JPH0695339B2 - Image processing method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an image processing method for converting an image.

[Conventional example]

Conventionally, as one of the image processing methods of this kind, an electronic plate making technique in the printing industry and a professional laboratory technique applying computer photographic image processing in the photographic laboratory industry are known.

For example, when a reproduced image is obtained by photoelectrically scanning an image original with a high-precision scanner such as a layout scanner or a laser color printer, a signal processing unit is provided in the middle of the process to correct the density expression (γ correction) for the input density signal. ), Gradation setting,
The following effects are obtained by performing processing such as color correction and cutout composition.

Color film fading restoration Adjusting the gradation of high night and shade, color expression exaggeration Relief of device failure, shooting mistake, development mistake, etc. Expressing creative images, expanding the image area and creating new designs.

When such special processing is performed, the density signal or brightness of the film original obtained by photoelectrically scanning the filter original with a color scanner, a high-precision scanner such as a laser color printer or a color image pickup tube, a color image pickup plate (for example, CCD), etc. It is converting a digital image consisting of signals. In particular, here, the texture screen processing for producing various texture patterns on the original image will be described.

Textured screen The term textured screen is a term used in the photographic industry and refers to a screen used to create various textured patterns on an image on photographic paper when printed. It is also used when you want to hide the particles as well as when you want to find the texture in the image. For this, there is also a method of using photographic paper that already has a pattern such as silk grain to make the particles less noticeable by the pattern, but the method using a texture screen when printing is more preferable. It is effective. However, since the photographic processing using this texture screen is generally performed as a darkroom operation, there are many technically difficult aspects of the apparatus and it has only been performed by some enthusiasts and professionals. The texture screen processing is to carry out the same effect as that using this texture screen by using the digital image processing technique.

Conventionally, this type of special effect processing using digital image processing includes an electronic plate making system (layout scanner) used in the printing industry. However, in order to perform the above texture screen processing with this apparatus, the following synthesis processing was performed.

The base film image from the film is read for each R, G, B color by a high-precision scanner, and after A / D conversion is stored in the memory as 8-bit film density data. Similarly, the texture image which becomes the texture is read by the high-precision scanner and the data is stored. In the case of texture image data, it may be stored as a monochrome image instead of a color image. Here, the film density data of the original image and the texture image stored in the memory are respectively ai (m, n),
Expressed as b (m, n). However, i represents the data of each color of R, G, B, (m, n) is the address of the image data, and the texture image is a monochrome image.

And then The synthetic image is obtained by the arithmetic processing. When such processing is performed, the original image component is As a result, an image with a low contrast and a flat feeling is produced. Further, in determining the values of α and β, the established values must be changed according to the state of the difference in shade of the texture image. This often depends on the experience and intuition of the creator, and because trial and error are repeated, not only is it time-consuming, but it is limited to a specific creator, and no one can do it easily. On the other hand, in order to perform high-speed calculations in digital image processing that handles a large amount of two-dimensional data such as film images, for example, NEXUS 6400 (Kashiwagi Laboratory), Tospix II
(Toshiba) often uses a dedicated image processing device (Image Processor). Unlike a general-purpose computer, such an image-only processing device does not have a floating-point arithmetic means.
To handle the screen as an 8-bit image memory In order to calculate
This must be done in a 16-bit configuration, and if one screen of image memory for result output is added, an image memory for at least 3 images is required.

As described above, in the processing using the conventional synthesis, the reproduced image becomes a flat tone. In addition, it takes time to repeat trial and error based on the creator's experience and intuition based on the ratio (α, β) of the original image and the texture image, and the usage is limited to a specific creator, so that it can be used by general consumers. Is difficult Further, when the image-only processing device is used, the image memory for two screens is required as the image memory for processing.

〔Purpose〕

SUMMARY OF THE INVENTION It is an object of the present invention to provide an image processing method which eliminates the above-mentioned drawbacks of the prior art, enables automatic composite image processing, and can obtain a composite image having a stereoscopic effect.

〔Example〕

An embodiment of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing an example of the flow of a density signal when the present invention is applied to a color scanner. An input signal obtained by photoelectrically scanning a film original loaded on the input drum 1 is converted into a density signal by the logarithmic conversion circuit 2 and then input to the input signal processing unit 4 via the AD converter 3. , The density value of each film is converted according to the type of the film, and the density value is converted so as to have a fixed relationship with the exposure amount of the film. This is because the film characteristic curve differs depending on whether the film is a negative film, a positive film or a reversal film, and in the case of a color image, R, G,
This is because each color of B is different. After that, the image is input to the image processing unit 5, which should be the heart of the system, and texture screen processing is performed. After that, the density signal output from the image processing unit 5 is processed by the output signal processing unit 6 to be converted into a laser light amount control signal, and is input to the converter 8 via the DA converter 7 and output from the light source 9. The laser light is modulated to cause the output drum 10 to reproduce a desired image.

FIG. 2 is a block diagram showing in more detail a part of the image processing section 5. The CPU 11 is a control processor, which holds all the real rights of the image processing unit 5, and controls the image processing based on a control program stored in the CPU memory 12. The parameter controller 13 controls the arithmetic unit 14, the parameter memory 15, and the parameter setting I / O 16 to perform initialization, setting, comparison, etc. of parameters necessary for processing. The processor 18 and the CPU 11 are connected by the image controller 17 via the CPU BUS.
The processor 18 operates according to the command of. The processor 18 is a core part of the image processing unit 5, and is any image memory 20 to 26 selected according to the instruction of the CPU 11.
And image data from I / O34 for image data
Arbitrary image memory with the calculated result selected
Send to 20-26 or I / O34 for image data. In addition, the data can be output to the flag map memory 19, which is a memory dedicated to the carrier, in order to carry the carrier to a necessary address by calculation. Since the image memories 20 to 26 are connected to any of the CPU BUS.VIDEO BUS, it is possible to read / write from the CPU 11 to the arbitrary image memories 20 to 26 and perform real-time calculation between the arbitrary memories. A high-speed RAM called Lucupup table 27-33 is connected to the exit of the image memory. This RAM has a structure of 256 x 8 bits, and 8 address lines of RAM (address 0 to 25 can be specified = 0 to 255 gradations) are directly connected to the output of each image memory, and 8 data lines. Is connected to the VIDEO BUS. The contents of RAM can be freely read and written by the CPU 11. The image data I / O 34 is an image input / output interface, which inputs image data from the input signal processing unit 4 in FIG. 1 and outputs image data to the output signal processing unit 6.

FIG. 3 shows a flow chart of the texture screen processing method performed in the image processing section 5, and FIG. 4 is a graph showing the states of the lookup tables 24 to 27, in which the horizontal axis is the input gradation and the vertical axis is the vertical axis. Indicates the output gradation. FIG. 4 (a) shows a standard state, in which 0 is written in address 0, 1 is written in address 1, 255 is written in address 255 ... Will be the same. Also, although not shown, 255 is at 0, 254 is at 1, ...
When a value of 0 is written in address 255, the input-output indicates that the density data is an inverted image.

The texture screen processing method performed by the image processing section 5 will be described in detail below with reference to FIG. 3 of the flow chart. Here, the density data of the digital image is ai (m,
n) will be used for explanation purposes. However, i is R, G, B, and T, and indicates R component image data, G component image data, B component image data, and texture image data of the original image, respectively. In addition, one pixel consists of 8 bits for each component.
The data having the highest density (darkest data), which can represent 256 gradations, has a value of 0, and the data having the lowest density (the lightest data) has a value of 255.

The flow chart of FIG. 3 will be described below.

Description of Step 1 S1 The processor 18 stores the R component image data of the original image in the image memory (1) from the input signal processing unit 4 via the image data I / O 34 according to the instruction of the CPU 11. Then G of the original image
The component image data, the B component image data, and the texture image data captured as a monochrome image are stored in the image memories (2), (3), (4) 21 to 23, respectively. At this time, the rack up tables (1) to (4) 27 to 30 are in the standard state of FIG. 4 (a). That is, the data is stored as it is in the image memory.

Explanation of Step 2 S2 The lower limit threshold Dmin of the image memory (4) is obtained using the area calculation counter of the processor 18. The lower limit threshold Dmin is a value in which the cumulative number of pixels from the image data value 0 to Dmin is equal to the number of pixels S ₀ in the vicinity of 5% of the total number of pixels. For example 512 x 512
If it is pixel data, the number of pixels from 0 to that value is 512
X512 x 0.05 = around 13107.

Parameter controller 13 and calculator 1 by CPU11
The value S ₀ , which is 5% of the total number of pixels of the image, is calculated by 4 and stored in the parameter memory 15, and the value of Dmin is also set to 0. Further, the flag map memory 19 is reset by the image controller 17 and the processor 18 and all values are set to 0.
Next, the processor 18 performs an operation to give 1 to the pixel of the density data Dmin (0 in this case) of the image memory (4) 23 in the flag map memory 19. Then, the number of flags (where 1 stands) is found in the flag map memory 19 by the area calculation counter, and the result Sx is sent to the calculator 14 via the parameter controller. Then, the result Sx is compared with the set value S _0, and if Sx ≧ S ₀ , this step is ended, and if Sv <So, 1 is added to Dmin and the processor 18 adds the image memory to the flag map memory 19 (4). The pixel of 23 density data Dmin (1 at this time) is calculated by giving 1 and Sx is calculated again by the area counter. Then, it is sent again to the arithmetic unit 14 to compare Sx and So. By repeating such calculation, from the density data 0 of the image memory (4) 23
A Dmin value is calculated when the number of Dmin pixels exceeds 5% of the total number of pixels.

Step 3 S3 Description The texture image data stored in the image memory (4) 23 is calculated as a _T (m, n) -Dmin, the level is corrected, and the result is again a _T '(m, n). As image (4) 23.

Step 4 S4 Description Similar to step 2, the upper limit threshold Dmax of the image memory (4) is obtained. The upper threshold Dmax is from the image data value 0
The cumulative number of pixels up to Dmax is 95% of the total number of pixels
It is a value equal to S ₁ . For example, if the data is 512 x 512 pixels, the number of pixels from 0 to that value is 512 x 512 x 0.95 = 2490
It is close to 37. By the function of the CPU 11, the parameter controller 13 and the arithmetic unit 14 calculate a value S ₁ which is 95% of the total number of pixels of the image and store it in the parameter memory 15 to set the value of Dmax to 0. Further, the flag map memory 19 is reset and the processor 18 gives a flag to the pixel of the density data Dmax (0 in this case) of the image memory (4) 23 in the flag map memory 19. Then, the area calculation counter determines the number of flags and sends the result Sx to the calculator 14. Therefore, the result Sx is compared with the set value S _1, and if Sx ≧ S ₁ , this step is terminated, and if Sx <S ₁ , 1 is added to Dmax and the flag map memory 19 is stored in the image memory (4) 23. The calculation is performed to give a flag to the pixel of the density data Dmax (which is 1 at this time), and the flag number Sx is calculated again. Then, it is sent again to the arithmetic unit 14 to compare Sx and Sl. By repeating this calculation in sequence, from the density data 0 of the image memory (4) 23
The Dmax value is calculated when the number of pixels of Dmax exceeds 95% of the total number of pixels.

Step 5 Explanation of S5 Upper threshold Dmax obtained in S4 and preset threshold
All pixels of the texture image data of the image memory (4) 23 are linearly converted by Dset so that the density data value Dmax becomes Dset. And stores the result in the image memory (4) 23 as a _T ″ (m, n).
Anything over 5 is set to 255. Specifically, it is performed using the look-up table (4) 30, and the look-up table (4) 30 is instructed by the CPU 11 as shown by the solid line in the standard state FIGS. 4 (a) to 4 (b). Rewrite.
Next, the function of the processor 18 causes the image memory (4) 23
Image data is converted into a look-up table through the look-up table (4) 30 and stored again in the image memory (4) 23. After the calculation, the look-up table (4) 30 is returned to the standard state of FIG. 4 (a). The setting threshold Dset used in the above description is desired to be about one-fourth of the input gradation, but it may be changeable in a conversational format via the parameter setting I / O 16 according to the creator's preference.

Needless to say, the texture data created in the process of steps S2 to S5 may be stored in the memory (external memory or internal memory) at the initial stage.

Description of Step 6 S6 Image memory (1),
The image data of (2), (3) (20 to 22) are subjected to pixel coordinate conversion by the following method and stored in the image memories (5), (6), (7) (24 to 26).

The pixel data a _R (m ₁ , n ₁ ) of the image memory ( ₁ ) is written in the pixel data a ′ _R (m, n) of the image memory (5). The relationship between m and m ₁ and n and n ₁ at this time is represented by texture image data a ″ _T (m ₂ , n ₂ ) of the image memory (4) 23.
Ask more. For example, when converting from ar (m ₁ , n ₁ ) to a _R ′ (m, n) The coordinate conversion is performed by the processor 18 so as to have the relationship of. Here, K is a constant that gives the strength of the coordinate conversion and is stored in the parameter memory 15. Also, K is not a preset value, but I / O16 for parameter setting.
Not only is it possible to set conversationally via
It is changed according to the value of m, n, that is, K = function (m,
It goes without saying that it can be n). In the case of 512 × 512 image data, the values of m and n take values of 2 to 511, respectively, and when m ₁ and n ₁ are smaller, they are set to ₁ , and when they are larger than 512, they are set to 512. The above description is about the image memory (5), but the image memories (6) and (7)
Is also the same.

Step 7 Explanation of S7 Image data of image memory (5) to (7) (24 to 26) Linear conversion However, i = R, G, B is calculated and the result is a ″ i (m, n)
Image memory (5) to (7) (24 to 26)
To store. Specifically, Lucupup table (5) ~
(7) Using 31 to 33, the standard state is rewritten by the instruction of the CPU 11 as shown by the solid lines in FIGS. 4 (a) to 4 (c). Next, by the function of the processor 18, the image data of each of the image memories (5) to (7) 24 to 26 is passed through the rewritten backup table (5) to (7) 31 to 33 and the image memory (5 ) ~
(7) Restore in 24-26. After such lookup conversion, the lookup tables (5) to (7) 31 to 33 are converted to the fourth order.
Return to the standard state of FIG.

Description of Step 8 S8 Image memory (5) ~
(7) Addition operation of the image data of 24-26 and the image data of the image memory (4) is performed. That is, a ″ i (m,
n) + a ″ _T (m, n) (where i = R, G, B) is performed and the result is stored as ai (m, n) in the respective image memories (5) to (7) 24-26.

Step 9 Explanation of S9 Image memory (5) to (7) 24 to 26 image data respectively (However, φ is all if φ or less), and the results are respectively stored in image memory (5) ~
(7) Store in 24 to 26. Specifically, as in step 6, the lookup table (5)-(7) 31-33 is converted as shown by the solid line in FIG. 4 (D).

After the conversion process, the lookup table (5)-(7) 31-
Return 33 to the state shown in FIG. 4 (a).

As a result, the texture screen processing according to the present invention is completed, and the result is the image memory (5) to (7) 24.
Stored in Nos. 26 to 26 and transferred to the output signal processing unit via the I / O 28 for inputting image data.

In the above embodiment, step 6, step 7, step 8 and step 9 are explained in this order, but the process of coordinate conversion (step 6) can be performed by changing the designation of the input / output memory in any of the processes of steps 7-9. . In the above embodiment, the steps 6 to 9 are explained for each image plane, but it goes without saying that the same result can be obtained even if the steps 6 to 9 are repeated for each pixel.

As described above, the inclination of the density change of the texture image (first image) is used to coordinate-convert and synthesize the input image (second image) to obtain a stereoscopic effect according to the density change of the texture image. An image is obtained.

Further, since the input image data is added in a compressed state, it is not necessary to provide a margin for the number of bits for one pixel of the memory, and the memory can be saved.

Further, it becomes possible to automate the synthesizing process, which can be operated by not only experienced users but general users.

Reference images 1, 2, 3 and 4 show the original input image, the texture image, the image obtained by combining the input image and the texture image except for the processing of step S6, and the combined image obtained in the present embodiment.

Although the image processing unit is used as the image processing unit in the embodiment, it is needless to say that the same effect can be obtained by using a general-purpose computer such as a minicomputer.

Further, in the present embodiment, the image is input from the film, but even if the image data is directly fetched from the subject from a still video camera, a video camera, etc., the magnetic floppy disk,
Similar effects can be obtained by inputting image data through a recording medium such as a magnetic tape, an optical disk, a bubble memory or the like.
In addition, I showed the composition with the texture screen image,
The present invention can of course be applied to the case of compositing original images.

〔effect〕

As described above, according to the present invention, in the image processing method for generating the third image by adding the pixel data of the first image and the pixel data of the second image, the coordinates of the coordinates of the third image are changed. Based on the pixel values in the vicinity of the corresponding coordinates of the first image, the coordinates of the second image corresponding to the coordinates of the third image are obtained, and the pixel data at the obtained coordinates of the second image are calculated. Since the pixel data of each coordinate of the first image is added to the pixel data of the written coordinate and the pixel data of the written coordinate is added to the coordinate of the third image, the pixel data of the third image is written. As a result, a third image having a three-dimensional effect corresponding to the first image can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of the flow of a density signal when the present invention is applied to a color scanner, FIG. 2 is a block diagram showing in more detail a part of the image processing section of FIG. 1, and FIG. The figure which shows the flow chart of an Example, FIG. 4 (a),
(B), (c), (d) is a figure showing the state of the look-up table used for implementing in the flow chart of FIG. 1 ... Input drum 2 ... Logarithmic conversion circuit 3 ... AD converter 4 ... Input signal processing unit 5 ... Image processing unit 6 ... Output signal processing unit 7 ... DA converter 8 ... Modulator 9 ... … Laser light source 10 …… Output drum 11 …… Control processor 12 …… CPU memory 13 …… Parameter controller 14 …… Calculator 15 …… Parameter memory 16 …… Parameter setting I / O 17 …… Image controller 18… … Processor 19 …… Flag map memory 20-26 …… Image memory 27-33 …… Ruckup table 34 …… I / O for image data

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Susumu Matsumura 770 Shimonoge, Takatsu-ku, Kawasaki-shi, Kanagawa Canon Inc. Tamagawa Plant (72) Inventor Hiroshi Omura 770 Shimonoge, Takatsu-ku, Kawasaki-shi, Kanagawa Canon Inc. Tamagawa Works (56) References Computer Graphics, Vol. 18, No. 3, 1984, p. 245-252, J. P. Lewis "Texture re Synthesis for Digital Painting"

Claims (1)

[Claims] 1. An image processing method for generating a third image by adding pixel data of the first image and the pixel data of the second image, wherein the first image corresponding to each coordinate which becomes the third image. Based on pixel values near the coordinates of the image, the coordinates of the second image corresponding to the coordinates of the third image are obtained, and the pixel data at the obtained coordinates of the second image are calculated as the first pixel data. An image characterized by being written in the coordinates that become the image of No. 3, and adding the pixel data of each coordinate of the first image to the pixel data of the written coordinates and writing in the coordinates that become the third image. Processing method.