JP4261201B2 - Image processing apparatus, image processing method, image processing program, and storage medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to an image processing apparatus., Image processing methodThe present invention relates to an image processing program and a storage medium.
[0002]
[Prior art]
For example, an image obtained by spatially dividing a single image into a plurality of frequency bands using wavelet transform is controlled for each frequency band, and only an image in a certain frequency band is emphasized by multiplication and then inverse conversion is performed. By doing so, there is a technique for performing natural image enhancement suitable for a visual impression (see, for example, Patent Document 1). In such a technique, since the image is spatially divided for each direction by using the wavelet transform, the image can be controlled for each direction.
[0003]
Further, there is a technique in which the sharpening degree of an image is adjusted by performing a filtering process for each frequency band and for each direction in a real space (see, for example, Patent Document 2).
[0004]
[Patent Document 1]
JP-A-6-274614
[Patent Document 2]
JP-A-9-247460
[0005]
[Problems to be solved by the invention]
However, as described in Patent Document 1, an image obtained by spatially dividing a single image into a plurality of frequency bands using wavelet transform is controlled for each frequency band, and an image in a certain frequency band is obtained. In the technique in which only the multiplication is performed and enhancement is performed and then the inverse transformation is performed, as described above, the image is decomposed in each direction by using the wavelet transformation. For example, the low frequency band image is enhanced. This requires a lot of line memory.
[0006]
On the other hand, according to the filtering process in the real space that performs the filtering process for each frequency band and in the direction in the real space as in the technique described in Patent Document 2, there is a degree of freedom in setting the frequency band of the filter. Therefore, it is advantageous to perform processing with a limited line memory.
[0007]
However, when performing real-space filtering processing, different filters must be used for each frequency band and for each direction, and product-sum operation is performed for all the different filters for each frequency band and for each direction. Therefore, there is a concern that the scale of hardware such as a circuit that performs a product-sum operation is increased. In particular, in order to perform real space filtering processing on an image in a low frequency band, a filter having a large matrix size is required, and there is a concern that the increase in hardware scale will be more remarkably caused.
[0008]
An object of the present invention is to obtain an image with good resolution and sharpness while suppressing an increase in hardware scale.
[0009]
[Means for Solving the Problems]
  The image processing apparatus according to claim 1 is a spatial frequency for an arbitrary spatial frequency component of the input image, using image input means for acquiring the input image and a plurality of directional filters each having a different directionality. First spatial frequency correction means for performing the correction independently for each direction determined by the directionality of the directional filter to obtain a first corrected image, and the frequency characteristics of the directional filter independent of the directionality AlsoThe amplitude peak is on the low frequency side.Second spatial frequency correction means for performing a spatial frequency correction on the spatial frequency component of the input image to obtain a second corrected image using a single filter; the first corrected image and the second corrected image; And image combining means for combining the two.
[0010]
  Therefore, spatial frequency correction for each direction is combined with spatial frequency correction that does not depend on directionality, and spatial frequency correction is performed for each direction or spatial frequency correction that does not depend on directionality according to the spatial frequency characteristics of the input image. Therefore, by appropriately setting the spatial frequency for performing spatial frequency correction for each direction, it is possible to suppress an increase in hardware scale and obtain an image with good resolution and sharpness.In general, when spatial frequency correction is performed for each direction, a filter with a large matrix size is required. By performing spatial frequency correction of low frequency components of the input image with a single filter, an increase in hardware scale is effective. Can be suppressed.
[0011]
According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the spatial frequency correction by the first spatial frequency correction unit and the spatial frequency correction by the second spatial frequency correction unit are performed in parallel.
[0012]
Therefore, the processing speed can be improved.
[0015]
  Claim 3The invention described isClaim 1 or 2In the image processing device described above, a feature amount acquisition unit that acquires a feature amount of the input image along the directionality of the directional filter, and spatial frequency correction is performed on the first correction image according to the feature amount. And third spatial frequency correction means for performing.
[0016]
  Therefore, the spatial frequency correction is performed for each direction by the first spatial frequency correction means.First corrected imageBy correcting the spatial frequency according to the feature amount acquired along the directionality of the directional filter,Input imageAn image with good resolution and sharpness can be obtained without impairing the above characteristics.
[0017]
  Claim 4The invention described isClaim 3In the image processing apparatus described above, a maximum value acquisition unit that acquires a maximum value of the feature amount for each direction in which the feature amount is acquired, and a second that performs spatial frequency correction on the second correction image according to the maximum value. 4 spatial frequency correction means.
[0018]
  Therefore, the spatial frequency was corrected by the second spatial frequency correction means.Second corrected imageBy correcting the spatial frequency according to the maximum value of the feature amount acquired along the directionality of the directionality filter, for example,Input imageEven in the case where the low frequency component is corrected by the second spatial frequency correction means, the increase in hardware scale is effectively suppressed,Input imageAn image with good resolution and sharpness can be obtained without impairing the above characteristics.
[0019]
  Claim 5The invention described isClaim 1 or 2In the image processing device described above, a feature amount acquisition unit that acquires the feature amount of the input image along the directionality of the directional filter, and a maximum value of the feature amount for each direction in which the feature amount is acquired And a fourth spatial frequency correction unit that performs spatial frequency correction on the second corrected image according to the maximum value.
[0020]
  Therefore, the spatial frequency was corrected by the second spatial frequency correction means.Second corrected imageBy correcting the spatial frequency according to the maximum value of the feature amount acquired along the directionality of the directionality filter, for example,Input imageIn the case where the low frequency component is corrected by the second spatial frequency correcting means,Input imageThe increase in hardware scale can be effectively suppressed without impairing the characteristics of the above.
[0021]
  Claim 6The invention described isAny one of claims 3 to 5In the image processing apparatus described above, the feature amount acquisition unit acquires an edge amount of the input image as the feature amount by a differential filter.
[0022]
Therefore, the edge of the image is not dulled.
[0023]
  The image processing method according to claim 7 acquires an input image, and performs spatial frequency correction on an arbitrary spatial frequency component of the input image by using a plurality of directional filters each having a different directionality in the direction. The first correction image is obtained independently for each direction determined by the directionality of the directional filter, and the frequency characteristics of the directional filter are not dependent on the directionality.The amplitude peak is on the low frequency side.Using a single filter, spatial frequency correction is performed on the spatial frequency component of the input image to obtain a second corrected image, and the first corrected image and the second corrected image are synthesized.
[0024]
  Therefore, spatial frequency correction for each direction is combined with spatial frequency correction that does not depend on directionality, and spatial frequency correction is performed for each direction or spatial frequency correction that does not depend on directionality according to the spatial frequency characteristics of the input image. Therefore, by appropriately setting the spatial frequency for performing spatial frequency correction for each direction, it is possible to suppress an increase in hardware scale and obtain an image with good resolution and sharpness.In general, when spatial frequency correction is performed for each direction, a filter with a large matrix size is required. By performing spatial frequency correction of low frequency components of the input image with a single filter, an increase in hardware scale is effective. Can be suppressed.
[0025]
  Claim 8The image processing program of the invention described isClaim 7The computer executes the image processing method described in (1).
[0026]
  Therefore,Claim 7The effects of the described invention can be obtained.
[0027]
  Claim 9The storage medium of the described invention isClaim 8Is stored in a readable manner.
[0028]
  Therefore,Claim 8The effects of the described invention can be obtained.
[0039]
DETAILED DESCRIPTION OF THE INVENTION
Next, a first embodiment of the present invention will be described with reference to FIGS. As an example, the image processing apparatus of the present embodiment is realized by a personal computer as shown in FIG.
[0040]
FIG. 1 is a block diagram schematically showing a hardware configuration of a personal computer (hereinafter referred to as a PC) as an image processing apparatus according to the present embodiment. As shown in FIG. 1, the PC 1 includes a CPU (Central Processing Unit) 2 that performs various types of information processing and drives and controls each unit included in the PC 1. Various information is transmitted to the CPU 2 via a bus 3. Are connected to a memory 4 such as a ROM (Read Only Memory) or a RAM (Random Access Memory). The memory 4 stores various control programs executed by the CPU 2, information used for various calculations by the CPU 2, and the like. In particular, the RAM functions as a work area for the CPU 2 because it has the property of storing various data in a rewritable manner.
[0041]
A media information reader 7 for reading information stored in a magnetic storage device 5 such as a hard disk or a portable medium 6 is connected to the bus 3 via a predetermined interface (not shown).
[0042]
The magnetic storage device 5 stores an operating system (hereinafter referred to as OS) and various application programs that operate on the OS including the image processing program of the present invention. The OS is a program for managing computer hardware and software. Typical examples of such an OS include Windows (registered trademark), UNIX (registered trademark), and the like. In the present embodiment, various programs operating on the operating system are application programs.
[0043]
The application program is read from the medium 6 by the media information reading device 7 and installed in the magnetic storage device 5, or is installed in the magnetic storage device 5 by downloading from the network 9 via the communication I / F 8 described later. Is. In the present embodiment, the image processing program read from the portable medium 6 by the media information reading device 7 is stored in the magnetic storage device 5 as one of the application programs. For this reason, in the present embodiment, a storage medium is realized by the portable medium 6. Since the magnetic storage device 5 stores an image processing program installed by reading from the medium 6, the magnetic storage device 5 also functions as a storage medium in this sense.
[0044]
The media information reader 7 uses, for example, various types of portable media 6 such as optical disks such as CDs and DVDs, magneto-optical disks, and flexible disks as storage media, and stores information stored in the portable media. An optical disk drive, a magneto-optical disk drive, a flexible disk drive, or the like is appropriately used depending on the type of the medium 6 to be read.
[0045]
In the PC 1, when the user turns on the power, the CPU 2 activates a program called a loader in the memory 4, reads the OS from the magnetic storage device into the RAM, and activates the OS.
[0046]
In addition to the input device 10 composed of a mouse and a keyboard, a display device 11 such as an LCD and a CRT, and a communication I / F 8 that communicates with the network 9, an optically read image is displayed on the bus 3. An image input device 12 such as a scanner device or a digital camera as an image input means for inputting as a signal is connected.
[0047]
Next, an image quality correction function realized by executing an image quality correction program including an image processing program will be described with reference to FIG. FIG. 2 is a functional block diagram showing an image quality correction function realized by executing an image quality correction program by the PC 1. As shown in FIG. 2, the image quality correction function includes a scaling processing unit 13 that scales an image input by the image input device 12 to a specified magnification, and a filtering process that performs desired spatial frequency correction on the scaled image. Means 14; γ conversion processing means 15 for performing γ conversion so that the image after spatial frequency correction has a desired density characteristic; halftone processing for performing halftone processing such as dither processing and error diffusion processing on the image after density characteristic conversion; This is realized by means 16. Although not particularly illustrated, the PC 1 has a function of converting an input image of R, G, B signals into output data of CMYK signals when the image input by the image input device 12 is a color image as an image quality correction function. Color correction means to be realized is also provided.
[0048]
Here, the scaling process by the scaling process unit 13, the γ conversion process by the γ conversion process unit 15, and the halftone process by the halftone process unit 16 are well-known techniques, so that the description thereof is omitted. The filtering processing performed by the filtering processing means 14 which is a feature in the embodiment will be described below with reference to FIGS. The function of the filtering processing unit 14 is realized by executing an image processing program by the PC 1.
[0049]
FIG. 3 is a functional block diagram showing various functions realized by the filtering processing performed by the filtering processing means 14. As shown in FIG. 3, the filtering processing means 14 smoothes the image after the scaling process by the smoothing filter 20 having the filter matrix shown in FIG. Although the description is omitted because it is a well-known technique, smoothing of an image uses a filter matrix of an appropriate size, and calculates the density value of the target pixel by the average value of the density values of the target pixel and surrounding pixels. It is a process to replace.
[0050]
Further, as shown in FIG. 5, the filtering processing means 14 includes a vertical direction (main scanning direction), a horizontal direction (sub-scanning direction), a right diagonal direction (upwardly rightward direction in FIG. 5), and a left diagonal direction (in FIG. 10). The four high-frequency emphasis filters 21, 22, 23, and 24 having directivity in four directions (upwardly to the left) perform spatial frequency correction in the vertical, horizontal, right diagonal, and left diagonal directions of the image after scaling. .
[0051]
In the present embodiment, a directional filter is realized by the four high-frequency emphasis filters 21, 22, 23, 24, and the vertical, horizontal, right diagonal, The first spatial frequency correction means and the first spatial frequency correction function are realized by the spatial frequency correction of the high frequency component in the diagonally left direction.
[0052]
Here, FIG. 5A shows a filter matrix included in the vertical high-frequency enhancement filter 21, and FIG. 6 shows a 600 dpi image by the vertical high-frequency enhancement filter 21 having the filter matrix shown in FIG. The spatial frequency characteristic is shown. As can be seen from FIG. 6, according to the high frequency emphasis filter 21 in the vertical direction of the present embodiment, the high frequency components of the image are emphasized by performing spatial frequency correction along the vertical direction.
[0053]
Similarly, the horizontal high-frequency emphasis filter 22 having the filter matrix shown in FIG. 5B, the diagonally high-frequency emphasis filter 23 having the filter matrix shown in FIG. 5C, and the filter shown in FIG. 5D. According to the left diagonal high-frequency emphasis filter 24 having a matrix, the high frequency components of the image are respectively corrected with spatial frequency correction along the horizontal, right diagonal, and left diagonal directions for enhancement.
[0054]
On the other hand, the filtering processing means 14 includes four edge amount calculation filters 25, 26, 27, and 28 in the vertical, horizontal, right diagonal, and left diagonal directions having the filter matrix shown in FIGS. Edge amounts in four directions of vertical, horizontal, right diagonal, and left diagonal are respectively calculated. Here, the feature amount acquisition means and the feature amount acquisition function are realized by the calculation of the edge amounts by the four edge amount calculation filters 25, 26, 27, and 28.
[0055]
The calculated edge amounts are converted into enhancement coefficients corresponding to the edge amounts by enhancement coefficient calculation units 29, 30, 31, and 32 corresponding to the four edge amount calculation filters 25, 26, 27, and 28, respectively. In the present embodiment, conversion to the enhancement coefficient is performed based on the relationship between the edge amount and the enhancement coefficient as shown in FIG.
[0056]
Then, the filtering processing means 14 uses the high frequency emphasis filters 21, 22, 23, and 24 in the vertical, horizontal, right diagonal, and left diagonal directions to move the high frequency components of the image along the vertical, horizontal, right diagonal, and left diagonal directions. And multiplying the emphasized image by the enhancement coefficients calculated by the enhancement coefficient calculation units 29, 30, 31, and 32 in the vertical, horizontal, right diagonal, and left diagonal directions in the vertical, horizontal, right diagonal, and left diagonal directions. By multiplying by the units 33, 34, 35, and 36, respectively, the high frequency components of the image are spatially frequency corrected for each direction. Here, the third spatial frequency correcting means and the third spatial frequency correcting function are realized. As a result, the high-frequency component of the image enhanced for each direction is enhanced according to the edge amount that is the feature amount for each direction of the image.
[0057]
Separately, the filtering processing means 14 causes the maximum value calculation unit 37 to execute each direction based on the respective edge amounts calculated by the edge amount calculation filters 25, 26, 27, and 28 in the vertical, horizontal, right diagonal, and left diagonal directions. The maximum value of the edge amount is calculated, and the calculated maximum value of the edge amount in each direction is converted into an enhancement coefficient corresponding to the maximum value by the enhancement coefficient calculation unit 38. Here, the maximum value acquisition unit and the maximum value acquisition function are realized by calculating the maximum value of the edge amount by the maximum value calculation unit 37. In the present embodiment, conversion to an enhancement coefficient based on the maximum value of the edge amount in each direction is performed based on the relationship between the edge amount indicated by the solid line in FIG. 8 and the enhancement coefficient.
[0058]
Further, the filtering processing means 14 performs the spatial frequency correction of the low frequency component of the image after the scaling process by the low frequency enhancement filter 39 having the filter matrix shown in FIG. Here, the second spatial frequency correction means and the second spatial frequency correction function are realized by the spatial frequency correction of the low frequency component of the image by the low frequency enhancement filter 39. Note that the low frequency component of the image subjected to spatial frequency correction by the low frequency enhancement filter 39 has frequency characteristics different from the frequency characteristics of the four high frequency enhancement filters 21, 22, 23, and 24 that are directional filters. The frequency characteristics are lower than the frequency characteristics of the two high-frequency emphasis filters 21, 22, 23, and 24.
[0059]
As can be seen from FIG. 9, the low frequency enhancement filter 39 does not have a specific directionality. In the present embodiment, a single filter independent of the directionality is realized by the low frequency enhancement filter 39. Yes. FIG. 10 shows the spatial frequency characteristics (main scanning direction) of the low-frequency enhancement filter 39 having the filter matrix shown in FIG. As shown in FIG. 10, it can be seen that the low frequency emphasis filter 39 of the present embodiment emphasizes the low frequency components of the image as compared with the high frequency emphasis filter 21 shown in FIG. This enhances the low frequency components of the image.
[0060]
The image in which the low frequency component is enhanced is subjected to spatial frequency correction by being multiplied by the enhancement coefficient calculated by the enhancement coefficient calculation unit 38 by the low frequency multiplication unit 40. Here, the fourth spatial frequency correction means and the fourth spatial frequency correction function are realized. As a result, the low-frequency component of the enhanced image is subjected to spatial frequency correction according to the maximum value of the edge amount that is the feature amount for each direction of the image.
[0061]
In the present embodiment, enhancement of high frequency components and enhancement of low frequency components of an image are performed in parallel.
[0062]
Then, the filtering processing means 14 uses the image smoothed by the smoothing filter 20, the high frequency enhancement amount multiplied by each high frequency multiplication unit 33, 34, 35, 36 and the low frequency enhancement amount multiplied by the low frequency multiplication unit 40. The adders 41, 42, 43, 44, and 45 sequentially add, and the added image is output as an enhanced image. Here, the image synthesizing means and the image synthesizing function are realized by synthesizing the frequency components of the image by the adders 41, 42, 43, 44, and 45.
[0063]
Thus, according to the present embodiment, the spatial frequency correction for each direction by the high frequency enhancement filters 21, 22, 23, and 24 and the spatial frequency correction independent of the directionality by the low frequency enhancement filter 39 are combined, Since the spatial frequency correction of the high frequency component of the image is performed for each direction, and the spatial frequency correction of the low frequency component of the image can be performed without depending on the directionality, the low frequency component of the image is regarded as the high frequency component of the image. Similarly, by emphasizing for each direction, a filter with a large matrix size that is required is not required, and an increase in hardware scale is suppressed, and high frequency components of an image are resolved and sharpened by performing spatial frequency correction for each direction. An image with good properties can be obtained.
[0064]
And according to this Embodiment, since the emphasis of the high frequency component of an image and the emphasis of the low frequency component of an image are performed in parallel, the processing speed of a filtering process can be improved.
[0065]
As described above, enhancing the high frequency components of the image more than necessary also enhances the noise of the image. However, by performing spatial frequency correction for each direction as in this embodiment, the noise can be reduced. It is possible to obtain an image with good resolution by emphasizing only a necessary direction while suppressing emphasis.
[0066]
Further, according to the present embodiment, the enhancement of the low frequency component of the image is performed by the single low frequency enhancement filter 39 so that the enhancement of the low frequency component of the image is the same as the enhancement of the high frequency component of the image. The amount of calculation can be reduced as compared with the case where each is performed independently.
[0067]
Furthermore, according to the present embodiment, the high-frequency component of the image emphasized for each direction is emphasized according to the edge amount that is the feature amount for each direction of the image. An image with good resolution and sharpness can be obtained without impairing the characteristics of the image.
[0068]
In addition, since the low frequency component of the image is emphasized according to the maximum value of the edge amount that is the feature amount for each direction of the image, the edge frequency can be corrected even when the low frequency component of the image is corrected. The increase in the hardware scale can be effectively suppressed without degrading the image quality and damaging the characteristics of the image.
[0069]
The edge amount calculation filters 25, 26, 27, and 28 of the present embodiment calculate the edge amount using a primary differential filter. However, the present invention is not limited to this, and for example, a primary differential filter and a secondary differential filter. It is good also as processing which takes the maximum value of.
[0070]
In addition, the conversion from the edge amount to the enhancement coefficient by the enhancement coefficient calculation units 29, 30, 31, and 32 differs depending on the direction, for example, the enhancement in the right diagonal / left diagonal direction is set to be weaker than the vertical / horizontal directions. May be converted. This makes it possible to emphasize only the necessary direction among the high frequency components of the image.
[0071]
In addition, in this embodiment, image processing is performed on an image input from an image input device such as a scanner device or a digital camera. However, the present invention is not limited to this. For example, another scanner device. Alternatively, an image obtained by reading the image may be input from a medium that stores the image acquired by a digital camera or the like. In this case, a media information reading device that reads an image from a medium that stores the image functions as an image input unit.
[0072]
Next, a second embodiment of the present invention will be described with reference to FIGS. The same parts as those in the first embodiment are denoted by the same reference numerals, and description thereof is also omitted.
[0073]
FIG. 11 is a functional block diagram showing various functions realized by the filtering processing performed by the filtering processing means 14 of the PC 1 of the present embodiment. The filtering processing means 14 of the present embodiment performs vertical, horizontal, and diagonal three-way filtering processing on the high-frequency component and medium-frequency component of the image, and a single processing for the low-frequency component of the image. Filtering processing is performed by the low frequency emphasis filter 91 (see FIG. 22).
[0074]
The filtering processing unit 14 smoothes the image by the smoothing unit 50 in the same manner as in the first embodiment, and the high frequency emphasizing unit 51, the medium frequency emphasizing unit 52, and the low frequency emphasizing unit 53 perform image smoothing. High-, medium-, and low-frequency components are emphasized, and an image obtained by combining the smoothed image and the image that emphasizes high-, medium-, and low-frequency components by sequential addition by the adders 54, 55, and 56 is subjected to image quality correction. Output as an image. The enhancement of the high, medium, and low frequency components of the image by the high frequency enhancement unit 51, the medium frequency enhancement unit 52, and the low frequency enhancement unit 53 is performed independently and in parallel. In the present embodiment, the image synthesizing means and the image synthesizing function are realized by the sequential addition by the adders 54, 55, and 56.
[0075]
The medium frequency component of the image subjected to spatial frequency correction by the intermediate frequency emphasizing unit 52 has a frequency characteristic different from the frequency characteristic targeted by the high frequency emphasizing unit 51 for spatial frequency correction, and the high frequency emphasizing unit 51 corrects the spatial frequency. The frequency characteristic is lower than the target frequency characteristic. Further, the low frequency component of the image subjected to the spatial frequency correction has a frequency characteristic different from the frequency characteristic targeted by the medium frequency enhancement unit 52, and the medium frequency enhancement unit 52 has the spatial frequency. The frequency characteristic is lower than the frequency characteristic to be corrected.
[0076]
Below, the high frequency emphasis part 51, the medium frequency emphasis part 52, and the low frequency emphasis part 53 are each demonstrated.
[0077]
FIG. 12 is a functional block diagram illustrating functions of the high frequency emphasizing unit 51. The high frequency emphasizing unit 51 spatially corrects the high frequency components of the image along the vertical direction by the vertical direction emphasizing unit 57, and corrects the high frequency components of the image along the horizontal direction by the horizontal direction emphasizing unit 58. The high-frequency component of the image is spatially frequency-corrected along the diagonal direction by the diagonal direction enhancement unit 59. The spatial frequency correction of the image by the vertical enhancement unit 57, the spatial frequency correction of the image by the horizontal enhancement unit 58, and the spatial frequency correction of the image by the diagonal direction enhancement unit 59 are performed independently. Further, the spatial frequency correction of the image by the vertical enhancement unit 57, the spatial frequency correction of the image by the horizontal enhancement unit 58, and the spatial frequency correction of the image by the diagonal direction enhancement unit 59 are performed in parallel. Thereby, the processing speed in the high frequency emphasis unit 51 can be improved.
[0078]
The high frequency emphasizing unit 51 uses the high frequency emphasis filters 60, 61, and 62 as directional filters having the filter matrix shown in FIGS. The high-frequency components of the image are emphasized by performing spatial frequency correction along each. The high frequency emphasis filters 60, 61, 62 have the frequency characteristics shown in FIG.
[0079]
Further, the high frequency emphasizing unit 51 uses the edge amount calculation filters 63, 64, and 65 having the filter matrix shown in FIGS. 14A to 14C to each of the edge amounts of the image along the three directions of vertical, horizontal, and diagonal. Is calculated. Here, the feature amount acquisition means and the feature amount acquisition function are realized by the edge amount calculation by the edge amount calculation filters 63, 64, 65.
[0080]
The high frequency emphasizing unit 51 converts the calculated edge amount into an emphasis coefficient by the corresponding emphasis coefficient calculation units 66, 67, and 68 in the same manner as in the first embodiment.
[0081]
Then, the high frequency multiplication units 69, 70, 71 multiply the high frequency components of the image vertically, horizontally, and diagonally by multiplying the image in which the high frequency components of the image are emphasized for each direction by the calculated enhancement coefficient. Spatial frequency correction is performed in each direction. Here, the third spatial frequency correcting means and the third spatial frequency correcting function are realized. As a result, the high-frequency component of the image enhanced for each direction is enhanced according to the edge amount that is the feature amount for each direction of the image.
[0082]
In addition, images spatially corrected by the high frequency multipliers 69, 70, and 71 are synthesized by sequentially adding them by the high frequency adders 72 and 73, and the synthesized image is output as an image in which the high frequency components are emphasized. .
[0083]
FIG. 15 is a functional block diagram illustrating functions of the intermediate frequency emphasizing unit 52. In the same way as the high frequency emphasizing unit 51, the medium frequency emphasizing unit 52 corrects the medium frequency component of the image along the vertical direction by the vertical direction emphasizing unit 74, and the horizontal direction emphasizing unit 75 horizontally converts the medium frequency component of the image Spatial frequency correction is performed along the direction, and the medium frequency component of the image is corrected along the diagonal direction by the diagonal direction enhancement unit 76. The spatial frequency correction of the image by the vertical enhancement unit 74, the spatial frequency correction of the image by the horizontal enhancement unit 75, and the spatial frequency correction of the image by the diagonal direction enhancement unit 76 are performed independently. The spatial frequency correction of the image by the vertical direction enhancement unit 74, the spatial frequency correction of the image by the horizontal direction enhancement unit 75, and the spatial frequency correction of the image by the diagonal direction enhancement unit 76 are performed in parallel. Thereby, the processing speed in the intermediate frequency emphasizing unit 52 can be improved.
[0084]
The medium frequency emphasizing filters 77, 78, and 79 as directional filters having directionality in the vertical, horizontal, and diagonal directions in the medium frequency emphasizing unit 52 have filter matrices shown in FIGS. is doing. Further, each of the medium frequency edge amount calculation filters 80, 81, and 82 in the vertical, horizontal, and diagonal directions has a filter matrix shown in FIGS.
[0085]
FIG. 18 shows the spatial frequency characteristics of the vertical medium frequency emphasizing filter 77 shown in FIG. As can be seen from FIG. 18, the frequency characteristic of the medium frequency emphasizing filter 77 is slightly lower than that of the high frequency emphasizing filters 60, 61, 62 having the frequency characteristic shown in FIG.
[0086]
Further, the medium frequency emphasizing unit 52 uses the medium frequency edge amount calculation filters 80, 81, and 82 having filter matrices shown in FIGS. 17A to 17C to respectively display images along the three directions of vertical, horizontal, and diagonal. The edge amount is calculated. Here, the feature amount acquisition means and the feature amount acquisition function are realized by calculating the edge amounts by the medium frequency edge amount calculation filters 80, 81, and 82.
[0087]
The medium frequency emphasizing unit 52 converts the calculated edge amount into an emphasis coefficient by the corresponding emphasis coefficient calculating units 83, 84, and 85, similarly to the high frequency emphasizing unit 51.
[0088]
Then, by multiplying the image obtained by enhancing the medium frequency component of the image in each direction by the medium frequency multiplication units 86, 87, and 88 by the calculated enhancement coefficient, the medium frequency component of the image is vertically, horizontally, and obliquely obtained. The spatial frequency is corrected in each direction. Here, the third spatial frequency correcting means and the third spatial frequency correcting function are realized. As a result, the medium frequency component of the image enhanced for each direction is enhanced according to the edge amount which is the feature amount for each direction of the image.
[0089]
In addition, the images subjected to spatial frequency correction by the intermediate frequency multipliers 86, 87, 88 are synthesized by sequentially adding them by the intermediate frequency adders 89, 90, and the synthesized image is output as an image in which the intermediate frequency components are emphasized. .
[0090]
In the present embodiment, the first spatial frequency correction unit and the spatial frequency correction of the high frequency component of the image by the high frequency enhancement unit 51 and the spatial frequency correction of the medium frequency component of the image by the medium frequency enhancement unit 52 are performed. A first spatial frequency correction function is realized.
[0091]
FIG. 19 is a functional block diagram illustrating functions of the low frequency emphasizing unit 53. The low frequency emphasizing unit 53 performs spatial frequency correction of the low frequency component of the image by the low frequency emphasizing filter 91 as a single filter having the filter matrix shown in FIG. Here, the second spatial frequency correction means and the second spatial frequency correction function are realized by the spatial frequency correction of the low frequency component of the image by the low frequency enhancement filter 91. Here, FIG. 22 shows the frequency characteristics of the low-frequency emphasis filter 91. As can be seen from FIG. 22, the frequency characteristic of the low frequency emphasizing filter 91 has a frequency characteristic on the lower frequency side as compared with the frequency characteristic of the medium frequency emphasizing filter 77 shown in FIG.
[0092]
Further, the low frequency emphasizing unit 53 calculates the edge amount of the image by the low frequency edge amount calculation filters 92, 93, 94 having the filter characteristics shown in FIGS. Here, the feature amount acquisition means and the feature amount acquisition function are realized by the calculation of the edge amount by the low frequency edge amount calculation filters 92, 93 and 94.
[0093]
The low frequency emphasizing unit 53 converts the calculated edge amount into an emphasis coefficient by the emphasis coefficient calculating unit 96 in the same manner as the high frequency emphasizing unit 51 and the medium frequency emphasizing unit 52.
[0094]
The low frequency multiplication unit 97 multiplies the image obtained by emphasizing the low frequency component of the image in each direction by the calculated enhancement coefficient, thereby reducing the low frequency component of the image in each of the vertical, horizontal, and diagonal directions. Correct each spatial frequency. Here, the fourth spatial frequency correction means and the fourth spatial frequency correction function are realized. As a result, the low-frequency component of the image is enhanced according to the edge amount that is the feature amount for each direction of the image.
[0095]
Thus, according to the present embodiment, the spatial frequency correction for each direction by the high frequency enhancement filters 60, 61, 62 and the medium frequency enhancement filters 77, 78, 79 and the directionality by the low frequency enhancement filter 91 are not dependent. In combination with spatial frequency correction, the high frequency component of the image and the medium frequency component can be corrected for each direction, and the low frequency component of the image can be corrected without having directionality. By emphasizing the low-frequency component of the image in each direction in the same way as the high-frequency component or medium-frequency component of the image, a large matrix size filter is not required, and the increase in hardware scale is suppressed. For medium frequency components, noise and line drawing jaggies are suppressed by performing spatial frequency correction for each direction. And, it is possible to obtain a good image of the resolution and sharpness.
[0096]
According to the present embodiment, enhancement of the high-frequency component of the image, enhancement of the medium-frequency component of the image, and enhancement of the low-frequency component of the image are performed in parallel, so that the processing speed of the filtering process is improved. be able to.
[0097]
As in the present embodiment, by performing spatial frequency correction for each direction with respect to the high frequency component and medium frequency component of the image, only the necessary direction is enhanced while suppressing noise enhancement, and an image with good resolution is obtained. Obtainable.
[0098]
Further, according to the present embodiment, the low frequency component of the image is emphasized by the single low frequency enhancement filter 91, so that the low frequency component of the image is emphasized independently for each direction. The amount of calculation can be reduced.
[0099]
In particular, as in the present embodiment, when the frequency characteristic of the low frequency emphasis filter 91 is set to a considerably low frequency side, the filter size tends to be remarkably large. By doing so, it is possible to increase the effect of suppressing the calculation amount without causing noise and jaggy in terms of image quality.
[0100]
Furthermore, according to the present embodiment, since the high frequency component and medium frequency component of the image emphasized for each direction are emphasized according to the edge amount that is the feature amount for each direction of the image, the edges are blunted. By doing so, an image with good resolution and sharpness can be obtained without impairing the characteristics of the image.
[0101]
In addition, since the low frequency component of the image is emphasized according to the maximum value of the edge amount that is the feature amount for each direction of the image, the edge frequency can be corrected even when the low frequency component of the image is corrected. The increase in the hardware scale can be effectively suppressed without degrading the image quality and damaging the characteristics of the image.
[0102]
Note that the low-frequency edge enhancement filter is not necessarily main-sub-equal as shown in FIG. 4, and may be a filter 91 'such as main-scanning 7 pixels and sub-scanning 5 pixels as shown in FIG. As shown in FIG. 23, by reducing the size in the sub-scanning direction, it is possible to reduce the number of line memories required for filtering processing by reducing the number of product-sum operations, thereby reducing the scale and cost of the arithmetic circuit. Reduction can be achieved more effectively.
[0103]
In the first and second embodiments, edge enhancement and smoothing of each frequency component of an image are performed in parallel. However, smoothing is not necessarily performed in parallel with edge enhancement. It is also possible to carry out before or after.
[0104]
Further, as shown in FIG. 24, an image output device 98 such as a printer for forming an image transmitted from the CPU is connected to the bus 3 of the PC 1 of the first and second embodiments, and the above filtering process is performed. The image whose image quality has been corrected may be printed out on a recording medium such as paper.
[0105]
Although illustration and description are omitted because it is a known technique, a printer used as the image output device 98 includes, for example, a printer engine that forms an image by an electrophotographic method or an inkjet method, and after image quality correction transmitted from the CPU 2 Operates based on images.
[0106]
In addition, in the first and second embodiments, the application example to the PC 1 is shown as the image processing apparatus. However, the present invention is not limited to this. For example, a digital copying machine (not shown) provided with a PC, a scanner, and a printer. Etc. may be used as an image processing apparatus, and an image based on an original image read by a scanner and corrected in image quality by a PC may be formed on a recording medium such as paper by a printer. In addition, for example, an MFP (multi-function printer) in which a digital copying machine including a PC, a scanner, and a printer is added with a communication function for transmitting / receiving data to / from another communication device installed remotely via a telephone line or the like. Function Printer) may be an image processing apparatus. Since the digital copying machine and the MFP are well-known techniques, illustration and description thereof are omitted.
[0107]
【The invention's effect】
  According to the image processing apparatus of the first aspect of the present invention, the spatial frequency correction for each direction is combined with the spatial frequency correction independent of directionality, and the spatial frequency correction is performed for each direction according to the spatial frequency characteristics of the input image. Since spatial frequency correction independent of directionality can be performed, the increase in hardware scale is suppressed by appropriately setting the spatial frequency for performing spatial frequency correction for each direction, and resolution and sharpness are improved. A good image can be obtained.In general, when spatial frequency correction is performed for each direction, a filter with a large matrix size is required. By performing spatial frequency correction of low frequency components of the input image with a single filter, an increase in hardware scale is effective. Can be suppressed.
[0108]
According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the spatial frequency correction by the first spatial frequency correction unit and the spatial frequency correction by the second spatial frequency correction unit are performed in parallel. By doing so, the processing speed can be improved.
[0110]
  Claim 3According to the described invention,Claim 1 or 2In the described image processing apparatus, the first corrected image obtained by correcting the spatial frequency for each direction by the first spatial frequency correcting unit is subjected to spatial frequency correction according to the feature amount acquired along the directionality of the directional filter. Thus, an image with good resolution and sharpness can be obtained without impairing the characteristics of the input image.
[0111]
  Claim 4According to the described invention,Claim 3In the described image processing apparatus, the second corrected image subjected to the spatial frequency correction by the second spatial frequency correcting unit is subjected to the spatial frequency correction according to the maximum value of the feature amount acquired along the directionality of the directional filter. Therefore, for example, even when the low frequency component of the input image is spatially corrected by the second spatial frequency correction unit, the resolution can be achieved without impairing the characteristics of the input image while effectively suppressing an increase in hardware scale. And an image with good sharpness can be obtained.
[0112]
  Claim 5According to the described invention,Claim 1 or 2In the described image processing apparatus, the second corrected image subjected to the spatial frequency correction by the second spatial frequency correcting unit is subjected to the spatial frequency correction according to the maximum value of the feature amount acquired along the directionality of the directional filter. Thus, for example, even when the low-frequency component of the input image is subjected to spatial frequency correction by the second spatial frequency correction unit, an increase in hardware scale can be effectively suppressed without impairing the characteristics of the input image.
[0113]
  Claim 6According to the described invention,Any one of claims 3 to 5In the described image processing apparatus, by acquiring the edge amount of the input image acquired by the differential filter as the feature amount, an image with good image quality can be obtained without dulling the edge of the input image.
[0114]
  Claim 7According to the image processing method of the described invention, the spatial frequency correction for each direction is combined with the spatial frequency correction independent of directionality, and the spatial frequency correction is performed for each direction according to the spatial frequency characteristics of the input image. Therefore, it is possible to perform spatial frequency correction that does not depend on the characteristics, and by appropriately setting the spatial frequency for performing spatial frequency correction for each direction, it is possible to suppress an increase in hardware scale and to achieve an image with good resolution and sharpness. Can be obtained.In general, when spatial frequency correction is performed for each direction, a filter with a large matrix size is required. By performing spatial frequency correction of low frequency components of the input image with a single filter, an increase in hardware scale is effective. Can be suppressed.
[0115]
  Claim 8According to the image processing program of the described invention,Claim 7The effects of the described invention can be obtained.
[0116]
  Claim 9According to the storage medium of the described invention,Claim 8The effects of the described invention can be obtained.
[Brief description of the drawings]
FIG. 1 is a block diagram schematically showing a hardware configuration of a PC according to a first embodiment of this invention.
FIG. 2 is a functional block diagram showing an image quality correction function realized by executing an image quality correction program by a PC.
FIG. 3 is a functional block diagram showing various functions realized by filtering processing performed by filtering processing means.
FIG. 4 is an explanatory diagram showing a filter matrix of a smoothing filter.
FIG. 5 is an explanatory diagram showing a filter matrix of a high frequency enhancement filter.
6 shows frequency characteristics in a 600 dpi image by a vertical high-frequency enhancement filter having the filter matrix shown in FIG.
FIG. 7 is an explanatory diagram showing a filter matrix of an edge amount calculation filter.
FIG. 8 is a graph showing a relationship between an edge amount and an enhancement coefficient.
FIG. 9 is an explanatory diagram showing a filter matrix of a low frequency enhancement filter.
FIG. 10 is a graph showing the frequency characteristics (main scanning direction).
FIG. 11 is a functional block diagram showing various functions realized by filtering processing performed by the filtering processing means of the PC according to the second embodiment of the present invention.
FIG. 12 is a functional block diagram illustrating functions of a high frequency emphasizing unit.
FIG. 13 is an explanatory diagram showing a filter matrix of a high frequency enhancement filter.
FIG. 14 is an explanatory diagram showing a filter matrix of an edge amount calculation filter.
FIG. 15 is a functional block diagram illustrating functions of an intermediate frequency emphasizing unit.
FIG. 16 is an explanatory diagram showing a filter matrix of a medium frequency enhancement filter.
FIG. 17 is an explanatory diagram showing a filter matrix of a medium frequency edge amount calculation filter;
FIG. 18 shows frequency characteristics of the vertical direction medium frequency emphasis filter shown in FIG.
FIG. 19 is a functional block diagram illustrating functions of a low frequency emphasizing unit.
FIG. 20 is an explanatory diagram showing a filter matrix of a low frequency enhancement filter.
FIG. 21 is an explanatory diagram showing a filter matrix of a low frequency edge amount calculation filter;
FIG. 22 shows frequency characteristics of the low frequency enhancement filter.
FIG. 23 is an explanatory diagram showing a filter matrix of another low-frequency edge enhancement filter.
FIG. 24 is a block diagram schematically showing a hardware configuration of a PC according to another embodiment;
[Explanation of symbols]
1 Image processing device
6 storage media
20, 21, 22, 23 Directional filter
39 Single filter
60, 61, 62 Directional filter
77, 78, 79 Directional filter
91 single filter

Claims (9)

An image input means for acquiring an input image;
A plurality of directional filters, each having a different directionality, are used to perform spatial frequency correction on an arbitrary spatial frequency component of the input image independently for each direction determined by the directionality of the directional filter. First spatial frequency correction means for obtaining a corrected image of
A second correction is performed by performing spatial frequency correction on the spatial frequency component of the input image using a single filter whose amplitude peak position is on the lower frequency side than the frequency characteristic of the directional filter without depending on directionality. Second spatial frequency correction means for obtaining an image;
Image synthesizing means for synthesizing the first corrected image and the second corrected image;
An image processing apparatus comprising:   The image processing apparatus according to claim 1, wherein the spatial frequency correction by the first spatial frequency correction unit and the spatial frequency correction by the second spatial frequency correction unit are performed in parallel. Feature quantity acquisition means for acquiring the feature quantity of the input image along the directionality of the directionality filter;
Third spatial frequency correction means for performing spatial frequency correction on the first corrected image according to the feature amount;
The image processing apparatus according to claim 1, further comprising: Maximum value acquisition means for acquiring a maximum value of the feature amount for each direction in which the feature amount is acquired;
Fourth spatial frequency correction means for performing spatial frequency correction on the second corrected image according to the maximum value;
The image processing apparatus according to claim 3, further comprising: Feature quantity acquisition means for acquiring the feature quantity of the input image along the directionality of the directionality filter;
Maximum value acquisition means for acquiring a maximum value of the feature amount for each direction in which the feature amount is acquired;
Fourth spatial frequency correction means for performing spatial frequency correction on the second corrected image according to the maximum value;
The image processing apparatus according to claim 1, further comprising:   The image processing apparatus according to claim 3, wherein the feature amount acquisition unit acquires an edge amount of the input image as the feature amount by a differential filter. Get the input image,
A plurality of directional filters, each having a different directionality, are used to perform spatial frequency correction on an arbitrary spatial frequency component of the input image independently for each direction determined by the directionality of the directional filter. Get the corrected image of
A second correction is performed by performing spatial frequency correction on the spatial frequency component of the input image using a single filter whose amplitude peak position is on the lower frequency side than the frequency characteristic of the directional filter without depending on directionality. Get an image,
An image processing method comprising combining the first corrected image and the second corrected image.   An image processing program for causing a computer to execute the image processing method according to claim 7.   A computer-readable storage medium storing the image processing program according to claim 8.

Images