JP2008293425A - Noise removal device, program, and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a noise removal device, a program and a method for achieving noise removal whose edge storage performance is high by directional filter processing, and for removing a pattern-shaped noise with low arithmetic quantity and a low circuit scale in a flat part. <P>SOLUTION: This noise removal device performs filter processing corresponding to the directions of edge components included in an input image to the image, and applies coring processing having different characteristics in the flat part and edge part of the image, and is provided with: an evaluation value calculation means for calculating an evaluation value concerning a plurality of directions; a direction discrimination processing means for discriminating the directions of the edge components from the evaluation value; a flatness discrimination processing means for discriminating the flat part or the edge part from the evaluation value; a directional filter processing means for operating filter processing according to the result of the direction discrimination processing means; and a coring processing means for operating coring processing according to the result of the flatness discrimination processing means. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to image processing, and more particularly to a noise removal apparatus, program, and method for removing image noise.

  In a noise removal method for removing a noise signal included in an image signal, smoothing processing using a low-pass filter is widely performed. However, when noise is reduced by a smoothing process using a low-pass filter, the edge component included in the image signal is also smoothed, resulting in a problem that the sharpness of the image is lowered.

  As a method for dealing with this problem, Patent Document 1 detects the direction of edge components included in an image, and applies a smoothing process along the direction to reduce noise without reducing the sharpness of the image. The method to do is shown.

  Further, some application examples of noise removal by the filter having the directionality as described above are disclosed. In Patent Document 2, a method for stabilizing a direction discrimination processing result by using an image that is referred to when detecting the direction of an edge component as an image obtained by performing low-pass filter processing on the original image, instead of the original image. In Patent Document 3, a method of applying a directional filter to the converted signal of each band is shown in noise removal based on multi-resolution conversion. Further, Patent Document 4 shows an example in which the coring threshold is adaptively changed as described later.

FIG. 10 is a diagram illustrating a noise removal method using a directional filter.
In FIG. 10, the input image 400 is subjected to noise removal to be an output image 404. In the input image 400, the direction determination processing unit 401 detects the edge direction of the target pixel, and the direction determination result 402 as a result is output to the directionality filter processing unit 403. The directional filter processing unit 403 performs filter processing on the input image 400 by determining a filter coefficient or a filter processing method based on the edge direction indicated by the direction determination result 402.

Next, details of the operation of the direction determination processing unit 401 will be described with reference to FIG.
A00 to A44 in FIG. 5 indicate pixels of a pixel block in the input image 400, and the center pixel A22 corresponds to the target pixel. In the direction determination process, an evaluation value is calculated for each direction in order to determine which of the directions 1 to 4 is the direction along the edge. Although various things can be considered as an evaluation value, the following evaluation values can be considered as those having a relatively small amount of calculation.
E1 = | A22−A23 | + | A22−A24 | + | A22−A21 | + | A22−A20 |
E2 = | A22−A13 | + | A22−A04 | + | A22−A31 | + | A22−A40 |
E3 = | A22−A12 | + | A22−A02 | + | A22−A32 | + | A22−A42 |
E4 = | A22−A11 | + | A22−A00 | + | A22−A33 | + | A22−A44 |

E1 to E4 are evaluation values corresponding to directions 1 to 4, respectively. The evaluation value increases when an edge exists in the direction to be evaluated, and decreases when there is no edge. Therefore, by detecting the minimum value from E1 to E4 and determining that there is no edge in the corresponding direction, the direction is determined as the direction along the edge.
JP 55-133179 A JP-A-8-202870 JP 2001-57677 A JP 2005-295488 A

  However, a filter with directionality has a high ability to preserve edges, but tends to preserve noise that is present in flat areas as minute edges, which causes pattern-like noise to remain in the output image. There is a problem of doing. In order to deal with this problem, it is conceivable to combine it with a coring process for the purpose of eliminating pattern noise. Here, the coring process is typically such that when the absolute value of the input signal is less than or equal to the threshold value, the signal is uniformly zero, and when the input signal is greater than the threshold value, the signal is obtained by subtracting the absolute value by the threshold value. It is processing.

  In this way, even if the noise remains as a fine pattern-like noise by the directional filter processing, the subsequent coring processing does not affect the edge component having a large value and removes the small noise. can do. However, when the amount of noise is large, the amount of residual pattern-like noise is also large, and it is necessary to increase the threshold value in order to remove this by coring processing. If it does so, the influence with respect to an edge component cannot be disregarded, and the effect which preserve | saved the edge by the directionality filter will be attenuated.

  Therefore, also in the coring process, it is necessary to distinguish between the edge portion and the flat portion, and to reduce the threshold value at the edge portion and increase the threshold value at the flat portion. Patent Document 4 is an example in which the coring threshold is adaptively changed in this way, but it is necessary to observe the change in the pixel value in the peripheral region of the pixel of interest in order to determine whether the edge portion or the flat portion, The amount of computation was large.

  As described above, the direction of the edge component included in the image is determined (direction determination process), and the smoothing process (filter process) along the direction is applied, thereby reducing noise without reducing the sharpness of the image. However, in order to effectively remove noise, it is necessary to discriminate between edge portions and flat portions (flatness discrimination processing) and change the coring threshold value for each. However, in the prior art, the above-described direction determination processing and flatness determination processing are calculated based on different indexes, and each processing is performed separately. As a result, there is a problem that the amount of calculation increases.

  The present invention has been made in view of the above problems, and achieves noise removal with high edge preservation performance by directional filter processing, and reduces pattern-shaped noise, which is a problem in a flat portion, with a low calculation amount and a low circuit scale. An object of the present invention is to provide a noise removing device, a program, and a method for removing the noise.

In order to solve the above problems, the present invention employs the following means.
The noise removal apparatus according to the present invention performs filtering on an input image according to the direction of an edge component included in the image, and has different characteristics between a flat portion and an edge portion of the image. A noise removal apparatus to which processing is applied, wherein an evaluation value calculation unit that calculates evaluation values related to a plurality of directions, a direction determination processing unit that determines a direction of an edge component from the evaluation values, and a flat portion from the evaluation values Flatness discrimination processing means for discriminating an edge portion, directional filter processing means for performing the filter processing based on the result of the direction discrimination processing means, and coring processing for performing the coring processing based on the result of the flatness discrimination processing means Means.

  According to the noise removal apparatus of the present invention, the evaluation value calculation is shared in the direction determination processing and the flat portion determination processing, so that pattern-like noise that is a problem in the flat portion is removed with a low calculation amount and a low circuit scale. Is possible.

  The noise removal program according to the present invention includes a filtering process according to a direction of an edge component included in an input image and a coring process having different characteristics between a flat portion and an edge portion of the image. Is a noise removal program that causes a computer to execute evaluation value calculation processing for calculating evaluation values for a plurality of directions, direction determination processing for determining the direction of an edge component from the evaluation values, and a flat portion from the evaluation values. Flatness discrimination processing for discriminating between the edge determination portion and the edge portion, and causing the computer to execute the filtering processing based on the result of the direction discrimination processing and the coring processing based on the result of the flatness discrimination processing. It is characterized by.

  An imaging system according to the present invention includes the above-described noise removing device.

  According to the imaging system of the present invention, it is possible to obtain an image suitable for a noise level of a captured image and a user's preference.

  The noise removal method according to the present invention performs a filtering process on the input image according to the direction of the edge component included in the image, and has different characteristics between the flat portion and the edge portion of the image. A noise removal method that applies processing, an evaluation value calculation step that calculates evaluation values related to a plurality of directions, a direction determination processing step that determines the direction of an edge component from the evaluation values, and a flat portion from the evaluation values A flatness discriminating process step for discriminating whether it is an edge portion, controlling the filtering process according to the result of the direction discriminating process, and controlling the coring process according to the result of the flatness discriminating process. .

  According to the noise removal apparatus, program, and method of the present invention, noise removal with high edge preservation performance is realized by directional filter processing, and pattern-like noise that is a problem in a flat portion is reduced with a small amount of computation and a low circuit scale. It can be removed.

[First Embodiment]
Hereinafter, a first embodiment of a noise removing device, a program, and a method according to the present invention will be described with reference to the drawings.
FIG. 1 is a functional block diagram showing the functions provided in the noise removal device according to the present embodiment.
The noise removal apparatus according to the present embodiment discriminates the direction of an edge component from an evaluation value calculation unit 103 that calculates evaluation values for a plurality of directions with respect to an input image, and an evaluation value given by the evaluation value calculation unit 103. Filter processing is performed based on the results of the direction discrimination processing unit 105, the flatness discrimination processing unit 107 that discriminates the flat portion or the edge portion from the evaluation value given from the evaluation value calculation unit 103, and the direction discrimination processing unit 105. A directional filter processing unit 101 and a coring processing unit 102 that performs a coring process based on the result of the flatness discrimination processing unit 107 are provided.

  In the noise removing apparatus having the above-described configuration, the input image 100 is subjected to filter processing in the direction along the edge direction by the directional filter processing unit 101, and then residual noise on the flat portion is detected by the coring processing unit 102. The output image 109 is removed more strongly than the edge portion. Here, the filter characteristic applied by the directional filter processing unit 101 is controlled by a direction discrimination result 106 that is an output of the direction discrimination processing unit 105.

  The evaluation value calculation unit 103 calculates an evaluation value 104 corresponding to each direction candidate for direction determination with reference to the input image 100, and the direction determination processing unit 105 performs direction determination based on the evaluation value 104. Process. The evaluation value 104 is also output to the flatness discrimination processing unit 107, and the flatness discrimination processing unit 107 displays a flatness discrimination processing result 108 indicating whether the target pixel of the image is located at the flat part or the edge part. It outputs to the coring process part 102, and controls the intensity | strength of a coring process.

Hereinafter, details of each processing unit will be described mainly with reference to FIG.
First, the evaluation value calculation unit 103 and the direction determination processing unit 105 will be described. Here, a case will be described as an example in which a 5 × 5 pixel block around a pixel of interest is used when determining the edge direction of a certain pixel of interest, and the edge direction to be determined is four directions. The block size can be set to other sizes such as 7 × 7 as required, and the number of directions can be increased to 8 directions or the like as required.

Assume that A00 to A44 in FIG. 5 are pixels of the pixel block in the input image 100, and the center pixel A22 corresponds to the target pixel position. In the direction discrimination, the evaluation value calculation unit 103 calculates an evaluation value for each direction in order to determine which of the directions 1 to 4 is the direction along the edge. Although various things can be considered as an evaluation value, in the present embodiment, the following evaluation values are used.
E1 = | A22−A23 | + | A22−A24 | + | A22−A21 | + | A22−A20 |
E2 = | A22−A13 | + | A22−A04 | + | A22−A31 | + | A22−A40 |
E3 = | A22−A12 | + | A22−A02 | + | A22−A32 | + | A22−A42 |
E4 = | A22−A11 | + | A22−A00 | + | A22−A33 | + | A22−A44 |

  E1 to E4 are evaluation values 104 corresponding to directions 1 to 4, respectively. The evaluation value 104 increases when an edge exists in the direction to be evaluated, and decreases when there is no edge. The direction determination processing unit 105 detects the minimum value from E1 to E4 of the evaluation value 104 and determines that the corresponding direction does not have an edge, thereby determining the direction as the direction along the edge, The direction determination result 106 is used.

Next, the evaluation value calculation unit 103 and the flatness determination processing unit 107 will be described.
Based on the evaluation value 104 calculated by the evaluation value calculation unit 103, the flatness determination processing unit 107 outputs a flatness determination processing result 108 indicating whether the pixel of interest of the image is located at the flat part or the edge part. To do. The evaluation values E1 to E4 in the present embodiment have a characteristic that the value increases when the direction crosses the edge. Therefore, the flatness discrimination processing unit 107 calculates the maximum value from E1 to E4, and uses the maximum value as the flatness discrimination processing result 108 representing the flatness. If this value is large, there will be a strong edge signal around the pixel of interest, and if this value is small, there will be no strong edge signal around the pixel of interest, and the characteristics of the flat part will be strong.

  The coring processing unit 102 receives the flatness discrimination processing result 108, sets a coring threshold value for the pixel of interest according to this value, and performs coring processing. When the flatness discrimination processing result 108 is a small value, the target pixel exists in the flat portion. Therefore, the coring threshold is increased, and when the flatness discrimination processing result 108 is a large value, it corresponds to the edge portion. Therefore, the coring threshold is reduced. The relationship between such a coring threshold and the flatness discrimination processing result 108 is shown in FIG. For example, the coring processing unit 102 stores the relationship shown in FIG. 2 set in advance as a coring threshold setting table, and sets the coring threshold corresponding to the calculated flatness discrimination processing result 108 to the coring threshold setting. A coring threshold is set by reading from the table.

  As described above, according to the noise removal method according to the present embodiment, by sharing the evaluation value calculation in the direction determination process and the flat part determination process, it is possible to reduce pattern-shaped noise that is a problem in the flat part with a low calculation amount. It becomes possible to remove it with a low circuit scale.

[Second Embodiment]
Next, a second embodiment of the present invention will be described with reference to FIG.
The noise removal apparatus and the program and method according to the present embodiment are different from the first embodiment in that the input image is divided into at least two band images, and an evaluation value is calculated for each band image. The point is that the direction determination process and the flatness determination process are performed using the evaluation value. Hereinafter, the noise removal method, the program, and the apparatus of the present embodiment will not be described with respect to the points common to the first embodiment, and different points will be mainly described.

FIG. 3 is a functional block diagram showing the functions provided in the noise removal apparatus according to the present embodiment. Here, a noise removal method based on multi-resolution conversion using a Laplacian pyramid will be described as an example.
The noise removal apparatus according to the present embodiment includes a multiresolution conversion unit that performs multiresolution conversion using a Laplacian pyramid, and an NR processing unit that performs noise removal on signals in each band.

The multi-resolution conversion unit includes a plurality of stages of filter processing / reduction processing units 201, 210, 218, and enlargement processing units 202, 211, 219 provided corresponding to the filter processing / reduction processing units 201, 210, 218, and the like. And a synthesizing unit that re-synthesizes the signal using the image after noise removal.
In such a configuration, the input image 200 is reduced by the filter processing / reduction processing unit 201 after the low-pass filter processing. By this processing, the image size is halved in both the horizontal direction and the vertical direction. The reduced image is supplied to the filter processing / reduction processing unit 210 at the next stage, and further reduced further.

  The image reduced in this way is enlarged to the original image size by the enlargement processing unit 202, and then a band image 204 is created by the subtractor 203. This band image 204 corresponds to a signal in a high frequency part cut off by the low-pass filter characteristic used in the filter processing / reduction processing unit 201. The NR processing unit 205 performs noise removal processing on the band image 204 containing a lot of high frequency components. Similarly in the next stage processing, a band image corresponding to the next stage is generated by the filter processing / reduction processing unit 210, the enlargement processing unit 211, and the subtractor 212, and the NR processing unit 213 performs noise removal processing on the band image. Apply. In this way, noise removal processing is applied to the sequentially band-divided signal.

  The output of the NR processing unit 221 is added to the signal on the low frequency side by the adder 224, enlarged by the expansion processing unit 225, and then sent to the processing on the high frequency side. In this way, the signals are sequentially recombined from the low frequency side, and finally the highest frequency signal is added by the adder 209 to form an output image 226.

FIG. 4 is a diagram illustrating details of the NR processing unit 205. The NR processing unit 213, the NR processing unit 221 and the like, which are NR processing units in other bands, have the same configuration as the NR processing unit 205, and a description thereof will be omitted.
The band image 204 is filtered in the direction along the edge direction by the directional filter processing unit 301, and then residual noise in the flat portion is more strongly removed than the edge portion by the coring processing unit 302. , The output signal 206. The filter characteristics applied by the directional filter processing unit 301 are controlled by a direction discrimination processing result 306 that is an output of the direction discrimination processing unit 305.

  In the direction discrimination processing unit 305, the direction discrimination processing is executed based on the evaluation value 304 calculated by the evaluation value calculation unit 303. The evaluation value 304 is also output to the flatness discrimination processing unit 307, and the flatness discrimination processing unit 307 displays a flatness discrimination processing result 308 that indicates whether the target pixel of the image is located at the flat part or the edge part. It outputs to the coring process part 302 and controls the intensity | strength of a coring process.

Similar to the evaluation value calculation unit 103 in the first embodiment, the evaluation value calculation unit 303 calculates an evaluation value for determining the edge direction of the target pixel.
Here, in the first embodiment, pixel blocks are generated from the input image 100 (A00 to A44), and the evaluation values E1 to E4 are calculated. On the other hand, in this embodiment, a pixel block is generated from a high-frequency image (first image) 207 including a band image 204 and a low-frequency image (second image) 208, and It is assumed that a similar evaluation value is generated.

In generating the pixel block, if the pixel block is generated only from the low-frequency image 208, direction determination with little influence of noise can be realized. Further, if a pixel block is generated from only the high frequency image 207, the influence of noise remains, but since it has a high frequency component including an edge component, direction determination corresponding to a fine structure can be realized. Alternatively, a pixel block can be generated by combining the high-frequency image 207 and the low-frequency image 208. For example, by performing a weighted average of the high frequency image 207 and the low frequency image 208, it is possible to generate an image (third image) having both characteristics. When the high-frequency image 207 and the low-frequency image 208 are combined, the characteristics of both the high-frequency image 207 and the low-frequency image 208 are used to cope with changes in the fine structure of the input image and to reduce noise. Stable direction discrimination that is not affected can be realized.
Note that the evaluation value calculation unit 303 may be configured so that the above-described method for generating a plurality of pixel blocks can be selected according to the characteristics of the band image 204 or the like.

  The evaluation value 304 is the same as E1 to E4 in the first embodiment, and corresponds to the direction 1 to the direction 4, respectively. When there is an edge in the direction to be evaluated, the evaluation value 304 becomes large and the edge It has a feature that it becomes small when no exists. The direction determination processing unit 305 detects the minimum value from E1 to E4 of the evaluation value 304 and determines that the corresponding direction does not have an edge, thereby determining the direction as the direction along the edge, The direction determination processing result 306 is assumed.

  The evaluation value 304 calculated in the evaluation value calculation unit 303 is used in the flatness discrimination processing unit 307 as in the first embodiment, and whether the target pixel of the image is located in the flat part or the edge part. The flatness discrimination processing result 308 representing is calculated. Similarly to the first embodiment, the coring processing unit 302 receives the flatness determination processing result 308, sets a coring threshold value for the pixel of interest according to this value, and performs coring processing.

As described above, according to the noise removal device according to the present embodiment, it is possible to more efficiently remove noise by combining noise removal based on multi-resolution conversion and directional filter processing. Become.
Even when noise removal based on multi-resolution conversion is combined with noise removal with high edge preservation performance by directional filter processing, in the NR processing in each band as in this embodiment, the direction determination processing and the flat portion are performed. By sharing the evaluation value calculation in the discrimination processing, it is possible to remove pattern-like noise that becomes a problem in the flat portion with a low calculation amount and a low circuit scale.

In the present embodiment, the evaluation value 304 used in the direction determination processing unit 305 and the flatness determination processing unit 307 is the same, but even if they are different, the same evaluation value calculation unit 303 is used. If it is possible to calculate with the above, the effect of realizing with a small circuit scale can still be obtained. For example, in the direction determination process, in order to realize the direction determination process corresponding to the fine structure of the image, the evaluation value is calculated using the high-frequency image 207, and the flatness determination process is stable and hardly affected by noise. In order to realize the discrimination processing, it is also possible to calculate using the low-frequency image 208. In this case, the circuit scale can be reduced by using the circuit of the evaluation value calculation unit 303 for both evaluation value calculations.
In the present embodiment, the noise removal method based on the multi-resolution conversion by the Laplacian pyramid has been described as an example as a method of performing noise removal on each band image by dividing into a plurality of band images. However, the present invention is not limited to this example. It is not something.

[Third Embodiment]
Next, a third embodiment of the present invention will be described with reference to FIG.
In the present embodiment, the noise removal device described in each of the above-described embodiments is incorporated in an imaging system such as a digital camera. The imaging system according to the present embodiment will be described below with reference to FIG. In addition, description is abbreviate | omitted about the point which is common in each above-mentioned embodiment, and a different point is mainly demonstrated.

  The imaging system according to the present embodiment includes a lens system 500, a CCD 501, an image processing unit 502, a noise removal unit 503, an image compression unit 504, a recording medium 505, a noise removal condition setting unit 506, and noise removal. A parameter setting unit 507 is provided as a main component.

  The image signal captured by the CCD 501 through the lens system 500 is subjected to white balance processing, edge enhancement processing, color signal processing, and the like in the image processing unit 502, and then noise is removed in the noise removal unit 503. The image signal from which noise has been removed is compressed into a JPEG format or the like by the image compression unit 504 and then stored in a recording medium 505 such as a memory card. The noise removing unit 503 applies the noise removing device described in each of the above embodiments.

  In removing noise, the image signal output from the image processing unit 502 may be an image composed of RGB signals, and the noise removing unit 503 may individually perform noise removal processing on the RGB image composed of three surfaces. The signal output from the processing unit 502 may be YCbCr, and the noise removal unit 503 may individually perform noise removal processing on the three-sided YCbCr image. Further, the image processing unit 502 may output RGB signals, and after the RGB signals are converted into YCbCr signals or the like in the noise removal unit 503, noise removal processing may be individually performed on the YCbCr image composed of three surfaces. . Further, it is not necessary to separate the noise removing unit 503 and the image processing unit 502 as in the present embodiment. For example, the edge enhancement processing of the image processing unit 502 and the processing order of the noise removing unit 503 can be switched. Moreover, the noise removal part 503 has a function with which the noise removal apparatus which concerns on any one of embodiment mentioned above is equipped.

  The operation of the noise removal unit 503 is controlled by the noise removal parameter setting unit 507. In the noise removal parameter setting unit 507, a plurality of parameter candidates corresponding to the required noise removal strength are stored as noise removal parameters at the time of factory shipment. The noise removal parameter setting unit 507 determines the noise removal strength based on a setting value related to the noise removal strength of the user, ISO sensitivity information from the noise removal condition setting unit 506, and the like, and selects an appropriate parameter from the plurality of parameter candidates. And the noise removal operation of the noise removal unit 503 is controlled.

The parameters related to noise removal may include selection of a configuration related to the present invention. For example, when the noise removal method corresponding to the second embodiment is applied to the noise removal unit 503, the following parameters may be stored in the noise removal parameter setting unit 507 as candidates.
In the second embodiment, a reference image used for direction determination is generated by a combination of a high-frequency image 207 including a high-frequency component and a low-frequency image 208 including a low-frequency component. As a selection related to this combination, whether only the target pixel position is a pixel of the high-frequency image 207, 3 × 3 pixels around the target pixel, or whether the entire pixel is the high-frequency image 207 or the low-frequency image 208 There can be such a choice. By this selection, a balance between direction determination corresponding to fine structural changes and stable direction determination not affected by noise is determined. Therefore, as shown in FIG. 7, a plurality of reference image generation methods as described above may be set as parameter candidates, and a reference image may be generated based on a preset condition. The noise removal parameter setting unit 507 selects an appropriate configuration from user settings, ISO sensitivity information, and the like, and controls the noise removal unit 503.

Note that the settings in each band of the multi-resolution need not be the same, and the settings in the high band and the settings in the low band may be different. For example, the conditions shown in FIG. 7 may be set in advance for each band image signal, and the reference image may be generated based on the conditions.
Alternatively, the number of reductions in multi-resolution conversion may be set in advance according to the noise level, and the number of reductions in multi-resolution conversion may be determined according to the noise level of the input image signal.
Furthermore, the noise removal unit 503 may select a noise removal method according to any embodiment according to a parameter instead of selecting any one of the above-described embodiments. For example, the number of times of resolution conversion in multi-resolution conversion is changed according to imaging conditions (ISO sensitivity, etc.).

  As described above, by adopting the noise removal method according to at least one of the embodiments and configuring the imaging system, stable direction determination that is not affected by noise while also supporting changes in the fine structure of the input image. Can be realized. In addition, by appropriately changing this configuration according to user settings, ISO sensitivity information, and the like, an imaging system capable of obtaining an image suitable for the noise level of the image to be captured and the user's preference can be obtained.

  In each of the above-described embodiments, hardware processing is assumed as the noise removal device, but it is not necessary to be limited to such a configuration. For example, a configuration in which processing is performed separately by software is also possible. In this case, the noise removing device includes a main storage device such as a CPU and a RAM, and a computer-readable recording medium on which a program for realizing all or part of the above processing is recorded. Then, the CPU reads out the program recorded in the storage medium and executes information processing / calculation processing, thereby realizing the same processing as the above-described noise removing apparatus. Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like.

Specific processing by the above software will be described below based on the flowchart shown in FIG.
First, the input image signal is sequentially reduced by performing multi-resolution conversion, and a plurality of reduced images are generated (S1).
Next, each reduced image is enlarged to the image size before the reduction (S2), and subtraction from each image before the reduction is performed, so that each band image to be subjected to noise removal is created (S3). ).
Next, noise removal processing is performed on each created band image (S4).
Then, each band image after noise removal is recombined, and the recombined image is output as an output image (S5).

Next, details of the noise removal processing (S4) will be described mainly with reference to FIG. Here, as an example, the noise removal process of the band image 204 illustrated in FIG. 3 among the plurality of band images created in the above-described process will be described.
First, in the noise removal process, an evaluation value is calculated for each direction in order to determine the direction along the edge (S11). In this process, a reference image is created using the high-frequency image 207 and the low-frequency image 208, and evaluation values are calculated for each direction of the reference image.
Next, a direction discrimination process is performed based on the evaluation value for each direction (S12). Specifically, the minimum value is detected among the evaluation values, and it is determined that no edge exists in the direction corresponding to the evaluation value. Thereby, it is determined that the direction corresponding to the evaluation value is the direction along the edge.
Next, a filter coefficient or the like is determined based on the edge direction indicated by the result of the direction discrimination process, and the filter process is performed on the band image 204 that is the target of noise removal (S13).
Next, flatness determination processing is performed based on the evaluation values for each direction (S14). Specifically, the maximum value among the evaluation values is detected, and if the evaluation value is large, a strong edge signal exists around the pixel of interest, and if the evaluation value is small, it is determined that there is no strong edge signal. Is done. Thereby, the flatness indicating whether the target pixel is located in the flat part or the edge part is determined.

Next, a coring process is performed on the band image that has been subjected to the filter process so that the minute signal becomes zero, and the band image 206 from which noise is removed is output (S15).

In the above example, the band image 204 created in the first stage processing of the multi-resolution conversion has been described. However, the same processing is applied to each band image created in the second and subsequent stages of multi-resolution conversion. Done.
Furthermore, in the above, the noise removal method corresponding to the second embodiment has been described, but other embodiments can be similarly processed by software.

It is the functional block diagram which expanded and showed the function with which the noise removal apparatus which concerns on the 1st Embodiment of this invention is provided. It is a graph which shows the relationship between the coring threshold value and flatness discrimination | determination result of the noise removal apparatus of FIG. It is the functional block diagram which expanded and showed the function with which the noise removal apparatus which concerns on the 2nd Embodiment of this invention is provided. FIG. 4 is a functional block diagram of an NR processing unit of the noise removal device shown in FIG. 3. It is a figure explaining operation | movement of a direction discrimination | determination process part. It is a schematic block diagram of the imaging system which concerns on the 3rd Embodiment of this invention. It is a condition setting table | surface which shows the conditions of the production | generation method of a pixel block. It is a flowchart of the noise removal method which concerns on this invention. It is a flowchart of the noise removal process in FIG. It is the functional block diagram which expanded and showed the function with which the conventional noise removal apparatus is provided.

Explanation of symbols

100 Input image 101, 301 Directional filter processing unit 102, 302 Coring processing unit 103, 303 Evaluation value calculation unit 104, 304 Evaluation value 105, 305 Direction discrimination processing unit 107, 307 Flatness discrimination processing unit 204 Band image 207 High Region image 208 Low region image

Claims (10)

A noise removing apparatus that applies a filtering process to an input image according to a direction of an edge component included in the image and applies a coring process having different characteristics between a flat portion and an edge portion of the image. And
Evaluation value calculation means for calculating evaluation values for a plurality of directions;
Direction discrimination processing means for discriminating the direction of the edge component from the evaluation value;
Flatness discrimination processing means for discriminating whether it is a flat portion or an edge portion from the evaluation value; and directional filter processing means for performing the filter processing according to a result of the direction discrimination processing means;
Coring processing means for performing the coring processing according to the result of the flatness discrimination processing means;
Noise elimination device with   The input image is divided into at least two band images, the evaluation value is calculated for each of the band images, and direction determination processing and flatness determination processing are performed using the evaluation values. Noise removal device.   The evaluation value calculation unit calculates the evaluation value using a first image including information on a frequency band of a band image to be subjected to noise removal and a frequency on a lower frequency side than the frequency band. The noise removal apparatus as described.   3. The noise removal according to claim 2, wherein the evaluation value calculation means calculates the evaluation value using a second image including information on a frequency lower than a frequency band of a band image to be subjected to noise removal. apparatus.   The evaluation value calculation means includes a first image including information on a frequency band of a band image to be subjected to noise removal and a frequency on a lower frequency side than the frequency band, and a lower frequency side than the frequency band of the band image. The noise removal apparatus according to claim 2, wherein the evaluation value is calculated using a third image calculated by performing a weighted average with a second image including frequency information.   The evaluation value calculation means includes: a first image including information on a frequency band of a band image to be subjected to noise removal and a frequency on a lower frequency side than the frequency band; and a frequency band of the band image to be noise removed At least one of a second image including frequency information on a lower frequency side and a third image calculated by performing a weighted average of the first image and the second image. The noise removal apparatus according to claim 2, wherein the evaluation value is selected using the selected image. The evaluation value calculation means calculates an evaluation value corresponding to the magnitude of the pixel value change in the direction of evaluation,
The direction determination processing means determines the direction based on the minimum value of the evaluation value,
The noise removal apparatus according to claim 1, wherein the flatness determination processing unit performs flatness determination based on a maximum value of the evaluation values. A noise removal program that causes a computer to execute filtering processing according to the direction of an edge component included in the input image and coring processing having different characteristics between a flat portion and an edge portion of the image. Because
Evaluation value calculation processing for calculating evaluation values for a plurality of directions;
Direction discrimination processing for discriminating the direction of the edge component from the evaluation value;
And flatness determination processing for determining whether the evaluation value is a flat portion or an edge portion,
The noise removal program which makes a computer perform the said filter process based on the result of the said direction discrimination | determination process, and the said coring process based on the result of the said flatness discrimination | determination process.   An imaging system comprising the noise removal device according to claim 1. This is a noise removal method that applies a filtering process to an input image according to the direction of the edge component included in the image, and applies a coring process having different characteristics between the flat part and the edge part of the image. And
An evaluation value calculation process for calculating evaluation values for a plurality of directions;
A direction determining process for determining the direction of the edge component from the evaluation value;
A flatness determination process for determining whether the evaluation value is a flat portion or an edge portion, and
A noise removal method of controlling the filtering process according to a result of the direction determining process step and controlling the coring process according to a result of the flatness determining process step.

Images