JPH1125254A - Methods for picture analysis and picture edition - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for analyzing a picture and the characteristic of its pattern, and a picture editing method for re-constituting a picture by operating its characteristics. SOLUTION: A two-dimensional Fourier transformation of an inputted original picture is operated for generating two-dimensional power spectrum data 41, and the data 41 are displayed as a map picture 142 of a two-dimensional plane for analyzing the characteristics of a picture. The obtained power spectrum is modulated so that a picture edition processing while leaving a basic spectrum pattern can be executed, and a picture giving impression similar to a pattern of the natural world can be easily generated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image analysis method and an image editing method, and more particularly to a design of various articles having patterns formed on the surface, a design of a television broadcast, a video image, or a background image used for a computer screen. Available to

[0002]

2. Description of the Related Art Conventionally, image processing has been performed in various places. For example, an article is given a design that enhances its merchantability, but the shape of the surface as well as the shape is important. As such patterns, artificial geometric patterns, pictorial patterns, characters, photographs, and the like are used in various combinations. In recent years, there is a preference for emotional stability, and patterns imitating the natural world have been used. For example, a natural stone pattern print is made on a kitchen top plate, and a forest-like image is drawn on the surface of a decorative plate attached to the inner wall surface of a room. Further, a random pattern based on analysis of a pattern in the natural world is also used.
The use of the / f fluctuation pattern has also been proposed. It is known that these natural patterns or figures based on their analysis have high comfort such as stability and intimacy given to humans, and can enhance the merchantability of articles provided with such patterns.

[0003]

By the way, in order to obtain a figure which gives human comfort based on the analysis of the natural world pattern as described above, it is necessary to analyze the natural world pattern from various viewpoints. Become. In other words, if it is possible to grasp what characteristics and regularities that are latent in natural patterns give human comfort, patterns synthesized based on this regularity can also give human comfort. Can be expected. However, even when analyzing patterns in the natural world, patterns in the natural world have extremely diverse characteristics,
However, it is not easy to extract the characteristics of the pattern,
I want that means. There has been a demand for a means that can obtain not a pattern in the natural world itself but another pattern utilizing characteristics of the pattern. Conventionally, various types of image processing software have been developed. However, although various kinds of effects such as blurring are provided, it is not sufficient to cope with editing and the like utilizing characteristics of an original image.

An object of the present invention is to provide an image analysis method capable of analyzing an image and analyzing characteristics of the pattern, and an image editing method capable of reconstructing an image by operating the characteristics. .

[0005]

An image analysis method according to the present invention comprises: an input process for reading an original image converted into electronic data;
Analysis processing for generating the two-dimensional power spectrum data by performing two-dimensional Fourier transformation on the input original image and analysis result display processing for displaying the generated two-dimensional power spectrum data as a two-dimensional plane map image are sequentially performed. It is characterized by performing.

[0006] As the original image converted into electronic data, a photograph or a printed matter is read and converted into data by an existing scanner or the like, a landscape or an article is photographed with an existing digital camera, or commercially available or distributed free of charge. Image files and the like can be used as appropriate. The processing of the two-dimensional Fourier transform is
What is necessary is just to implement | achieve by means, such as executing on a computer, using an existing algorithm. The display of the map image on the two-dimensional plane may be performed by using a display processing function set in an existing computer, displaying the screen on a computer display device, or printing out by a printing device.
These input processing, analysis processing, and analysis result display processing may be realized by recording the program as a computer program on a medium such as a magnetic disk using an existing programming tool, and reading and executing the program as appropriate in a computer device. . Of course, it may be realized by dedicated hardware.

In this way, the pattern of the original image is analyzed as a repetition of a minute pattern by the two-dimensional Fourier transform, and can be extracted as a two-dimensional frequency power spectrum. The power spectrum obtained in this way is
It shows what appearance pattern distribution is due to the repetition of minute patterns in the pattern of the original image, and those showing similar spectral patterns have a similar image impression. Therefore, by using the analysis result according to the present invention, it is possible to identify an image that is not a natural pattern itself but gives a similar impression.

In the image analysis method according to the present invention, the original image is a color image, and in the input processing, the original image is subjected to primary color separation to generate color image data for each primary color, and together with the original image, In addition to performing a process of inspecting the number of colors and a process of generating monochrome image data by converting the original image to monochrome, the analysis process is performed on the color image data of each primary color and the monochrome image data, respectively. It is desirable to generate two-dimensional power spectrum data for each primary color and monochrome. In this way, the color original image can be efficiently processed, and the overall brightness can be handled by the monochrome image data in addition to the processing for each primary color data.

In the image analysis method of the present invention, each of the primary colors is red, green, and blue, and the analysis result display processing is performed by using four of the three primary color two-dimensional power spectrum data and the monochrome two-dimensional power spectrum data. It is preferable that the origins are displayed in a frame of two rows and two columns where the origins are gathered at the center of each other. In this way, color correspondence can be easily realized by using image information processing in red, green, and blue (RGB) performed by a general computer, and the two-dimensional power spectrum data 41 can be reduced in size and visibility. Can be displayed. In the present invention, the primary color separation may be performed not only by the RGB method but also by the CYM method (sky blue; cyan, yellow; yellow, pink; magenta) frequently used in the printing field.

In the image analysis method of the present invention, at the time of the analysis result display processing, one-dimensional power spectrum data is generated from the two-dimensional power spectrum data. It is desirable to display a power spectrum graph in which is represented by the frequency. In this way, it is easier to grasp the relationship between the frequency and the power spectrum than to the two-dimensional map image, and the outline of the characteristics of the original image can be easily identified. However, the details are more accurate with a two-dimensional map image. In addition, in the case where color separation is performed for primary colors and monochrome, it is desirable to display the power spectrum curves of each primary color and monochrome in the one-dimensional power spectrum data.

An image editing method according to the present invention is an image editing method for generating a new image by referring to a map image generated by the above-described image analysis method, wherein the input processing, the analysis processing, and the analysis result display processing are performed. Subsequently, two-dimensional power spectrum modulation processing for modulating the two-dimensional power spectrum data and image restoration processing for restoring a new image from the modulated two-dimensional power spectrum data are performed.

The two-dimensional power spectrum modulation processing includes tilt modulation, division modulation, map modulation, and the like, which will be described later. In the image restoration process, an inverse transform of the two-dimensional Fourier transform performed in the above-described analysis process may be used.
Each of these processes may be appropriately implemented using existing computer technology, as in the case of the image analysis method described above.

In this way, the pattern of the original image is analyzed as a superposition of various linear stripe patterns by the two-dimensional Fourier transform, and can be extracted as a two-dimensional frequency power spectrum. By performing modulation on the power spectrum obtained in this way, it is possible to perform processing such as editing an image while leaving a basic pattern. Therefore, by performing the editing according to the present invention, it is possible to easily create an image or the like that is not a natural pattern itself but gives a similar impression.

In the image editing method according to the present invention, it is preferable that the two-dimensional power spectrum modulation process includes a gradient modulation process of adding or subtracting a bias value linearly inclined in a frequency space with respect to the two-dimensional power spectrum. In this way, since the operation is performed linearly from the low band to the high band of the frequency space, the pattern and color of the restored image can be changed while utilizing the features of the original image.

In the image editing method of the present invention, it is preferable that the two-dimensional power spectrum modulation processing includes a division modulation processing for increasing or decreasing a power level of each band in each of a plurality of divided frequency spaces. If you do this,
An operation can be performed for a specific frequency band, and the details of the pattern can be limitedly operated.

In the image editing method according to the present invention, it is preferable that the two-dimensional power spectrum is generated for each primary color and monochrome, and the division modulation processing is performed for each primary color and monochrome two-dimensional power spectrum.
In this way, the division modulation can be performed for each primary color and the entire light and dark (monochrome), and more detailed operations can be performed.

In the image editing method of the present invention, it is preferable that the two-dimensional power spectrum modulation process includes a frequency map operation process for operating the two-dimensional power spectrum by directly editing the map image. For direct editing of the map image, a means can be used in which a pointer on the screen is set at an arbitrary position with an existing so-called mouse or the like, and an adjustment operation is performed with a button or the like. The adjustable action area can be a point or an area having a predetermined spread, and can be realized by applying a technique used for a so-called brush effect of an existing paint tool. In this way, more detailed editing can be performed on the two-dimensional map image, extreme changes to the map image are possible, and special effects can be obtained.

In the image editing method of the present invention, the power level at the position is increased or decreased by designating an arbitrary position in the map image at the time of the map operation processing, and an action area is set around the designated position. It is desirable to perform a frequency map modulation parameter setting process. In this way, the operation of the above-described range of the effect of the brush effect can be performed, and the direct operation of the frequency map can be performed in more detail.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. In the present embodiment, the method of the present invention is executed by each functional element realized by a software program executed on an existing personal computer, and the computer device executing the program executes the method of the present invention. It functions as an image processing device. In FIG. 1, an image processing device 10 includes an input processing unit 20,
An analysis processing unit 30, an analysis result display unit 40, a two-dimensional power spectrum modulation unit 50, and an image restoration unit 60 are provided.

(Input processing unit 20) The input processing unit 20 is for reading an original image file 11 in which an original image converted into electronic data is recorded, and includes a file input unit 21, a color number inspection unit 22, a monochrome A conversion unit 23 and an original image display unit 24 are provided. The file input unit 21 takes in an external file into the computer device, and appropriately uses an existing configuration set in the computer, such as a mechanical device that inputs a file from an external storage device of the computer device and a control program thereof. Is what you do. The input data of the original image file 11 is held in the main storage area or the like as color image data 25 of three primary colors of RGB.

The program according to the present embodiment operates on an existing window system, and some of the functional elements such as the file input unit 21 are activated by a menu operation as shown in FIG. In FIG. 2, when the file is not opened, only the “file” menu 110 is displayed on the menu bar 101 of the program window 100. By selecting the "Open" menu 111 from the "File" menu 110 and opening a file selection dialog (not shown) common to the system, the file input unit 21 is opened.
Is started. Note that if the “end” menu 112 is selected in the “file” menu 110, the program ends.

The color number inspection unit 22 inspects the number of colors based on the data from the file input unit 21. Monochrome converter 23
Generates monochrome image data 26 by processing the data from the file input unit 21. For monochrome conversion, RG
A method of converting an image signal of the B system into an image signal of the US NTSC standard and extracting a Y signal from a YIQ display system (Y: lightness, I: saturation, Q: color tone) can be used. Specifically, Y = 0.228R + 0.587G + 0.114B.

The original image display unit 24 displays the input color image data 25 of the three primary colors of RGB on a display of a computer device or the like, and appropriately uses an existing configuration such as a display program of the computer device. is there. In FIG. 3, a window 113 named “sample file” is displayed in the program window 110, and a display image 114 is displayed therein. The display image 114 is displayed by the original image display unit 24 and corresponds to the color image data 25.

(Analysis Processing Unit 30) The analysis processing unit 30 generates two-dimensional power spectrum data by performing two-dimensional Fourier transform on the color image data 25 of the three primary colors RGB generated from the input original image. Two-dimensional Fourier transform unit
31, a direction designation input unit 32 is provided. In FIG. 3, when the display image 114 is displayed, the program window 1 is displayed.
At 00, a menu bar 102 is displayed, and the analysis processing unit 30
It is activated by selecting an “FFT” menu 121 from a “processing” menu 120 on the menu bar 102.
Note that the “Direction” menu 1
By selecting 22, the direction designation input unit 32 is activated.

The two-dimensional Fourier transform unit 31 uses the color image data 25 as a two-dimensional signal, approximates the two-dimensional signal by superposition of linear waves, and distributes the intensity of each of the two-dimensional spatial frequencies (X and Y components) ( Examine the power spectrum). At this time, the two-dimensional Fourier transform can be processed by an existing algorithm. Since the target RGB color image data 25 of the three primary colors is discrete data for each pixel, the conversion process uses DFT (Discrete Fourier Transform). More specifically, this can be realized by performing one-dimensional DFT in each of the vertical and horizontal directions.
FT (fast Fourier transform) may be used.

The direction designation input unit 32 displays a direction designation dialog 130 (see FIG. 4) on the display. The direction designation dialog 130 receives an external input from a keyboard, a mouse, or the like, and controls the operation of the two-dimensional Fourier transform unit 31. It is specified. As shown in FIG. 4, in the direction designation dialog 130, “all directions” 131, “projection” 132, and “horizontal one-dimensional” 133 can be selected and designated for the direction in which the Fourier transform is performed.

When "all directions" 131 is selected, the two-dimensional Fourier transform unit 31 performs the two-dimensional Fourier transform described above.
If "Projection" 132 is selected, enter the "Direction" numerical value 1
34 becomes possible, and based on this setting, only the component in the designated direction is extracted from the result of the two-dimensional Fourier transform. When “horizontal one-dimensional” 133 is selected, one-dimensional Fourier transform in the X direction is sequentially performed for each Y-axis position of the image. These processes are performed for each of the three primary color image data 25 and the monochrome image data.
26, two-dimensional power spectrum data 41 of three primary colors (RGB) and monochrome (M) is generated and stored in the analysis result display unit 40.

(Analysis result display section 40) Analysis result display section 40
Is activated when the processing of the two-dimensional Fourier transform unit 31 is completed, and holds two-dimensional power spectrum data 41 of three primary colors (RGB) and monochrome (M) in an area secured on the main memory. Power spectrum data
41 is displayed as a two-dimensional plane map image 142 and a one-dimensional power spectrum graph 143. For this purpose, the analysis result display unit 40 includes a two-dimensional power spectrum map display unit 42 and a one-dimensional power spectrum graph display unit.
43 and.

A two-dimensional power spectrum map display section 42
Displays the four primary color (RGB) and monochrome (M) two-dimensional power spectrum data 41 as map images 142. As shown in FIG. 5, the program window
A window 141 named “sample file PwrSpecMap” is displayed in 110, and a map image 142 is displayed therein. The map image 142 is displayed in a two-row, two-column frame format in which the origins are gathered at the center of each other, a monochrome (M) region at the upper right, a red (R) region at the upper left, a green (G) region at the lower left, and a right The lower part is a blue (B) area. In each area, the horizontal axis represents the frequency of the X component, and the vertical axis represents the frequency of the Y component. Each displayed dot represents a power level in color (blue is low, becomes stronger from green to yellow, and becomes red, Is the strongest).

A one-dimensional power spectrum graph display section 43
The one-dimensional power spectrum data is generated by calculating the average in all directions from the two-dimensional power spectrum data 41 by the least squares method or the like, and the one-dimensional power spectrum data has the vertical axis as the level and the horizontal axis as the frequency. This is displayed as a power spectrum graph 143. As shown in FIG. 5, "Sample file ..." is displayed in the program window 110.
A window 144 (namely, “PwrSpec” is omitted in the display) is displayed, and a power spectrum graph 143 is displayed therein.

(Two-dimensional power spectrum modulator 50) The two-dimensional power spectrum modulator 50 modulates the two-dimensional power spectrum data 41.
A division modulation setting unit 52, a frequency map operation unit 53, a frequency map modulation parameter setting unit 54, and a modulation processing unit 55 are provided. In addition, as shown in FIG.
Is displayed, the menu bar 103 is displayed.The inclination modulation setting unit 51, the division modulation setting unit 52, the frequency map modulation parameter setting unit 54, and the modulation processing unit 55 are displayed from the `` modulation '' menu 150 of the menu bar 103. Each menu 151-154
It is activated by selecting. However,
The frequency map operation unit 53 is always valid when the map image 142 is displayed.

The tilt modulation setting section 51 is provided with a “modulation” menu 15
It is activated by selecting the `` tilt modulation '' menu 152 from 0, and accepts an external input by a keyboard, a mouse or the like by a tilt modulation dialog 150 displayed on the display,
This specifies the modulation processing in the modulation processing unit 55.
As shown in FIG. 6, the gradient modulation dialog 160 has a value 161 of “current gradient”, an input frame 162 of “new gradient”, and an input frame 163 of “center frequency”. By designating these “new slope” and “center frequency”, a sloped straight line passing through the center is defined, and designation is performed such that each data of the two-dimensional power spectrum data 41 is modulated according to the straight line. . In this processing, there is no change at the center frequency, but the frequencies before and after the center frequency are specified to increase or decrease (depending on the slope) as the distance from the center frequency increases.

The division modulation setting section 52 is provided with a “modulation” menu 15
It is activated by selecting the “split modulation” menu 153 from 0, and accepts an external input by a keyboard, a mouse, or the like by a split modulation dialog 170 displayed on the display,
This specifies the modulation processing in the modulation processing unit 55.
As shown in FIG. 7, the divided modulation dialog 170 includes a black-and-white (monochrome M) modulation slider group 171 and RGB-color modulation slider groups 172, 173 and 174. Each of the slider groups 171 to 174 has 11 modulation sliders corresponding to each frequency band of frequencies 1 to 1024, and by operating each with a mouse or the like, it is possible to specify power increase / decrease for each frequency band. ing. Each of the slider groups 171 to 174 is provided with a reset button 175 so that the slider once set for each of RGB colors and monochrome can be returned to the center with one touch. In the split modulation dialog 170, "half width" 176, "maximum magnification" 177,
A numerical box of “effect range” 178 is provided, and the amount of power operation for each band that can be increased or decreased with each slider, the influence range, and the like can be finely adjusted.

The frequency map operation section 53 is always effective when the above-described map image 142 is displayed. The position of the map image 142 is designated by using a mouse or the like, and an increase or decrease operation is performed. This is for directly operating the monochrome (M) two-dimensional power spectrum data 41. In FIG. 5, for example, by moving the mouse pointer 145 to a blue (B) area and operating the left button of the mouse, the power of the relevant portion can be increased, and by operating the right button, the power of the relevant portion can be reduced. . Such an operation may be appropriately configured using existing graphical user interface technology.

The frequency map modulation parameter setting unit 54
The “modulation” menu 150 is activated by selecting a “frequency map modulation parameter” menu 154 from the menu 150. When the frequency map operation unit 53 directly operates the two-dimensional power spectrum data 41 by, for example, operating the mouse button 1 click. For this purpose, an external input by a keyboard, a mouse, or the like is received by a frequency map modulation parameter setting dialog 180 displayed on the display. As shown in FIG. 8, the frequency map modulation parameter setting dialog 180 includes “half width” 181,
Numerical boxes for “maximum magnification” 182 and “influence range” 183 are provided, and the amount of power increase / decrease at a designated position which can be increased / decreased by one operation and the influence range can be finely adjusted.

The modulation processing section 55 is provided with the above-described gradient modulation setting section.
The two-dimensional power spectrum data 41 is read and frequency-modulated based on the modulation specification by the division modulation setting unit 52, frequency map operation unit 53, and frequency map modulation parameter setting unit 54. This processing can be appropriately configured using existing digital signal filtering technology. After rewriting the two-dimensional power spectrum data 41, the modulation processing unit 55 restarts the analysis result display unit 40 to display the map image 142 and the power spectrum graph 143 again. As shown in FIG. 9, when the modulation processing by the modulation processing unit 55 is completed, the map image 142 of the modulated two-dimensional power spectrum data 41 is displayed in a new window 146 by the two-dimensional power spectrum map display unit 42. The power spectrum graph 143 of the modulated two-dimensional power spectrum data 41 is displayed in a new window 147 by the one-dimensional power spectrum graph display unit 43.

(Image Restoring Unit 60) The image restoring unit 60 restores an image from two-dimensional power spectrum data, and includes a two-dimensional inverse Fourier transform unit 61, a restored image display unit 62, and a file output unit 63. I have.

The two-dimensional inverse Fourier transform unit 61 performs an operation reverse to that of the above-described two-dimensional Fourier transform unit 31, and converts the modulated two-dimensional power spectrum data 41 into RG data.
This is to restore the color image data 64 of the three primary colors B.
The two-dimensional inverse Fourier transform unit 61 calculates the map image 142
Is displayed on the menu bar 103 (see FIG. 9), an “image restoration” menu 155 to an “execute” menu 156 are displayed.
Is executed by selecting.

The restored image display section 62 displays the restored color image data 64 of the three primary colors RGB on a display of a computer device or the like, and appropriately uses an existing configuration such as a display program of the computer device. is there. In FIG. 10, a window 115 named “sample file” is displayed in the program window 110, and a display image 116 is displayed therein. The display image 116 is displayed by the restored image display unit 62 and corresponds to the color image data 64.

The file output unit 63 performs the reverse process of the file input unit 21 and uses an existing configuration for saving the color image data 64 as an external restored image file 12. The file output unit 63 is executed by selecting the “Save” menu 119 from the “File” menu 117 on the menu bar 103 (see FIG. 10) when the above-described map image 142 is displayed. By selecting the “Print” menu 118 from the “File” menu 117, the color image data is output to an external printer (not shown).
It is possible to output 64 and print it.

According to the present embodiment, the following effects can be obtained. The original image is read, and the input original image is subjected to two-dimensional Fourier transform to obtain two-dimensional power spectrum data 41.
And generate the generated two-dimensional power spectrum data 41
Is displayed as a two-dimensional plane map image 142, the pattern of the original image is analyzed as a repetition of minute patterns, and can be extracted as a two-dimensional frequency power spectrum. The power spectrum obtained in this way is
It shows what appearance pattern distribution is due to the repetition of minute patterns in the pattern of the original image, and those showing similar spectral patterns have a similar image impression. Therefore, the map images 142 of various images
By comparing patterns, it is possible to perform analysis such as identifying images that give similar impressions even at first glance, and identifying images that give different impressions even at first glance. Can be.

In the present embodiment, the input processing section 20 performs primary color separation of the original image to obtain color image data 25 for each primary color.
And the original image is converted to monochrome to generate monochrome image data 26, and the analysis processing unit 30 analyzes the color image data 25 and monochrome image data 26 of each primary color, and analyzes each primary color and monochrome. Since the two-dimensional power spectrum data 41 is generated, it is possible to efficiently process the color original image, and it is possible to cope with the entire brightness with monochrome image data in addition to the processing for each primary color data.

In particular, since the primary colors are red, green, and blue, color correspondence can be easily realized by using image information processing in red, green, and blue (RGB) performed by a general computer. Then, the analysis result display section 40
Displays four two-dimensional power spectrum data 41 of three primary colors (RGB) and monochrome (M) as a map image 142 having a two-row, two-column frame whose origins are gathered at the center of each other. 41 can be displayed compactly and with good visibility.

In this embodiment, the analysis result display section 40 generates one-dimensional power spectrum data from the two-dimensional power spectrum data 41, and converts the one-dimensional power spectrum data into a power having a vertical axis representing a level and a horizontal axis representing a frequency. Since it is displayed on the spectrum graph 143, it is possible to grasp the relationship between the frequency and the power spectrum in a two-dimensional map image 14.
It is easier than just 2 and the outline of the characteristics of the original image can be easily identified. In particular, the power spectrum graph 143
In, since a total of four power spectrum curves corresponding to the primary color and the monochrome which are color-separated, respectively, are displayed, it is possible to easily grasp the outline of each primary color and the monochrome.

In the present embodiment, two-dimensional power spectrum data 41 is extracted by input processing, analysis processing, and analysis result display processing, and then two-dimensional power spectrum modulation processing for modulating the two-dimensional power spectrum data 41 is performed. Further modulated two-dimensional power spectrum data 41
By performing image restoration processing to restore a new image from
Processing such as editing an image while leaving the basic spectral pattern can be performed. Therefore, by performing the editing according to the present invention, it is possible to create an image that gives a similar impression even if the images are seemingly different, or perform an editing such as creating an image that gives a different impression even if the images are seemingly similar. It can be carried out. As a result, for example, it is possible to easily create an image that is not a natural pattern itself but gives a similar impression.

In the present embodiment, the two-dimensional power spectrum modulation processing includes a slope modulation processing for adding / subtracting a bias value that is linearly sloped in the frequency space with respect to the two-dimensional power spectrum 41. Since the operation is linear over the range from to, the pattern and color of the restored image can be changed while utilizing the characteristics of the original image.

In the present embodiment, the two-dimensional power spectrum modulation processing includes division modulation processing for increasing or decreasing the power level of each band in each of a plurality of divided frequency spaces. , The details of the pattern can be limitedly manipulated. In particular, since the division modulation processing is performed for each primary color and monochrome two-dimensional power spectrum, the division modulation can be performed for each primary color and the entire light and dark (monochrome), and more detailed operations can be performed.

In the present embodiment, the two-dimensional power spectrum modulation processing includes a frequency map operation processing for operating the two-dimensional power spectrum data 41 by directly editing the map image 142. In addition to this, further detailed editing can be performed, extreme changes to the map image 142 are possible, and special effects can be obtained. In particular, the map image 14
Since the power level at the position is increased or decreased by designating an arbitrary position of 2, the operation can be intuitively and easily performed. In addition, since the frequency map modulation parameter setting processing for setting the action area around the designated position is performed, the above-described operation of the range of the effect of the brush effect can be performed, and the direct operation of the frequency map can be performed in more detail. it can.

It should be noted that the present invention is not limited to the above-described embodiment, and a specific system may be realized by appropriately referring to existing software technology. Modifications and the like shown in the section of "Means" are included in the present invention. further,
The above-described configuration uses a program executed by an existing computer system to execute a central processing unit (C) of the system.
PU) and an execution module developed on the main memory. At this time, the program is stored in a portable recording medium (a magnetic disk such as a so-called floppy disk or a removable hard disk, another magnetic storage medium such as a magnetic tape or a magnetic drum), an optical disk such as a CD-ROM, a magneto-optical disk, a ROM chip or an IC. A semiconductor storage medium such as a memory card) may be incorporated in the system. Then, in the system, the incorporated program is stored and held in an external storage device of the system, and is read into the main memory upon execution, and is executed by the CPU. As the external storage device, various fixed recording media (magnetic storage media such as so-called hard disks,
For example, a flash memory or a semiconductor memory element such as a battery-backed semiconductor RAM can be used.
If the program is installed at the time of manufacturing the device,
Read-only recording medium (semiconductor ROM, etc.) for the system
And a program may be recorded in advance so that it can be read out at the time of execution.

[0050]

According to the present invention, the pattern of the original image can be analyzed as a repetition of minute patterns by the two-dimensional Fourier transform, and can be extracted as a two-dimensional frequency power spectrum. By performing the comparison, it is possible to perform an analysis such as identifying an image that gives a similar impression even if the images are seemingly different, and an image that gives a different impression even if the images are seemingly similar. . Further, by performing modulation on the power spectrum obtained in this way, it is possible to perform processing such as editing an image while leaving a basic spectral pattern, and to give a similar impression although it is not a pattern in the natural world itself. Images and the like can be easily created.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a schematic diagram showing a file input operation of the embodiment.

FIG. 3 is a schematic diagram showing an analysis operation of the embodiment.

FIG. 4 is a schematic diagram showing a direction instruction of an analysis operation of the embodiment.

FIG. 5 is a schematic diagram showing a modulation operation of the embodiment.

FIG. 6 is a schematic diagram showing a tilt modulation setting of the embodiment.

FIG. 7 is a schematic diagram showing division modulation settings according to the embodiment.

FIG. 8 is a schematic diagram showing frequency map modulation parameter settings of the embodiment.

FIG. 9 is a schematic diagram showing an image restoration operation according to the embodiment.

FIG. 10 is a schematic diagram showing a save operation of the embodiment.

[Explanation of symbols]

 10 Image processing unit 20 Input processing unit 30 Analysis processing unit 40 Analysis result display unit 50 2D power spectrum modulation unit 60 Image restoration unit 11 Original image file 114 Display of original image 25 Color image data 26 Monochrome image data 41 2D power spectrum Data 142 Two-dimensional map image 143 One-dimensional power spectrum graph 64 Reconstructed color image data 116 Display of reconstructed image 12 Reconstructed image file

Claims (10)

[Claims] An input process for reading an original image converted into electronic data, an analysis process for performing two-dimensional Fourier transform on the input original image to generate two-dimensional power spectrum data, and a generated two-dimensional power spectrum And an analysis result display process for displaying data as a two-dimensional plane map image. 2. The image analysis method according to claim 1, wherein the original image is a color image, and in the input process, primary image color separation is performed on the original image to generate color image data for each primary color. In addition to performing a process of inspecting the number of colors of the original image and a process of generating the monochrome image data by converting the original image to monochrome, the analysis process includes the color image data of each primary color and the monochrome image data. A two-dimensional power spectrum data for each primary color and each monochrome. 3. The image analysis method according to claim 2, wherein each of the primary colors is red, green, and blue, and the analysis result display processing is performed on the two-dimensional power spectrum data of the three primary colors and the monochrome two-dimensional power spectrum. An image analysis method, wherein four data are displayed in a two-row, two-column frame whose origins are gathered at the center of each other. 4. The image analysis method according to claim 1, wherein one-dimensional power spectrum data is generated from the two-dimensional power spectrum data during the analysis result display processing. An image analysis method, wherein data is displayed as a power spectrum graph in which a vertical axis represents a level and a horizontal axis represents a frequency. 5. An image editing method for generating a new image while referring to a map image generated by the image analysis method according to claim 1, wherein the input processing and the analysis are performed. Following processing and analysis result display processing,
An image editing method, comprising: performing a two-dimensional power spectrum modulation process for modulating the two-dimensional power spectrum data; and an image restoration process for restoring a new image from the modulated two-dimensional power spectrum data. 6. The image editing method according to claim 5, wherein the two-dimensional power spectrum modulation processing includes a gradient modulation processing for adding or subtracting a bias value linearly inclined in a frequency space to the two-dimensional power spectrum. An image editing method, characterized in that: 7. The image editing method according to claim 5, wherein the two-dimensional power spectrum modulation process includes a division modulation process for increasing or decreasing a power level of each band in each of a plurality of divided frequency spaces. An image editing method characterized by including: 8. The image editing method according to claim 7, wherein the two-dimensional power spectrum is generated for each primary color and monochrome, and the division modulation processing is performed for each primary color and monochrome two-dimensional power spectrum. An image editing method characterized by being performed. 9. The image editing method according to claim 5, wherein the two-dimensional power spectrum modulation process manipulates the two-dimensional power spectrum by directly editing the map image. An image editing method comprising a frequency map operation process. 10. The image editing method according to claim 9, wherein the power level of the position is increased or decreased by specifying an arbitrary position of the map image at the time of the map operation processing, and the vicinity of the specified position is increased. A frequency map modulation parameter setting process for setting an action area in the image editing method.