradiation field detected by the irradiation field detecting means, An insertion unit that inserts a predetermined pattern around the irradiation field when the background region is not detected by the background region detection unit; and an encoding unit that encodes the X-ray image in which the predetermined pattern is inserted by the insertion unit. And an image processing apparatus. 3. An irradiation field detecting means for detecting an X-ray irradiation area of an input X-ray image, a converting means for performing a discrete wavelet transform on the X-ray image, and an element in the irradiation field detected by the irradiation field detecting means. A blank region detection unit that detects a blank region; and an insertion unit that inserts a predetermined pattern into the highest frequency component of the conversion coefficient converted by the conversion unit when the blank region is not detected by the blank region detection unit. A quantization unit that quantizes the transform coefficient into which the predetermined pattern has been inserted by the insertion unit; and an encoding unit that encodes the X-ray image quantized by the quantization unit. Image processing device. 4. The image processing apparatus according to claim 1, wherein the predetermined pattern includes information on a subject of the X-ray image. 5. An extraction unit for extracting a region of interest from the irradiation field based on the plain region detected by the plain region detection unit, wherein the encoding unit is extracted by the extraction unit. The image processing apparatus according to claim 1, wherein the pixel value of the region of interest is bit-shifted up and encoded. 6. An irradiation field detecting step of detecting an X-ray irradiation area of an input X-ray image, a plain area detecting step of detecting a plain area in the irradiation field detected in the irradiation field detecting step, An insertion step of inserting a predetermined pattern into a transparent area detected in the transparent area detection step, and an encoding step of encoding an X-ray image in which the predetermined pattern is inserted in the insertion step. Image processing method. 7. An irradiation field detecting step of detecting an X-ray irradiation area of the input X-ray image, a plain area detecting step of detecting a plain area in the irradiation field detected in the irradiation field detecting step, An insertion step of inserting a predetermined pattern around the irradiation field when the transparent area is not detected in the transparent area detection step; and an encoding step of encoding an X-ray image in which the predetermined pattern is inserted in the insertion step. And an image processing method. 8. An irradiation field detection step of detecting an X-ray irradiation area of an input X-ray image, a conversion step of performing a discrete wavelet transform of the X-ray image, and an element in the irradiation field detected in the irradiation field detection step. A blank area detection step of detecting a blank area, and, if the blank area is not detected in the blank area detection step, an insertion step of inserting a predetermined pattern into the highest frequency component of the conversion coefficient converted by the conversion step. A quantizing step of quantizing the transform coefficient into which the predetermined pattern has been inserted in the inserting step, and an encoding step of encoding the X-ray image quantized in the quantizing step. Image processing method. 9. The image processing method according to claim 6, wherein the predetermined pattern includes information on a subject of the X-ray image. 10. An extraction step for extracting a region of interest from the irradiation field based on the plain region detected in the plain region detection step, wherein the encoding step includes extracting the region of interest from the irradiation field. 10. The image processing method according to claim 6, wherein the pixel value of the region of interest is bit-shifted up and encoded. 11. A computer-readable storage medium storing a program for executing the image processing method according to claim 6. Description: