JP2004154567A - Image processing apparatus, method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus that eliminates speckle noise. <P>SOLUTION: The image processing apparatus sends ultrasound from an ultrasonic probe at a number of steering angles (e.g. θ1, θ2, and θ3) and generates images based on reflected signals from an object. The apparatus has an image generating section 104 that generates an image based on signals within a frequency band designated for each steering angle, θ1, θ2, and θ3; an image synthesis section 106 that combines images produced separately for each steering angle to obtain a senthesized image; and a frequency band selection section 105 that designates different frequency bands for at least two steering angles. <P>COPYRIGHT: (C)2004,JPO

Description

  The present invention relates to an image processing apparatus that non-destructively displays the inside of an inspection target such as an ultrasonic diagnostic apparatus.

2. Description of the Related Art Image processing apparatuses that irradiate a sound signal to a subject, receive a reflected signal from the subject, and image the inside of the subject based on the received signal are widely used in the industrial field and the clinical field. An ultrasonic diagnostic apparatus is a typical example of such an apparatus.
The processing speed of these image processing devices has been improved due to the recent improvements in the performance of computers, and it has become possible to measure and image subjects in real time. In fields requiring high temporal resolution, it is possible to provide a tomographic image of an inspection object in real time, and to make a biological function diagnosis using an image.

  The ultrasonic diagnostic apparatus transmits a sound signal to a subject, receives a reflected wave of the signal, and generates an image based on the signal. If there is a part in the subject that makes it difficult for the sound signal to pass through, or if there is a part (boundary) that reflects almost all of the sound signal, it is located behind the part (far away from the direction in which ultrasonic waves are transmitted and received). Cannot be obtained, and a low-luminance area called a shadow may occur.

Further, since the sound signal is a wave, interference occurs. When the object distribution is distributed with a distance difference of one wavelength or less of the sound signal from the source of the sound signal, the randomness may appear strongly due to the object distribution. False information resulting from such a cause is called speckle noise (also called sesame salty noise or spot-like noise).
In order to remove the influence of shadows or speckle noise, one composite image may be created from image data acquired from a plurality of angles (steer angles). This technique is called compound scanning. In general, compound scanning is performed by taking an average of image data obtained from a plurality of directions.

1A to 1C are diagrams showing a general compound scanning method. FIG. 1A shows a state in which an ultrasonic beam is irradiated from an ultrasonic array probe in a plurality of directions.
When there is an inspection object b1 as shown in FIG. 1B, when scanning from multiple directions c1, c2, and c3 by electric steering using the ultrasonic array probe 13 as shown in FIG. 1C, Image data c4, c5, and c6 as shown in FIG. 1C can be obtained. Each of the image data includes (i) a portion c7 in which the contour of the inspection object is drawn with high brightness, (ii) a portion c8 in which the reflected signal is drawn with low brightness because the reflected signal is weak, and (iii) a sound. There is a shadow portion c9 where a signal does not reach or a reflected signal cannot be detected. Since the images c4 to c6 are obtained from different angles, the appearances of the (i) high-luminance part, (ii) low-luminance part, and (iii) shadow part are different.

  By combining images obtained from these multiple directions, it is possible to generate an image with less missing information. As shown in an example of the synthesis result c11 shown in FIG. 1C, it can be seen that a sufficient effect can be obtained only by the conventional method to improve the above (ii) and (iii). In general, compound scanning is performed by taking an average of image data obtained from a plurality of directions. With this method, the shadow area can be narrowed, and white noise can be removed.

The following documents relate to these conventional techniques.
JP-A-09-094248 Japan Electronic Machinery Manufacturers Association "Revised Medical Ultrasound Equipment Handbook" Published by Corona Co., Ltd., January 20, 1997

However, there is a problem that speckle noise cannot be sufficiently removed by compound scanning.
As described above, speckle noise is caused by interference between reflected waves when a distribution of scatterers having a wavelength equal to or less than the wavelength exists in the inspection object. That is, the appearance pattern is determined depending on the positional relationship between the ultrasonic array probe and the inspection object and the wavelength of the sound signal used. Acquisition of image data from different directions changes the manner of interference of reflected waves, and thus the appearance pattern of speckle noise changes, but generally the angle range that can be steered by an ultrasonic array probe is at most about 30 ° When the data from the three directions shown in FIG. 1C is acquired, the angle difference between the image data is only 15 °, so that there is no significant change in the appearance pattern of the speckle noise. I can't. In this case, the removal of speckle noise is insufficient.

  FIG. 2A shows a state of scanning an inspection object having a small depth, and FIG. 2B shows a state of scanning an inspection object having a large depth. As shown in FIGS. 2A and 2B, when the depth of the inspection object (the distance from the ultrasonic array probe) increases, the inspection object must be arranged in an area where the scan areas at three steering angles overlap. Therefore, the angle difference is further reduced, and the change in the appearance pattern of speckle noise is also reduced. In this case, speckle noise cannot be removed.

  An object of the present invention is to provide an image processing apparatus, method, and program for removing speckle noise. Another object of the present invention is to provide an image processing apparatus, method, and program capable of improving image quality such as clearly displaying an outline portion of an image.

  An image processing apparatus of the present invention is an image processing apparatus that transmits an ultrasonic wave from an ultrasonic probe and generates an image based on a reflected signal generated from a reflected wave from an object, and includes at least two of the reflected signals. Designating means for designating a frequency band, generating means for generating an image for each frequency band based on a signal of the specified frequency band among the reflected signals, and synthesizing means for synthesizing the generated image for each frequency band. Prepare. According to this configuration, by generating a plurality of images from signals having different frequency bands included in the reflected signal, the difference in the speckle noise appearance pattern of the generated image for each frequency band is increased, and further, the generated images Is synthesized, the influence of speckle noise can be reduced, and speckle noise can be removed or reduced.

  Here, the ultrasonic probe transmits ultrasonic waves at a plurality of steer angles, the specifying unit specifies different frequency bands for at least two steer angles of the plurality of steer angles, and the generating unit A configuration in which an image is generated for each frequency band based on a signal of a designated frequency band among the reflected signals for each of the steering angles, and the combining unit combines the generated images for each of the steering angles. Is also good. According to this configuration, the difference in the speckle noise appearance pattern of the generated image for each steer angle is increased by making the frequency band of the reflected signal as the generation source different for the reflected signals of at least two steer angles. Thus, it is possible to reduce or eliminate the influence of the speckle noise by reducing the influence after the combination.

  Here, the designating means may be configured to designate a different frequency band for a specific steer angle among the plurality of steer angles. According to this configuration, a specific steering angle can be realized with a simple configuration in which a different frequency band is specified.

  Further, the designation means stores a plurality of sets including at least a first set and a second set, each set including an angle group indicating a plurality of steer angles and a band group indicating a frequency band corresponding to the steer angle. The image processing apparatus controls the steer angle of the ultrasonic probe based on the angle group in one set stored in the correspondence table, and the designating unit controls the angle group in the set. And a frequency band for each steer angle is specified based on the band group corresponding to the first group. The correspondence table indicates that the steer angle difference of the angle group in the first group is smaller than the steer angle difference of the angle group in the second group. In this case, the band difference of the first band group may be set to be larger than the band difference of the second band group. According to this configuration, even if the steering angle difference is relatively small, the difference in the appearance pattern of the speckle noise is increased by increasing the difference in the frequency band, and the speckle noise can be removed or reduced. .

Here, the synthesizing unit sets the number of images generated for each steering angle by the generation unit to M, the number of pixels of each image to N, and the i-th pixel value of the m-th image among the generated images to f_m. (I) The value which does not exceed the value that f_m (i) can take may be S, and the i-th pixel value f_g (i) of the synthesized image may be synthesized according to the following pixel value calculation.
f_g (i) = S-M / ((1 / (S-f_0 (i))) + (1 / (S-f_1 (i))) + ... + (1 / (S-f_ (M-1)) (I))))
According to this configuration, in addition to removing or reducing speckle noise, the value of S is used as a reference to remove the influence of a value that is greatly different from a plurality of pixels to be combined, and the vicinity of the reference value is obtained in the image of the combined result. There is an effect that a portion having the pixel value of can be highlighted.

Further, the image processing apparatus further includes a region determining unit that determines a first region including a contour and another region based on an image for each steering angle generated by the generating unit, and the combining unit includes: The first combining operation is performed on the image in the first region with respect to the image for each steering angle generated by the generating unit, and the first combining operation is performed on the pixels in the region other than the first region. A configuration for performing a different operation may be adopted. According to this configuration, an optimal pixel calculation method is selected for each region between the first region and the other regions, and an optimal pixel calculation method is selected for each region such as a region including an outline, a low-luminance region, and other regions. It can be synthesized by pixel calculation, and a clear display can be obtained.
Further, the image processing method and the program of the present invention have the same configuration as described above.

As described above, according to the image processing apparatus of the present invention, there is an effect that speckle noise can be removed or reduced.
In addition to removing or reducing speckle noise, it is also necessary to remove the influence of values that are far apart from each other among the plurality of pixels to be combined, and to highlight parts having pixel values near the reference value in the combined image. There is an effect that can be.

Furthermore, the optimum pixel calculation method can be selected for each region, and synthesis can be performed by optimum pixel calculation for each region such as a region including an outline, a low-luminance region, and other regions, and a clear display can be obtained. .
In addition, it is possible to clearly display an image of the contour and its vicinity.
For each region of the image, synthesis can be performed according to the characteristics of the region.

  Further, the image processing method and the program according to the present invention have the same effects as described above.

(Embodiment 1)
FIG. 3 is an external view of the ultrasonic diagnostic apparatus 10 according to the first embodiment of the present invention. When displaying a tomographic image by compound scanning, the ultrasonic diagnostic apparatus 10 generates an image based on reflected signals in different frequency bands, and further synthesizes them to remove speckle noise. The hardware configuration includes a display device 11, a main body device 12, and a probe 13.

The display device 11 is a CRT or the like, and displays the obtained tomographic image, contour, measurement result, and the like in gray scale, color, or the like. A transparent touch panel may be provided on the front surface of the CRT in order to obtain an operator's instruction on a display image using a touch pen or the like.
The probe 13 is a probe including an ultrasonic vibrator and an acoustic lens for transmitting and receiving ultrasonic waves.

The main body device 12 is a computer including a transmission / reception circuit for electronic scanning by ultrasonic waves, a signal / image processing circuit including a DSP and a CPU, etc., and a switch group for interacting with an operator, a trackball, a liquid crystal display unit. And an operation panel having a mouse and the like.
FIG. 4 is a block diagram showing a main functional configuration of the ultrasonic diagnostic apparatus 10 shown in FIG. The ultrasonic diagnostic apparatus 10 includes an ultrasonic array probe 101, a transmission / reception control unit 102, a steering angle instruction unit 103, an image generation unit 104, a frequency band selection unit 105, an image synthesis unit 106, an image storage unit 107, a frame memory 108, An image display unit 109 is provided.

  The ultrasonic array probe 101 has the function of the probe 13 shown in FIG. 3, and actually transmits and receives a sound wave in response to control information from the transmission and reception control unit 102. At that time, the ultrasonic array probe 101 can electrically steer as shown in FIG. 1A by adjusting the delay amount at the time of transmission / reception in accordance with an instruction from the transmission / reception control unit 102. is there. Although a linear array probe is shown here, a convex (convex) array probe may be used. The actual setting of the steer angle is performed by an input from the steer angle instruction unit 103. With this steering function, data obtained by observing the same subject from different angle directions can be obtained.

The transmission / reception control unit 102 performs transmission / reception of sound in the ultrasonic array probe 101 and control of the steer angle according to the instruction of the steering angle instruction unit 103.
The steer angle instructing unit 103 instructs the transmission / reception control unit 102 on a plurality of steer angles in the compound scan.
For example, the steer angle instruction unit 103 internally stores the steer angle table T1 and the pointer P shown in FIG. 5, and sequentially designates a plurality of steer angles according to the entry pointed to by the pointer P. The steer angle table T1 stores an entry number (No. in the figure) and a steer angle group (a plurality of steer angles θ1, θ2, θ3) in association with each other. In the steer angle table T1 of FIG. 7, the steer angle groups are set so that the steer angle difference increases in the order of entry numbers 1, 2, and 3. The entry number held by the pointer P is set to a default value or selectively set by an operator.

The image generation unit 104 performs an amplification operation, an interpolation operation, and the like on a signal in the frequency band selected by the frequency band selection unit 105 among the signals included in the sound signal for each steering angle received under the control of the transmission / reception control unit 102. To generate an image.
The frequency band selection unit 105 selects a frequency band for each steer angle specified by the steer angle instruction unit 103 and specifies the frequency band to the image generation unit 104. At that time, the frequency band selection unit 105 selects a frequency band such that at least two are different. This is to cause the image generation unit 104 to generate an image based on signals of different frequency bands among the reflected signals, and that the appearance patterns of speckle noise are different in different frequency bands.

  As a specific configuration example, the frequency band selection unit 105 stores therein a frequency band table T2 and a pointer Q as shown in FIG. 6A, and a pointer correspondence table T3 as shown in FIG. 6B. Then, the frequency bands are sequentially designated according to the entry indicated by the pointer Q. This frequency band table T2 stores entry numbers (No. in the figure) and frequency band groups (a plurality of frequency bands ω1, ω2, ω3) in association with each other. Here, ω1 is a fundamental band having the same frequency as the sound signal transmitted from the ultrasonic array probe 101, ω2 is a second harmonic band having a frequency twice the fundamental wave, and ω3 is a third harmonic having a frequency three times the fundamental wave. It is a harmonic band.

  The pointer correspondence table T3 is a table in which the pointer P and the pointer Q are associated. In this example, pointers are set so that a steer angle group with a small steer angle difference (entry No. 1 in FIG. 5) corresponds to a frequency band group with a large band difference (entry No. 1 in FIG. 6). I correspond. That is, assuming that one entry in FIG. 5 and one entry in FIG. 6 are a set linked by the pointer correspondence table T3, the angle group of the first set (No. 1 in FIG. 5 and No. 1 in FIG. 6). Is smaller than the steer angle difference (15 °) of the angle group in the second set (No. 3 in FIG. 5 and No. 4 in FIG. 6), the first band group ( The band difference of the second band group (No. 4 in FIG. 6) is set to be larger than the band difference of the second band group (No. 4 in FIG. 6). This is because the smaller the steer angle difference, the smaller the difference in the speckle noise appearance pattern, so the larger the frequency band difference, the larger the speckle noise appearance pattern difference. is there.

The image synthesizing unit 106 synthesizes the image generated for each steering angle by the image generating unit 104 by, for example, weighting and averaging the pixels at the same position, and then combining the synthesized image with the image storage unit 107 and the frame memory 108. To be stored.
The image storage unit 107 is provided for reading out a composite image when offline and displaying the composite image on the image display unit 109 via the frame memory 108.

The frame memory 108 is a frame memory that stores a synthesized image from the image synthesis unit 106 during real-time processing or a synthesized image from the image storage unit 107 during offline processing.
The image display unit 109 corresponds to the display device 11 shown in FIG.

  FIG. 7 is a diagram illustrating the relationship between the steer angle specified by the steer angle specifying unit 103, the frequency band selected by the frequency band selection unit 105, and the image generated by the image generation unit 104.

The uppermost row (first row) in the figure shows the steer angle transmitted from the ultrasonic array probe 101 (probe 13), and in actuality, the ultrasonic array probe 101 is fixed at one position and the sound wave is transmitted in three directions. Are transmitted, but three are described in the direction for convenience.
The second row shows the frequency band ω1 (fundamental wave) of the transmission wave s1 from the ultrasonic array probe 101.

  In the third stage, the image generation unit 104 uses the signal r11 of the frequency band ω1 for the reflected signal of the steer angle θ1, and uses the signal r22 of the frequency band ω2 for the reflected signal of the steer angle θ2. In the reflected signal, the images r11, d22, and d33 are generated using the signal r33 in the frequency band ω3. This corresponds to the case where the frequency band of entry number no. 1 in the frequency band table T2 shown in FIG.

The fourth row shows the three generated images r11, r22, and r33. Since these three images have different steer angles and different frequency bands, the difference in speckle noise generation patterns is also large.
The fifth row shows the synthesis result d1 of those images. As a result, speckle noise can be sufficiently removed. That is, since there is a large difference between the speckle noise generation patterns of the three generated images, it is possible to remove or reduce the speckle noise in the synthesized image. This combination is performed, for example, by taking weighted averages of the corresponding pixels at the same position in the three images using α1, α2, and α3.

Here, α1: α2: α3 may be, for example, 1: 10: 100. This is because the ratio of the amplitude of the fundamental wave, the second harmonic, and the third harmonic included in the reflected signal is approximately 100: 10: 1.
If α1: α2: α3 is set to 1: 1: 1, the removal of speckle noise is insufficient, but a contour enhancement component due to a harmonic component is added to the image of the fundamental wave. .

FIG. 8 is another example illustrating the relationship between the steering angle, the frequency band, and the generated image. 7 differs from FIG. 7 in that the frequency band for the steering angle θ3 is not ω3 but ω1.
In the fourth row of the figure, of the three images r11, r22, and r31 generated by the image generation unit 104, two images r11 and r31 generated from the reflected waves at the steer angles θ1 and θ3 have the same frequency band ω1. Are the same, but since the difference between the steering angles (the angle difference between θ1 and θ3) is large, the difference between the patterns of occurrence of speckle noise is also large. In the two images r11 and r22 generated from the reflected waves at the steer angles θ1 and θ2, the difference between the steer angles (the angle difference between θ1 and θ2) is not large, but the frequency band is different between ω1 and ω2, so the speckle noise The difference in the pattern of occurrence of is large. The same applies to the two images r22 and r31 corresponding to the steering angles θ2 and θ3.

FIG. 9 is another example illustrating the relationship between one steering angle, a frequency band, and a generated image. 7 differs from FIGS. 7 and 8 in that the steer angle of the transmission wave is one instead of three. That is, the transmission wave s1 is transmitted at one steering angle θ2.
In the fourth row of the figure, among the three images r21, r22, and r23 generated by the image generation unit 104, the three images d21, d22, and d23 generated from the reflected waves at the steer angle θ2 all have a steer angle. Although the same, the frequency bands are different from ω1, ω2, ω3, so the speckle noise generation pattern is also different. In this regard, if the steer angles and the frequency bands of a plurality of images are different from each other as shown in FIGS. 7 and 8, it is easy to sufficiently remove speckle noise. However, even if the frequency bands are different as shown in FIG. Noise can be reduced.

  FIGS. 10A and 10B are explanatory diagrams illustrating that the appearance pattern of speckle noise is different when the frequency band is different. FIG. 10A schematically shows a state of a reflected wave (echo) in the fundamental wave band. The subjects A, B, and C in the figure are located at the same interval as the wavelength of the fundamental wave. The vertical axis (downward) of each waveform represents the distance (subject depth d) from the ultrasonic array probe 101, and the horizontal axis represents luminance. Each of the echoes A, B, and C from the subjects A, B, and C show respective waveforms when it is assumed that only the reflected wave from the subject can be received by the ultrasonic array probe 101. The synthesized echo indicates a waveform used for image generation as a received wave.

Similarly, FIG. 10B schematically shows the appearance of the second harmonic (echo).
Each of the echoes synthesized in FIGS. 10A and 10B has high luminance portions n1 and n2 stochastically generated by interference, and the high luminance component causes speckle noise as speckles (speckles). It becomes.
However, the subject depth d1 of the combined echo of the fundamental wave band in FIG. 10A and the subject depth d2 of the combined echo of the second harmonic band in FIG. Occurs. As a result, the speckle distribution pattern (or the manner in which the fluctuation component appears) in the image in which the echo is generated in the fundamental wave band is significantly different from that in the second harmonic band. Therefore, by combining the generated images, the influence of the fluctuation component can be reduced (captured as white noise).

The operation of the ultrasonic diagnostic apparatus 10 according to Embodiment 1 of the present invention configured as described above will be described.
FIG. 11 is a flowchart showing processing up to generation of a composite image in the ultrasonic diagnostic apparatus 10, and shows the procedure of the processing shown in FIGS.
In the figure, a loop 1 (steps 80 to 85) is the same number of times as the number of steer angles to be scanned, and therefore, only one process will be described. When the steer angle instructing unit 103 reads out the steer angle θi from the steer angle table T1 (FIG. 5) and instructs the transmission / reception control unit 102 to the steer angle (step 81), the transmission / reception control unit 102 controls the ultrasonic array probe 101. Then, an ultrasonic wave is transmitted at the steer angle, a reflected wave is received (step 82), and a received signal is output to the image generation unit 104. At this time, the frequency band selection unit 105 reads out and selects the frequency band ω corresponding to θi from the frequency band table T2 (FIG. 6A), and instructs the image generation unit 104 (step 83). The image generation unit 104 generates an image based on the signal of the frequency band specified by the frequency band selection unit 105 among the reception signals from the transmission / reception control unit 102 (step 84). By the processing of this loop 1, an image is generated for each steer angle specified by the steer angle instructing unit 103, based on the signal of the frequency band selected by the frequency band selecting unit 105 among the reflected signals.

As described above, according to the ultrasonic diagnostic apparatus in the present embodiment, the present invention takes into account the generation principle of speckle noise, and provides a difference in frequency band for each steer angle, for each steer angle. It is possible to increase the difference in the speckle noise appearance pattern of the generated image, reduce the effect after combination, and remove or reduce speckle noise.
Note that the steer angle instruction unit 103 may be configured to specify a frequency band different from the others for a specific steer angle among a plurality of steer angles, or may be configured to receive an operator input. . In this way, a specific steer angle can be realized by a simple configuration in which a different frequency band is designated. The same applies to the frequency band selection unit 105.

Further, the number of steer angles at the time of scanning is three, but is not limited to this. The number of steering angles at the time of scanning is determined by the number of steering angles θ1, θ2, θ3,... In the steering angle table T1. A desired number of steer angles may be written in the steer angle table T1. At the same time, the same number of frequency bands may be written in the frequency band table T2.
Further, as shown in FIG. 9, images may be generated from received waves in different frequency bands included in reflected waves at one steer angle, and the generated images may be combined.

Further, in the above-described embodiment, the configuration has been described in which the image combining unit 106 combines the images generated by the image generating unit 104 using the weighting factors α1, α2,. In 104, an image may be generated after weighting the signal of each frequency band. At this time, the weighting coefficient may be a value that is inversely proportional to the signal strength for each frequency band, similarly to α1, α2,.
(Embodiment 2)
FIG. 12 is a block diagram illustrating a configuration of an ultrasonic diagnostic apparatus according to Embodiment 2 of the present invention. This apparatus is different from the configuration shown in FIG. 4 in that a threshold setting unit 201 and a threshold mask generation unit 202 are newly added, and that an image synthesizing unit 200 is provided instead of the image synthesizing unit 106. .

  In contrast to the image combining unit 106 according to the first embodiment that combines images by weighted averaging, the present embodiment introduces a calculation method that combines pixels so as to emphasize pixel values near pixel values that are contours. In addition, an area including an outline in an image can be selected from a plurality of synthesis operation methods according to whether the operation method is selectable depending on whether the area is an area including an outline or another area. It is configured to display more clearly. In FIG. 12, components denoted by the same reference numerals as in FIG. 4 have the same functions, and therefore, different points will be mainly described.

  The image synthesizing unit 200 is configured to synthesize the image by the weighted average by the image synthesizing unit 106 according to the first embodiment. , The maximum value detection unit 203, the averaging calculation unit 204, the harmonic averaging calculation unit 205, the pixel value calculation unit 206, and the combining method selection unit 207 are selected. Prepare.

  The maximum value detection unit 203 detects a maximum value for a corresponding pixel of a plurality of pieces of image data generated by the image generation unit 104. That is, the maximum value detection unit 203 focuses on the pixel values at the same position in the plurality of image data, selects the pixel value having the largest pixel value, and sets the pixel value at the corresponding position in the composite image. . This method is effective in that, when it is clear that a signal representing a subject exists at that position, the subject can be depicted with higher luminance.

The averaging operation unit 204 calculates the averaging of the corresponding pixels of the plurality of pieces of image data generated by the image generation unit 104, and sets the obtained average as the pixel value of the pixel in the combined image. The averaging is the most commonly used technique when generating a composite image by compound scanning. The pixel value f_g (i) of the composite image obtained by the averaging is expressed by the following equation.
f_g (i) = (f_0 (i) + f_1 (i) + ... + f_ (M-1) (i)) / M
Here, the number of image data to be synthesized (the number of pixels thereof is N) is M, and the i-th (i is 1 to N) -th of the m-th (m is 0 to (M-1))-th image data Let the pixel be f_m (i) and the i-th pixel value of the composite image be f_g (i).

  In a composite image obtained by averaging, pixel values among a plurality of images are uniformly reflected on the pixel values of the composite image, and a seemingly natural composite result is obtained. Further, since an LPF (Low Pass Filter) effect is also obtained by the superposition effect, the effect of general white noise following a Gaussian distribution is reduced by performing this processing. Speckle noise can be expected to have the same effect when the appearance pattern differs between a plurality of image data, but on the other hand, blurring also occurs in the outline of the subject to be clearly drawn, so use of this method alone is not desirable. .

The harmonic average calculation unit 205 calculates a harmonic average for a corresponding pixel of the plurality of image data generated by the image generation unit 104, and sets the average as a pixel value of the pixel in the image of the synthesis result. The pixel value f_g (i) of the composite image based on the harmonic mean is expressed by the following equation.
f_g (i) = M / ((1 / f_0 (i)) + (1 / f_0 (i)) + ... + (1 / f_ (M-1) (i)))
In the calculation of the harmonic mean, the calculation cannot be performed when the pixel value is 0 or a negative value. However, when the pixel value is 0, a process of replacing the pixel value with a non-zero positive value (for example, 1) is performed. The pixel value does not generally include a negative value, but if present, performs processing such as replacing the pixel value with a non-zero positive value. According to the harmonic averaging, when there is a value (large value) that is extremely different from the value, the influence of this value on f_g (i) can be reduced.

The pixel value calculation unit 206 calculates the pixel value of the composite image for the corresponding pixel of the plurality of image data generated by the image generation unit 104 according to the following expression f_g (i), and obtains the image of the composite result. Is the pixel value of the pixel.
f_g (i) = SM / ((1 / (S-f_0 (i))) + (1 / (S-f_1 (i))) + ... + (1 / (S-f_ (N-1)) (I))))
Here, S has a value that does not exceed a value f_m (i) can take and is a reference value. That is, the pixel value close to S is reflected so as to be emphasized in f_g (i). When performing this pixel operation, f_m (i) must be a positive value other than S, and if f_m (i) is equal to S, the value is changed to a positive value close to S other than S ( (For example, S + 1).

  The synthesis method selection unit 207 determines whether any one of the four calculation units 203 to 206 is to be used for the pixels in the region including the contour in the image for each steer angle generated by the image generation unit 104. For the pixels in the other area, a different operation unit is selected and executed. Whether or not the area includes an outline depends on the mask data generated by the threshold value setting unit 201 and the threshold value mask generation unit 202. That is, the synthesizing method selecting unit 207 selects an image synthesizing method for each pixel from the four arithmetic units 201 to 204 based on the mask data generated by the threshold mask generating unit 202.

  The threshold setting unit 201 sets a threshold for each image generated by the image generation unit 104. The threshold can be set in two ways: (a) when the user sets an arbitrary value using an input device (keyboard, trackball, various switches, etc.) attached to the ultrasonic diagnostic apparatus; If the value is determined in advance by parameters such as a frequency, an amplification factor at the time of image generation, a used band, and a combining method, (c) an arbitrary region of image data to be combined is designated, and a pixel value of the region is determined. There are cases where a value such as an average value, an intermediate value, a maximum value, or a minimum value is set as a threshold.

The threshold mask generation unit 202 determines the magnitude of the pixel value of the image data according to the threshold set in the threshold setting unit 201, and creates mask data containing the result. The mask data here indicates whether or not a pixel exceeds a threshold value, and the threshold value is set as a value indicating a limit of a pixel value in an outline of the image and its vicinity.
FIG. 13 is a diagram illustrating an example of a method of generating mask data.

Here, only one threshold value Th is given, and the threshold mask generation unit 202 generates a composite image from three pieces of image data (f_1, f_2, f_3) having N pixel values. And
When a certain pixel f_1 (i) (i is 1 to N) of the image data f_1 is larger than Th, the mask data T_f1 (i) for this image data is set to 1. This means that it is a pixel in the contour and the area near the contour. If it is less than Th, T_f1 (i) = 0. This means that it is the other area. The threshold mask generation unit 202 performs this processing for all the pixel values, and also performs the same processing for the three pieces of image data, so that the respective mask data T_f1, T_f2, Generate T_f3.

Further, as shown in the lower part of the figure, the threshold mask generation unit 202 generates mask data T_g (i) to be used for synthesis from the three mask data T_f1, T_f2, and T_f3. In the three mask data shown in the figure, the mask value takes a value of 1 (white portion) or 0 (black portion). Here, the result of the logical sum of the three mask data is used as the mask data for synthesis.
This is an example, and if it is desired to obtain an image that is equal to or larger than the threshold value in all the image data, the masking data for synthesis may be obtained by logical product and the number of bits per pixel of the mask data may be increased. The mask data may have a value other than 0 and 1 to be multi-valued. In that case, three or more areas such as an outline in the image and an area in the vicinity thereof and a low-luminance area can be determined.

The operation of the ultrasonic diagnostic apparatus according to the present embodiment configured as described above will be described.
FIG. 14 is a flowchart illustrating the combining operation of the image combining unit 200. In the figure, the loop 1 (steps 101 to 107) is performed the same number of times as the number N of pixels of the image generated by the image generation unit 104, and therefore, only one of the loops will be described below.

  The combining method selecting unit 207 determines whether the mask data T_g (i) corresponding to the pixel to be combined is 1 (region 1 including the contour) or T_g (i) = 0 (other region 0). It is determined whether there is any data (step 102), and according to the determination result (step 103), the calculation for area 1 is selected (step 105) or the calculation for area 0 is selected (step 104). An arithmetic unit corresponding to the selected operation performs a pixel synthesis operation (step 106).

  As described above, the combining method selecting unit 207 selects a combining method based on the value of the threshold mask data T_g (i) generated by the threshold mask generating unit 202. In other words, when T_g (i) is composed of binary values of 0 and 1, for example, in the region of T_g (i) = 0, the average of f_1 (i), f_2 (i), f_3 (i), T_g (i) In the area where) = 1, the maximum value is detected.

  Further, when T_g (i) = 1, it means that there is a high luminance value indicating the presence of the inspection object at the location i, so that three pieces of image data f_1 (i), f_2 (i), By selecting the highest luminance of f_3 (i), it is possible to more clearly depict the subject. When T_g (i) = 0, the location i does not correspond to the contour portion in the subject, and therefore, the luminance value at that location is likely to represent noise. When the averaging process is performed on such a signal, an LPF effect is obtained, and a noise component is suppressed in the composite image.

When the threshold mask data T_g (i) has a value other than 0 or 1 (three or more values), a combining method may be selected for each value.
Here, the selection of each operation unit will be supplementarily described.
In the case of the harmonic averaging process, when there is extremely different (large) data in a series of data groups, the influence can be reduced.

As a first example, when there is a data set of {80, 90, 100, 1000}, 317.5 when the averaging unit 204 is selected, 115.6 when the harmonic averaging unit 205 is selected, and pixels When the value calculation unit 206 is used, the value is 88.0 when S = 75.
As a second example, when there is a data set {80, 90, 100, 1}, the averaging operation unit 204 is 67.8, the harmonic average operation unit 205 is 3.869, and the pixel value operation unit 206 is S = When it is 75, it becomes 88.6.

  In the data set of the above two examples, there are three values around 90, and a value of 1000 or 1 is set as a value far apart. That is, assuming that the value around 90 is a value close to the correct answer, and looking at the above calculation result, when the averaging operation unit 204 (the averaging operation unit) is used, it is affected by the value of 1000 or 1 The value deviates from around 90.

According to the harmonic average calculation unit 205, good results are obtained for the data set of the first example {80, 90, 100, 1000}, but the harmonic average is strongly affected by a value near 1. Therefore, in the data set of the second example {80, 90, 100, 1}, the values are far apart.
However, as described in the second embodiment, it is possible to set a certain threshold value and select a value of a pixel value to be processed to some extent. Therefore, avoid processing a pixel value having a value close to 1 Is possible.

  When the pixel value calculation unit 206 is used, a stable result is output for both data sets. This is because a value close to 90 of 75 was selected as the reference value S. In this formula, a process is performed in which a harmonic mean is obtained by reversing the magnitude relationship between pixel values, and the magnitude of the result is again reversed. Accordingly, the equation of the pixel value calculation unit 206 is obtained by modifying the characteristic of the harmonic mean (a value close to 1 (reference value) is emphasized) so as to emphasize a value close to the reference value S.

  The user can arbitrarily select these synthesis methods. In addition, the method of determining the reference value S when using the pixel value calculation unit 206 is arbitrary. However, if a luminance value depicting the inspection object can be obtained quantitatively, the reference value S is used as the reference value S. Thus, the influence of a value far apart can be removed, and the roughness of the screen can be suppressed.

  As described above, according to the ultrasonic diagnostic apparatus in the present embodiment, by setting the same value as the pixel value to be emphasized as the reference value S in the pixel value calculation unit 206, A portion having a pixel value near the reference value can be highlighted.

  In the present embodiment, in addition to the removal or reduction of the speckle noise in the first embodiment, the effect of clearing the image by the enhancement by the pixel value calculation unit 206 is synergistically achieved. The effect of emphasis can be obtained independently of the effect of the first embodiment. That is, if a value of a pixel value representing an outline is set as the reference value S, a region including the outline in the image of the synthesis result can be emphasized and displayed clearly. In addition, if a value of a pixel in a specific area to be emphasized other than the contour is set as the reference value S, the area can be emphasized to improve the image quality and to display finely.

  The threshold value setting unit 201 sets one or two or more values as boundaries of a plurality of regions as threshold values, and the threshold mask generation unit 202 sets mask data indicating a region including an outline or a low-luminance region. Can be determined, that is, the region can be determined. Further, by selecting one of the calculation units 203 to 206 for each region indicated by the mask data in the synthesis method selection unit 207, an optimal pixel calculation method can be determined for each region. By selecting and synthesizing by optimal pixel calculation for each area such as an area including an outline, a low-luminance area, and other areas, a clear display can be obtained. In the present embodiment, in addition to the removal or reduction of speckle noise in the first embodiment, synthesizing that a clear image is obtained by selecting a pixel operation for each region by the threshold setting unit 201 and the threshold mask generation unit 202. However, the effect of selecting the pixel operation for each region can be obtained independently of the effect of the first embodiment.

  It is desirable that the maximum value detection unit 203 performs image adjustment on the plurality of image data generated by the image generation unit 104, such as leveling the average pixel level. This is because the intensity (amplitude) of the received signal differs for each frequency band (fundamental wave, second harmonic, third harmonic) included in the reflected signal. That is, since the amplitude ratio of each of the received signals of the fundamental wave, the second harmonic, and the third harmonic is about 100: 10: 1, the image data may be leveled to absorb the amplitude ratio. For example, the maximum value detection unit 203 may perform leveling by weighting each image data so as to be inversely proportional to the intensity of the received signal, and perform maximum value detection on the leveled image data.

Such leveling of the image data may be similarly performed in the averaging operation unit 204, the harmonic averaging operation unit 205, and the pixel value operation unit 206. That is, the averaging operation unit 204 may be configured to obtain the pixel value f_g (i) of the synthesized image by averaging with weighting, similarly to the image synthesis unit 106 in the first embodiment. In that case, the averaging unit 204 may obtain the pixel value f_g (i) of the composite image. Here, the weight coefficient for the m-th (m is 0 to (M-1)) image data is αm.
f_g (i) = (α_0 · f_0 (i) + α_1 · f_1 (i) + ... + α_ (M-1) · f_ (M-1) (i)) / M
In addition, the harmonic average calculation unit 205 may be configured to obtain the pixel value f_g (i) of the composite image by the harmonic average with weighting represented by the following equation.
f_g (i) = M / ((1 / (α_0 · f_0 (i)) + 1 / (α_1 · f_1 (i)) + ... + 1 / (α_ (M-1) · f_ (M-1) (i) ))
Further, the pixel value calculation unit 206 may be configured to obtain the pixel value f_g (i) of the composite image by performing a pixel calculation with weighting represented by the following equation.
f_g (i) = SM / (p_0 (i) + p_1 (i) + ... + p_ (M-1) (i))
p_0 (i) = 1 / (S-α_0 ・ f_0 (i))
p_1 (i) = 1 / (S-α_1 ・ f_1 (i))
p_ (M-1) (i) = 1 / (S-α_ (M-1) ・ f_ (M-1) (i))
When the image generation unit 104 generates image data by weighting using the weighting coefficient α, the weighting in the image synthesis unit 200 is unnecessary.

(Embodiment 3)
In this embodiment, a method of storing an RF (Radio Frequency) signal as a basis for generating image data and generating a higher-quality composite image in an offline state will be described.

  FIG. 15 shows a schematic configuration of an ultrasonic diagnostic apparatus according to Embodiment 3 as an example to which the present invention is applied. The apparatus differs from the configuration in FIG. 12 in that a new RF signal storage unit 301 is added, and that the image generation unit 104 sets a plurality of frequency bands (for example, the aforementioned ω1 , Ω2, ω3). Hereinafter, the different points will be mainly described.

The RF signal storage unit 301 stores an RF signal received by the transmission / reception control unit 102, and has a capacity sufficient to store at least an amount of RF signal necessary for synthesizing one composite image when offline. Have.
FIG. 16 is a diagram showing an example of the flow of processing at the time of offline. At the time of off-line, the RF signal from the ultrasonic array probe 101 (that is, the received reflected wave) is stored in the RF signal storage unit 301. At the time of off-line, all information necessary for image synthesis is stored in the RF signal storage unit 301.

The image generation unit 104 reads the RF signal from the RF signal storage unit 301, generates image data by changing various parameters (such as a frequency band), and stores the image data in an image data memory (not shown). When generating the image data, the image generating unit 104 performs weighting according to the frequency band in order to level the average pixel level of the image data.
FIG. 16 shows a state in which image data is generated by the fundamental wave ω1, the second harmonic ω2, and the third harmonic ω3 for each of the RF signals having the steer angles in three directions (θ1, θ2, θ3). A total of nine pieces of image data are stored.

The nine pieces of image data thus generated are combined by the image combining unit 106. FIG. 16 shows two combining methods.
In the synthesis example 1, first, a frequency compound (synthesis of images in different frequency bands at the same steering angle) is performed, and then the synthesis results are further synthesized to obtain a final synthesized image.

In synthesis example 2, the order is changed, images of the same frequency band having different steer angles are first synthesized, and the synthesis results are further synthesized (frequency compound) to obtain a final synthesized image. .
Both synthesis examples are mathematically equivalent processing, and the final result should not depend on the processing order.However, in the actual processing, variable components such as digit loss and rounding error are mixed. Since the transmission to the subsequent stage is different, the effect on the final image appears. The order in which the synthesis processing is performed depends on which method is selected as the synthesis method. Therefore, a memory or the like in which the relationship between the processing method and the processing order is registered in advance is provided. A configuration in which the user specifies the combination method and order each time may be adopted.

In the case of the image synthesis at the time of off-line, the processing amount increases as described above, but since the number of data used for generating the composite image increases, the effect of removing noise can be further increased.
However, if the number of image data used for the synthesis increases, the synthesized image may be blurred accordingly. In this case, an HPF (High Pass Filter), an edge enhancement filter, or an equivalent By using a synthesizing method having an effect, the blur of the synthesized image may be restored.

  It is needless to say that the functional block diagrams shown in FIGS. 4, 9 and 13 of each embodiment and the flowchart in FIG. 11 can be realized as a program in the DSP or CPU in the main unit 12 shown in FIG. This program can be distributed through a recording medium such as a CD or a telecommunication line, and is independently traded as package software or download software.

  The present invention is suitable for an image processing apparatus that transmits an ultrasonic wave from an ultrasonic probe and generates an image based on a reflected signal generated from a reflected wave from an object, for example, an ultrasonic diagnostic apparatus for medical use and the like. ing.

FIG. 4A is a diagram illustrating a state in which an ultrasonic beam is irradiated from an ultrasonic array probe in a plurality of directions. (B) It is a figure showing an inspection object. (C) is a diagram showing image data acquired from three directions. (A) is a figure which shows a mode that a test object with a shallow depth is scanned. (B) It is a figure showing signs that an inspection object with a deep depth is scanned. FIG. 1 is an external view of an ultrasonic diagnostic apparatus according to Embodiment 1 of the present invention. FIG. 2 is a block diagram illustrating a main functional configuration of the ultrasonic diagnostic apparatus. FIG. 4 is a diagram showing a steering angle table and a pointer P. (A) is a diagram showing a frequency band table and a pointer Q. (B) is a diagram showing a pointer correspondence table. FIG. 3 is a diagram illustrating a relationship between a steering angle, a frequency band, and a generated image. FIG. 11 is another example of a diagram for explaining a relationship among a steering angle, a frequency band, and a generated image. FIG. 11 is still another example of a diagram for explaining a relationship between one steer angle, a frequency band, and a generated image. (A) A state of a reflected wave (echo) in a fundamental wave band is schematically shown. (B) The appearance of the reflected wave (echo) in the second harmonic band is schematically shown. It is a flowchart figure which shows a process until a generation of a synthetic image in an ultrasonic diagnostic apparatus. FIG. 9 is a block diagram illustrating a configuration of an ultrasonic diagnostic apparatus according to Embodiment 2 of the present invention. FIG. 4 is a diagram illustrating an example of a method for generating mask data. 6 is a flowchart illustrating a combining operation of an image combining unit. FIG. 13 is a diagram illustrating a schematic configuration of an ultrasonic diagnostic apparatus according to Embodiment 3. FIG. 7 is a diagram illustrating an example of a flow of processing at an offline time.

Explanation of reference numerals

T1 Steer angle table T2 Frequency band table T3 Pointer correspondence table 10 Ultrasonic diagnostic device 11 Display device 12 Main unit 13 Ultrasonic array probe 101 Ultrasonic array probe 102 Transmission / reception control unit 103 Steer angle instruction unit 104 Image generation unit 105 Frequency band selection Unit 106 Image synthesis unit 107 Image storage unit 108 Frame memory 109 Image display unit

Claims (20)

An image processing apparatus that transmits an ultrasonic wave from an ultrasonic probe and generates an image based on a reflected signal generated from a reflected wave from an object,
Designation means for designating at least two frequency bands in the reflected signal;
Generating means for generating an image for each frequency band based on the signal of the designated frequency band among the reflected signals,
And a synthesizing unit for synthesizing the generated image for each frequency band. The combining means is M for the number of images generated for each frequency band by the generating means, N for the number of pixels of each image, and f_m (i) for the i-th pixel value of the m-th image among the generated images. , F_m (i) does not exceed a possible value, and S is used to synthesize the i-th pixel value f_g (i) of the synthesized image according to the pixel value calculation shown in (Equation 1). The image processing apparatus according to claim 1.
(Equation 1) f_g (i) = S−M / ((1 / (S−f — 0 (i))) + (1 / (S−f_1 (i))) +... + (1 / (S−f_ ( M-1) (i)))) The image processing apparatus further includes an area determination unit configured to determine a first area including an outline and another area based on the plurality of images generated by the generation unit,
The synthesizing unit performs a first synthesizing operation on a plurality of images generated by the generating unit with respect to pixels in a first area, and performs a first synthesizing operation on pixels in an area other than the first area. 2. The image processing apparatus according to claim 1, wherein a different operation is performed. The area determination means includes:
Threshold determination means for performing threshold determination on a pixel-by-pixel basis for each image generated by the generation means,
4. The image processing apparatus according to claim 3, further comprising: a region data generating unit configured to generate a region data indicating a distribution of the first region and the other region by taking a logical sum of a result of the threshold determination for each image. Image processing device.   The image processing apparatus according to claim 4, wherein the threshold value determination unit performs the threshold value determination using a limit of a value that can be taken by a pixel near the contour as a threshold value. The combining means includes:
At least two of the first to fourth computing units;
For a pixel in the first area, one operation unit is used as the first synthesis operation, and for pixels in an area other than the first area, another operation unit is used as an operation different from the first synthesis operation. A selecting unit for selecting and calculating for each pixel according to the area data,
The first arithmetic unit combines pixels with a maximum value for a plurality of images generated by the generation unit,
The second arithmetic unit synthesizes pixels by averaging,
The third arithmetic unit synthesizes pixels by harmonic averaging,
The fourth arithmetic unit calculates the number of the plurality of images as M, the number of pixels of each image as N, and the i-th pixel value of the m-th image among the generated images as f_m (i), f_m. The value of the i-th pixel value f_g (i) of the synthesized image is synthesized according to the pixel value calculation shown in (Equation 2), where S is a value not exceeding a value that (i) can take. Image processing device.
(Equation 2) f_g (i) = S−M / ((1 / (S−f — 0 (i))) + (1 / (S−f_1 (i))) +... + (1 / (S−f_ ( M-1) (i)))) The ultrasonic probe transmits ultrasonic waves at a plurality of steering angles,
The specifying means specifies different frequency bands for at least two steer angles of the plurality of steer angles,
The generating means, for each of the steering angles, generates an image for each frequency band based on the signal of the specified frequency band among the reflected signals,
The image processing apparatus according to claim 1, wherein the combining unit combines the generated images for each of the steering angles. The image processing apparatus according to claim 7, wherein the specifying unit specifies a frequency band different from others for a specific steer angle among the plurality of steer angles. The designation means is a set consisting of an angle group indicating a plurality of steer angles and a band group indicating a frequency band corresponding to the steer angle, and stores a plurality of sets including at least the first and second sets. Have a table,
The image processing apparatus controls the steer angle of the ultrasonic probe based on the angle group in one set stored in the correspondence table,
The specifying means specifies a frequency band for each steer angle based on a band group corresponding to the angle group in the set,
The correspondence table indicates that, when the steering angle difference of the angle group in the first group is smaller than the steering angle difference of the angle group in the second group, the band difference of the first band group is equal to the band of the second band group. The image processing apparatus according to claim 7, wherein the setting is set to be larger than the difference. The combining means is M for the number of images generated for each steering angle by the generating means, N for the number of pixels of each image, and f_m (i) for the i-th pixel value of the m-th image among the generated images. , F_m (i) as a value that does not exceed a possible value, as S, synthesize the i-th pixel value f_g (i) of the synthesized image according to the pixel value calculation shown in (Equation 3). The image processing apparatus according to claim 1.
(Equation 3) f_g (i) = S−M / ((1 / (S−f — 0 (i))) + (1 / (S−f_1 (i))) +... + (1 / (S−f_ ( M-1) (i)))) The image processing apparatus further includes an area determination unit configured to determine a first area including a contour and an area other than the first area based on the image for each steering angle generated by the generation unit,
The synthesizing unit performs a first synthesizing operation on a pixel in a first area with respect to an image for each steering angle generated by the generating unit, and performs a first synthesizing operation on a pixel in an area other than the first area. The image processing apparatus according to claim 7, wherein an operation different from the synthesis operation is performed. The area determination means includes:
Threshold determination means for performing threshold determination on a pixel-by-pixel basis for each image generated by the generation means,
12. An area data generating means for generating area data indicating distribution of a first area and other areas by taking a logical sum of a result of the threshold determination for each image. Image processing device.   13. The image processing apparatus according to claim 12, wherein the threshold value determination unit performs the threshold value determination using a limit of a value that can be taken by a pixel near the contour as a threshold value. The combining means includes:
At least two of the first to fourth computing units;
For a pixel in the first area, one operation unit is used as the first synthesis operation, and for pixels in an area other than the first area, another operation unit is used as an operation different from the first synthesis operation. A selecting unit for selecting and calculating for each pixel according to the area data,
The first arithmetic unit combines pixels with a maximum value for a plurality of images generated by the generation unit,
The second arithmetic unit synthesizes pixels by averaging,
The third arithmetic unit synthesizes pixels by harmonic averaging,
The fourth calculation unit is configured to determine the number of images generated for each steering angle by the generation unit as M, the number of pixels of each image as N, and the i-th pixel value of the m-th image among the generated images as f_m. (I) A value not exceeding a possible value of f_m (i) is set as S, and the i-th pixel value f_g (i) of the composite image is synthesized according to the pixel value calculation shown in (Equation 4). The image processing device according to claim 11.
(Equation 4) f_g (i) = S−M / ((1 / (S−f — 0 (i))) + (1 / (S−f_1 (i))) +... + (1 / (S−f_ ( M-1) (i)))) The image processing apparatus further includes a storage unit that stores the reflection signal received by the ultrasonic probe,
The generating unit and the synthesizing unit generate and synthesize an image in real time based on the reflection signal received by the ultrasonic probe, or generate and synthesize an image off-line based on the reflection signal stored in the storage unit. The image processing apparatus according to claim 14, wherein the image processing apparatus is configured to be able to select one of the following. The specifying means specifies a plurality of frequency bands for each steering angle in the case of the off-line,
The generating means, in the case of the off-line, generates an image for each of a plurality of frequency bands for each steer angle,
The combining means generates a first combined image and a second combined image,
The first synthesized image is obtained by synthesizing images having different frequency bands of the same steering angle, and further synthesizing the synthesis result for each steering angle,
The image processing apparatus according to claim 15, wherein the second synthesized image is obtained by synthesizing images of the same frequency band having different steer angles, and further synthesizing the synthesis result for each frequency band. An image processing method for transmitting an ultrasonic wave from an ultrasonic probe and generating an image based on a reflection signal from an object,
A designating step for designating at least two frequency bands in the reflected signal;
A generation step of generating an image for each frequency band based on a signal of a specified frequency band among the reflected signals,
Combining a generated image for each frequency band. The ultrasonic probe transmits an ultrasonic wave at a plurality of steer angles, In the specifying step, at least two steer angles of the plurality of steer angles, specify different frequency bands,
In the generating step, for each of the steering angles, generate an image for each frequency band based on the signal of the specified frequency band of the reflected signal,
The image processing method according to claim 17, further comprising: combining the generated images for each of the steering angles in the combining step. Transmitting ultrasonic waves from an ultrasonic probe, a program executable by a computer in an image processing apparatus that generates an image based on a reflected signal from an object,
A designation step for designating at least two frequency bands in the reflected signal;
A generation step of generating an image for each frequency band based on a signal of a specified frequency band among the reflected signals,
And a synthesizing step of synthesizing the generated image for each frequency band. The ultrasonic probe transmits an ultrasonic wave at a plurality of steer angles, In the specifying step, at least two steer angles of the plurality of steer angles, specify different frequency bands,
In the generating step, for each of the steering angles, generate an image for each frequency band based on the signal of the specified frequency band of the reflected signal,
20. The non-transitory computer-readable storage medium according to claim 19, further comprising: synthesizing the generated image for each steering angle in the synthesizing step.

Images