JP7290968B2 - Image processing device and program - Google Patents

Description

The present invention relates to an image processing apparatus and program for inputting a moving image for each frame image and performing image processing using wavelet degeneration based on wavelet packet decomposition.

Conventionally, as a technique for removing noise by inputting a moving image frame by frame and using wavelet degeneration based on wavelet packet decomposition, among the frequency decomposition coefficients obtained by wavelet transforming the input image, the components below a predetermined noise level are cored. It is known to remove noise by

Also, as an image processing device using wavelet degeneration, a high-frequency noise image and a low-frequency noise image are extracted, and these images are divided and integrated according to whether they are images of luminance components or images of color difference components. A technique for reducing the noise level is known (see Patent Document 1, for example).

Patent No. 5352942 specification

David L. Donoho, Iain M. Johnstone, Gerard Kerkyacharian and Dominique Picard: “Wavelet Shrinkage: Asymptopia?”, Journal of the Royal Statistical Society. Series B (Methodological), Vol. 57, No. 2, pp. 301-369 , 1995.

As described above, techniques for removing noise from an image using wavelet degeneration are conventionally known, such as in Non-Patent Document 1 and Patent Document 1.

On the other hand, in recent years, for example, transmission of large-capacity content services such as 4K and 8K SHV (Super Hi-Vision) as ultra-high-definition moving images exceeding high-definition has been considered as next-generation terrestrial broadcasting. When performing high-compression encoding for the transmission of such ultra-high-definition video, it is assumed that video images that are difficult to high-compression-encode due to reasons such as the power of spatial high-frequency band components being large may be input. be. In this case, since the encoding bit rate may not be sufficient for the enormous amount of information that the input moving image has as transmission capacity, image corruption may occur due to encoding related to compression or transmission.

Therefore, it is possible to reduce the amount of information contained in the input video by performing conventional noise removal processing such as Non-Patent Document 1 and Patent Document 1 before high-compression encoding processing for ultra-high-definition video. .

However, if the conventional noise removal processing is used to remove noise from an ultra-high-definition moving image, image corruption is suppressed, but there is a problem that edge components and the like tend to blur.

For this reason, when reducing the amount of information by using wavelet degeneracy in the preceding stage of compression encoding processing for moving images such as ultra-high-definition moving images, it is difficult to compress and encode moving images that are difficult to compress and encode while suppressing blurring of edge components. A technique is desired that suppresses image corruption that occurs in video and improves the coding quality of moving images.

Therefore, an object of the present invention is to perform compression encoding while suppressing blurring of edge components by inputting a moving image frame by frame before compression encoding processing and performing image processing using wavelet degeneration based on wavelet packet decomposition. An object of the present invention is to provide an image processing device and a program for improving image quality related to processing.

An image processing apparatus according to one aspect of the present invention performs image processing using wavelet degeneration based on wavelet packet decomposition by inputting a moving image for each frame image before compression encoding processing. Based on the continuous signal power of the band group, it is configured to adaptively set a degeneracy function that performs at least band limitation while suppressing blurring of edge components and the like. Furthermore, the image processing apparatus of another aspect according to the present invention performs wavelet packet decomposition for each frame image of the moving image to be processed in order to enable both band limitation and noise removal while suppressing blurring of edge components. Efficiently sets the degeneracy function based on both the continuous signal power and the noise power of each frequency band group based on .

That is, the image processing apparatus of the present invention is an image processing apparatus that inputs a moving image for each frame image and performs image processing using wavelet degeneration based on wavelet packet decomposition. ) first-order wavelet packet decomposition processing and extracting frequency decomposition coefficients of each frequency band; The frequency resolution coefficients of the horizontal high-frequency band, the vertical high-frequency band, and the diagonal high-frequency band, which are equal to or greater than a predetermined threshold, are regarded as continuous signal components, and the average value, standard deviation, or median of the frequency resolution coefficients equal to or greater than the predetermined threshold is calculated. A continuous signal power detector for detecting a value as a continuous signal power of each frequency band group in the restricted band; A restricted band degeneration function setting unit that sets a degeneration function for a band group, and performs wavelet degeneration processing by applying the degeneration function to frequency resolution coefficients of each frequency band obtained by performing the n-th order wavelet packet decomposition processing. a wavelet degeneracy unit that band-limits the frequency resolution coefficients within the restricted band by , Among the frequency resolution coefficients of each frequency band obtained by performing the n-th order wavelet packet decomposition processing, the power that is equal to or greater than a predetermined ratio in the power of all the frequency resolution coefficients in the spatial diagonal highest frequency band is defined as a noise component. and a noise power detection unit for detecting noise power, wherein the degeneracy function setting unit for limited band, based on the continuous signal power, the noise power, and the limit amount parameter, in the limited band It is characterized by setting a degeneracy function for each frequency band group .

Further, in the image processing apparatus of the present invention, a passband degeneration function setting unit for setting a degeneration function for a frequency band group in the passband based on the noise power, with a frequency band group other than the restricted band as the passband. wherein the wavelet degeneration unit applies the degeneration functions set by the restricted band degeneration function setting unit and the passband degeneration function setting unit to perform wavelet degeneration processing, thereby reducing the restricted band band limitation and noise elimination of the frequency resolution coefficients within the passband, and noise elimination of the frequency resolution coefficients within the passband.

Further, in the image processing apparatus of the present invention, the limiting amount parameter includes a first band limiting amount for band limiting the entire band within the limited band, and a second band limiting amount based on the continuous signal power. and an amount.

Furthermore, the program of the present invention constitutes a program for causing a computer to function as the image processing apparatus of the present invention.

According to the present invention, it is possible to suppress image corruption that occurs in moving images that are difficult to encode while suppressing blurring of edge components and the like, and it is possible to improve the quality of moving image encoding.

1 is a block diagram showing a schematic configuration of an image processing apparatus according to one embodiment of the present invention; FIG. FIG. 4 is a diagram showing frequency bands obtained by second-order wavelet packet decomposition related to image processing of one example in the image processing apparatus of one embodiment according to the present invention; 4 is a flow chart showing image processing of an example in the image processing apparatus of one embodiment according to the present invention; 3(a) and 3(b) are explanatory diagrams of degeneration functions for a passband and each restricted band, respectively, according to an example of image processing in an image processing apparatus according to an embodiment of the present invention; FIG. 8A and 8B are explanatory diagrams of degeneration functions for a passband and each restricted band, respectively, related to image processing of a modified example in the image processing apparatus according to one embodiment of the present invention; FIG. FIG. 4 is a diagram showing frequency bands obtained by third-order wavelet packet decomposition related to image processing of one example in the image processing apparatus of one embodiment according to the present invention;

An image processing apparatus 1 according to an embodiment of the present invention will be described in detail below with reference to the drawings.

〔Device configuration〕
FIG. 1 is a block diagram showing a schematic configuration of an image processing apparatus 1 of one embodiment according to the present invention. The image processing apparatus 1 of the present embodiment is configured as an apparatus for inputting a moving image frame by frame before compression encoding processing and performing image processing using wavelet degeneration based on wavelet packet decomposition. A continuous signal power detection unit 12 , a noise power detection unit 13 , a restriction band degeneration function setting unit 14 , a passband degeneration function setting unit 15 , a wavelet degeneration unit 16 and a reconstruction unit 17 are provided. In the example shown in FIG. 1, the wavelet packet decomposition order number (nth order) can be externally specified for the image processing apparatus 1, and each necessary component in the image processing apparatus 1 can be specified externally. , it is also possible to have a configuration in which processing is performed while being transmitted as parameters.

The wavelet packet decomposition unit 11 receives a frame image of a moving image as an input image, and performs nth-order wavelet packet decomposition processing on the input image in accordance with the set wavelet packet decomposition order (n≧2). The frequency-resolved coefficients of each frequency band obtained by are output to the wavelet degeneracy unit 16 . Of the frequency resolution coefficients of each frequency band obtained by the n-th order wavelet packet decomposition, the frequency resolution coefficients of a predetermined spatial high frequency band group are extracted as a restricted band by the continuous signal power detection unit 12, and the n-th order wavelet Among the frequency resolution coefficients of each frequency band obtained by packet decomposition, the frequency resolution coefficients in the spatial diagonal high frequency band are extracted by the noise power detector 13 .

The spatial high frequency band group and the spatial low frequency band group will be described more specifically with reference to FIG. FIG. 2 is a diagram showing frequency bands obtained by second-order wavelet packet decomposition related to image processing of one example in the image processing apparatus 1 of one embodiment according to the present invention. FIG. 2 shows an example in which a frame image of a so-called 8K video captured by a broadcast camera and having an image size (horizontal resolution 8K (7680 pixels)×vertical resolution 4K (4320 pixels)) is used as an input image. By performing second-order wavelet packet decomposition in the horizontal and vertical directions, respectively, the second-order spatial low-frequency band LL ^m , horizontal high-frequency band LH ^m , and vertical high-frequency band represented by horizontal frequency × vertical frequency are shown in the figure. It is decomposed into a band HL ^m and a diagonal (horizontal/vertical) high frequency band HH ^m (m is the group number of each frequency band group C ^m ). The frequency band group ^C1 belonging to the spatial low frequency band group is a passband for removing only noise components, and the frequency band groups ^C2 , ^C3 , and ^C4 belonging to the spatial high frequency band group are edge components. It is determined in advance as a restricted band for which band restriction is performed while suppressing blurring such as .

That is, in the example shown in FIG. 2, for example, assuming that an 8K frame image is degenerated into a 4K frame image, the frequency band group ^C1 is the passband, and the frequency band groups ^C2 , ^C3 , and ^C4 are the restriction bands. It is defined as Regarding the frequency band group C ^m obtained by n-th order wavelet packet decomposition, which frequency band group is to be the pass band or the restriction band is predetermined so that it is uniquely determined by the value of n in the n-th order wavelet packet decomposition. shall be Note that Parseval's law holds between each level in the n-th order wavelet Hackett decomposition.

As shown in FIG. 1, the continuous signal power detector 12 detects each spatial high frequency band group C ^m (= C ² , C ³ , C ⁴ ), the frequency resolution coefficients of the horizontal high-frequency band LH ^m , vertical high-frequency band HL ^m , and diagonal high-frequency band HH ^m are extracted, and frequency resolution coefficients equal to _or greater than a predetermined threshold Th are extracted. is regarded as a continuous signal component. Then, the continuous signal power detection unit 12 detects the average value, standard deviation, or median value of the frequency resolution coefficients equal to or greater than the predetermined threshold ^value T _h as the continuous signal power P _cm of each frequency band group in the restricted band, and limits it. It is output to the band degeneration function setting unit 14 .

The noise power detection unit 13 determines, in accordance with the wavelet packet decomposition order (nth order), of each frequency band group C ^m (=C ¹ , C ² , C ³ , C ⁴ ), which is the output value from the wavelet packet decomposition unit 11. A spatial diagonal high frequency band (HH ⁴ where n=2 in this example) is extracted. Then, the noise power detection unit 13 detects the power of all components (all frequency resolution coefficients) in the spatial diagonal highest frequency band (in this example, n=2, HH ⁴ ) obtained by n-th order wavelet packet decomposition. A power exceeding a predetermined ratio (upper α% power) is regarded as a noise component, detected as a noise power _PN , and output to the degeneration function setting unit 14 for the limit band and the degeneration function setting unit 15 for the pass band. .

The degeneracy function setting unit for limited band 14 sets the continuous signal power P ^cm _of each frequency band group C ^m (=C ² , C ³ , C ⁴ ) in the limited band obtained from the continuous signal power detection unit 12 and the noise power Based on the noise power P _N obtained from the detection unit 13 and the limiting amount parameter specifying the band limiting amount, the degeneration function S ^m (=S ² , S ³ , ^S4 ). The limiting amount parameter includes a first band limiting amount (for example, 6 [dB]) for band limiting overall within the limited band, and and a second band-limiting amount (for example, 20 [dB]) based on the continuous signal power P _c ^m .

On the other hand, based on the noise power P _N obtained from the noise power detection unit 13, the passband degeneration function setting unit 15 degenerates the frequency band group C ^m (=C ¹ ) in the passband used in the wavelet degeneration unit 16. Set S ^m (=S ¹ ).

The wavelet degeneracy unit 16 applies degeneracy functions set by the restriction band degeneracy function setting unit 14 and the passband degeneracy function setting unit 15 to the frequency resolution coefficients of each frequency band input from the wavelet packet decomposing unit 11. Apply to perform the wavelet shrinkage process. As a result, the wavelet degeneracy unit 16 performs band limitation and noise removal on the frequency resolution coefficients within the restricted band, noise removal on the frequency resolution coefficients within the passband, and the frequencies in each frequency band after band limitation and noise removal The decomposition coefficients are output to the reconstruction unit 17 .

The reconstruction unit 17 generates an output image by performing wavelet reconstruction processing using the frequency resolution coefficients of each frequency band band-limited and noise-removed by the wavelet degeneration processing, and outputs it to the outside.

[Device operation]
Hereinafter, image processing of an example will be described more specifically as an operation of the image processing apparatus 1 of the present embodiment with reference to FIGS. 1 to 3. FIG.

FIG. 3 is a flow chart showing an example of image processing in the image processing apparatus 1 of one embodiment according to the present invention.

First, the image processing apparatus 1 inputs a frame image of a moving image as an input image through the wavelet packet decomposition unit 11, and according to the set wavelet packet decomposition rank (nth order) (n≧2), the input image is subjected to nth order decomposition. A wavelet packet decomposition process is performed to obtain frequency decomposition coefficients for each frequency band (step S1).

Subsequently, the image processing apparatus 1 uses the continuous signal power detection unit 12 to determine each spatial high-frequency band group C ^m that is a restricted band among the output values from the wavelet packet decomposition unit 11 according to the wavelet packet decomposition order (nth order). The frequency resolution coefficients of the horizontal high-frequency band LH ^m , vertical high-frequency band HL ^m and diagonal high-frequency band HH ^m in (=C ² , C ³ , C ⁴ ) are extracted and the frequencies equal to or higher than a predetermined threshold _Th are extracted. Consider the decomposition coefficients as continuous signal components. Then, the continuous signal power detection unit 12 detects the average value ^{, standard deviation, or median value of the frequency resolution coefficients equal to or greater than the predetermined threshold Th as the continuous signal power P cm} _{of each} frequency band group in the restricted band ( step S2).

Here, the continuous signal power P _cm of each frequency band group in the restricted band is obtained by each frequency band LH ^m , HL ^m , HH in each spatial high frequency band group C ^m (=C ² , C ³ , C ⁴ ) ^. The frequency resolution coefficients (signal components) of ^m are determined by utilizing the fact that the components above the threshold _Th are highly likely to be edge components that are continuous in the horizontal, vertical and diagonal directions. For example, to describe with reference to ^the example shown in FIG ^. 2 ^, the continuous signal power detection unit ¹² detects frequency resolution coefficients (signal components ), the average value (or standard deviation or median value) of the power of the components above the threshold T _h is obtained as the continuous signal power P _c ⁴ , and each continuous signal in the second-order frequency band groups C ² and C ³ The powers P _c ² and P _c ³ are similarly determined based on threshold values T _h (however, different threshold values may be used for the spatial high-frequency band groups C ² , C ³ and C ⁴ ).

Subsequently, in the image processing apparatus 1, the noise power detection unit 13 detects the spatial diagonal maximum frequency band (n=2 in this example), which is the output value from the wavelet packet decomposition unit 11, according to the wavelet packet decomposition order (nth order). HH ⁴ ) is extracted, the power of the predetermined upper α % of the power of all components (all frequency resolution coefficients) in the spatial diagonal highest frequency band HH ⁴ is regarded as a noise component, and the noise power P _N is detected (step S3).

In general, the power of signal components is unevenly distributed in the low frequency band. On the other hand, the power of noise components such as thermal noise is almost constant over the entire frequency band group. For this reason, noise components are dominant in the components within the spatial diagonal maximum frequency band resulting from the n-th order wavelet packet decomposition. Therefore, in this example, the noise power detection unit 13 detects the power of the upper α% of the powers of all components in the spatial diagonal maximum frequency band (in this example, n=2, HH ⁴ ) by second-order wavelet packet decomposition. is detected as the noise power _PN . For example, if α=95, the powers of all the components in the spatial diagonal highest frequency band ^HH4 are sorted, and the top 95% value is taken as the noise power _PN .

Subsequently, the image processing apparatus 1 _uses the degeneration function setting unit 14 for the limit band ^and the degeneration function setting unit ₁₅ for the pass band to set A degenerate function is set (step S4).

More ^{specifically , the} restricted band degeneration function setting unit 14 calculates ^the continuous signal power P _c ^m , the noise power P _N obtained from the noise power detection unit 13, and a limiting amount parameter specifying a band limiting amount determined in consideration of the continuous signal power P _c ^m . A degeneracy function S ^m (=S ² , S ³ , S ⁴ ) for each frequency band group in the restricted band is set.

For example, as shown in FIG. 4(b), the limited band degeneration function setting unit 14 sets the first band limiting amount for band limiting as a whole within the limited band as a limiting amount parameter to 6 [dB]. In order to limit the band while suppressing blurring of continuous signals (edges, etc.), the second band limiting amount based on the continuous signal power P _c ^m is set to 20 [dB]. In this case, the band-limited degeneracy function ^setting unit ¹⁴ sets y= ⁰ . 5x, y=0.1x corresponding to the second band limiting amount 20 [dB], and continuous signal power _P ^cm of each frequency band group C ^m (=C ² , C ³ , C ⁴ ) in the limiting band , and the noise power P _N , and the degeneration functions S ² , S ³ and S ⁴ are set for the wavelet degeneration unit 16 . Note that the slope coefficient _m0 shown in FIG. 4B is naturally determined by the second band limit amount of 20 dB. Since _m0 is not set to zero, blurring can be suppressed.

In addition, the passband degeneration function setting unit 15 uses the noise power P _N obtained from the noise power detection unit 13 as a degeneration function for the frequency band group C ^m (=C ¹ ) in the passband used in the wavelet degeneration unit 16. Set S ^m (=S ¹ ).

For example, as shown in FIG. 4(a), the degenerate function ^S1 indicated by the output y for the input x is determined based on the noise power _PN that suppresses the noise component on y=x without band limitation. The degeneration function S ¹ is set for the wavelet degeneration unit 16 .

Subsequently, in the image processing apparatus 1, the wavelet degeneracy unit 16 performs the restriction band degeneracy function setting unit 14 and the passband degeneracy function setting for the frequency decomposition coefficients of each frequency band input from the wavelet packet decomposing unit 11. By applying the degeneration function set by the unit 15 and executing wavelet degeneration processing, band restriction and noise removal of the frequency resolution coefficients within the restricted band and noise removal of the frequency resolution coefficients within the passband are performed (step S5 ).

Finally, the image processing device 1 generates an output image by performing wavelet reconstruction processing using the frequency resolution coefficients of each frequency band band-limited and noise-removed by the wavelet degeneration processing by the reconstruction unit 17. Output to the outside (step S6).

As described above, the image processing apparatus 1 of the present embodiment performs the above processing for each frame image of a moving image input before compression encoding processing, thereby adaptively and efficiently removing edge components and the like. Both bandlimiting and denoising while suppressing blur are achieved.

(Modification)
In the above-described embodiment, the image processing apparatus 1 is provided with the noise power detection unit 13 and the passband degeneration function setting unit 15, and the degeneration function illustrated in FIG. 4 is set in the wavelet degeneration unit 16. However, the noise power detection unit 13 and the passband degeneration function setting unit 15 may be omitted. The degeneracy function of this variant is illustrated in FIG.

As shown in FIG. 5(a), for the degeneration function ^S1 indicated by the output y for the input x, since y=x without band limitation, installation of the degeneration function setting unit 15 for the passband can be omitted. , the processing load can be further reduced.

Also, as shown in FIG. 5B, when the noise power detection unit 13 is omitted, the degeneration function setting unit 14 for the limited band sets the degeneration functions S ² , S ³ , For ^S4 , y = 0.5x corresponding to the first band limiting amount 6 [dB], y = 0.1x corresponding to the second band limiting amount 20 [dB], and each frequency band in the limited band can be determined based on the continuous signal power _P ^cm of the group C ^m (=C ² , C ³ , C ⁴ ), and the degeneration functions S ² , S ³ , S ⁴ are set for the wavelet degeneration unit 16 do. In the example shown in FIG. 5B as well, the slope coefficient _m0 is naturally determined by the second band-limiting amount of 20 dB, and since _m0 is not set to zero, blurring can be suppressed.

(Application example)
In the above-described embodiment and modified example, an example of operating with second-order wavelet packet decomposition with n=2 has been mainly described, but n≧3 can also be used. For example, as shown in FIG. 6, it is also possible to operate with third-order wavelet packet decomposition with n=3. FIG. 6 shows an example in which a frame image of a so-called 8K video captured by a broadcast camera and having an image size (horizontal resolution of 8K (7680 pixels)×vertical resolution of 4K (4320 pixels)) is used as an input image. By performing third-order wavelet packet decomposition in the horizontal and vertical directions, respectively, the spatial low-frequency band LL ^m , horizontal high-frequency band LH ^m , and vertical high-frequency band represented by horizontal frequency × vertical frequency are shown in the figure. It is decomposed into a band HL ^m and a diagonal (horizontal/vertical) high frequency band HH ^m (m is the group number of each frequency band group C ^m ). The frequency band groups C ¹ to C ⁹ belonging to the spatial low frequency band group ^shown in FIG ^. may be determined in advance as a restricted band targeted for noise removal while suppressing blurring of edge components and the like.

As shown in FIG. 6, when performing third-order wavelet packet decomposition, the continuous signal power detector 12 detects each frequency band group C ^m (=C ¹⁰ to C ¹⁶ ) in the restricted band as in the example shown in FIG. , and the noise _power detection unit 13 uses the spatial diagonal highest frequency band (n = 3 in this example, HH ¹⁶ ) by the third-order wavelet packet decomposition to ^obtain the noise power P _N can be obtained. For this reason, the limit band degeneration function setting unit 14 and the pass band degeneration function setting unit 15 set the degeneration functions in the limit band and the pass band to the wavelet degeneration unit 16 in the same manner as in the example shown in FIG. can be set.

The image processing apparatus 1 in the above embodiment can be configured by a computer, and a program for functioning each processing unit of the image processing apparatus 1 can be preferably used. Specifically, a control unit for controlling each processing unit of the image processing apparatus 1 can be configured by a central processing unit (CPU) in a computer, and a program necessary for operating each processing unit can be At least one memory can be used as a storage unit that stores data as appropriate. That is, the functions of each processing unit of the image processing apparatus 1 can be realized by causing the CPU of such a computer to execute the program. Furthermore, a program for realizing the function of each processing unit of the image processing apparatus 1 can be stored in a predetermined area of the storage unit (memory) described above. Such a storage unit can be configured with a RAM or ROM inside the device, or can be configured with an external storage device (eg, hard disk). Also, such a program can be made up of a part of software (stored in a ROM or an external storage device) on an OS used in a computer. Furthermore, a program for causing such a computer to function as each processing unit of the image processing apparatus 1 can be recorded on a computer-readable recording medium. Also, each processing unit of the image processing apparatus 1 can be configured as a part of hardware or software, and can be realized by combining them.

Although the present invention has been described with reference to specific embodiments, the present invention is not limited to the above-described embodiments, and can be modified in various ways without departing from the technical idea thereof. For example, in the examples of the embodiments described above, an example of performing noise removal mainly using wavelet degeneration based on second-order or third-order wavelet packet decomposition has been described. It is also possible to perform noise removal processing using Therefore, the image processing apparatus 1 according to the present invention is not limited to the above-described embodiment examples, but is limited only by the description of the claims.

INDUSTRIAL APPLICABILITY According to the present invention, it is possible to suppress blurring of edge components and the like while suppressing image corruption that occurs in moving images that are difficult to encode.

REFERENCE SIGNS LIST 1 image processing device 11 wavelet packet decomposition unit 12 continuous signal power detection unit 13 noise power detection unit 14 degeneration function setting unit for restricted band 15 degeneration function setting unit for pass band 16 wavelet degeneration unit 17 reconstruction unit

Claims (4)

An image processing device for inputting a moving image for each frame image and performing image processing using wavelet degeneration based on wavelet packet decomposition,
a wavelet packet decomposition unit that performs n (n≧2)-th order wavelet packet decomposition processing on the frame image and extracts frequency decomposition coefficients of each frequency band;
A predetermined spatial high-frequency band group is extracted from each of the frequency bands as a restricted band, and a frequency resolution coefficient equal to or greater than a predetermined threshold for the horizontal high-frequency band, the vertical high-frequency band, and the diagonal high-frequency band within the restricted band. is regarded as a continuous signal component, and the average value or standard deviation of the frequency resolution coefficients equal to or greater than the predetermined threshold, or the median value is detected as the continuous signal power of each frequency band group in the restricted band;
a degeneracy function setting unit for a restricted band that sets a degeneracy function for each frequency band group in the restricted band based on at least the continuous signal power and a restriction amount parameter that specifies the amount of band restriction;
A wavelet for band-limiting the frequency resolution coefficients within the restricted band by applying the degeneration function to the frequency resolution coefficients of each frequency band obtained by performing the n-th order wavelet packet decomposition processing and executing the wavelet degeneration processing. a degenerate portion;
a reconstruction unit that generates an output image by performing wavelet reconstruction processing using the frequency-resolved coefficients after the wavelet degeneration processing;
Among the frequency resolution coefficients of each frequency band obtained by performing the n-th order wavelet packet decomposition processing, the power of all the frequency resolution coefficients in the spatial diagonal highest frequency band is regarded as a noise component that is equal to or greater than a predetermined ratio. , and a noise power detection unit for detecting as noise power ,
The degeneracy function setting unit for the limited band is characterized in that it sets the degeneracy function for each frequency band group in the limited band based on the continuous signal power, the noise power, and the limit amount parameter. Image processing device. A passband degeneration function setting unit that sets a frequency band group other than the restricted band as a passband and sets a degeneration function for the frequency band group in the passband based on the noise power,
The wavelet degeneration unit applies the degeneration functions set by the restricted band degeneration function setting unit and the passband degeneration function setting unit to perform wavelet degeneration processing, thereby performing frequency decomposition within the restricted band. 2. The image processing apparatus according to claim 1 , wherein band limiting and noise elimination of coefficients and noise elimination of frequency-resolved coefficients within said passband are performed. The limiting amount parameter includes a first band limiting amount for band limiting as a whole within the limited band and a second band limiting amount based on the continuous signal power, The image processing apparatus according to claim 1 or 2 . A program for causing a computer to function as the image processing apparatus according to any one of claims 1 to 3 .

Images