JP2001043361A - Image processing method and image processing using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To quantitatively generate a high definition image at low cost by making the input image of low resolution into that of high definition through interpolation processing and restoring an original object by applying deconvolution processing to this high definition image with a dot image distribution function. SOLUTION: An image processing means 50 samples an input image 102 with low resolution and stores it in an image storage means 5. A high definition processing means 6 generates a high definition image 103 by applying interpolation processing to the input image 102. With this processing, the resolution is doubled and the boundary of gradation data is averaged as well. A deconvolution processing means 7 restores the original object by applying deconvolution processing to the high definition image 103 and outputs it as an output image 104. In deconvolution processing, the original object is restored while using the dot image distribution function.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method for generating a high-definition image, and more particularly to an image processing method capable of inexpensively and quantitatively generating a high-definition image and an image processing method using the same. Related to the device.

[0002]

2. Description of the Related Art As a conventional image processing method for obtaining a high-definition image, there is an image processing method for increasing the definition by interpolating low-resolution pixels. FIG. 7 is an explanatory diagram showing an example of such a conventional image processing method.

[0003] "480 x 64" as shown in FIG.
As shown in FIG.
The image is processed into a high-definition image of about 1280 × 960 ″, which is about twice.

[0004] By increasing the definition of an image by the interpolation process, it is possible to smooth oblique lines. For example, FIG. 8 is an explanatory diagram showing an example in which the oblique line portions before and after the interpolation processing are enlarged.

It can be seen that the line after interpolation shown in FIG. 8B is smoother than the line before interpolation shown in FIG. 8A.

The optical component has an optical transmission characteristic corresponding to the transmission characteristic of the electric system. FIG. 9 is an explanatory diagram showing an example of such an optical transfer characteristic. In FIG. 9, 2
Denotes an object, 3 denotes a lens which is an optical component, and 4 denotes a camera which is a photographing device.

Light from the object 2 is condensed by the lens 3 and is incident on the camera 4. The camera 4 captures an image of the object 2. Under these circumstances, when the contrast is taken on the vertical axis and the resolution is taken on the horizontal axis, "CL11" in FIG.
The transfer characteristics are as shown in FIG.

As shown by "CL11" in FIG. 9, the contrast decreases as the resolution of the photographing of the camera increases (high definition), and it can be seen that the relationship is similar to the relationship between the gain and the frequency in the electric system. .

That is, as shown in FIG.
In the case of the image photographed in the above, and the image photographed in twice the resolution "2N", the latter has a much lower contrast.

Therefore, conventionally, such sharpening of contrast is corrected by sharpening using a sharpening filter or the like. FIG. 10 is an explanatory diagram for explaining the sharpening process by the sharpening filter.

When the transfer characteristic indicated by "CL21" in FIG. 10 is sharpened by a sharpening filter, "CL2" in FIG.
The characteristic is such that the contrast is enhanced in the portion of the resolution “2N” as shown in FIG. 2, so that the decrease in the contrast can be corrected.

Further, instead of increasing the definition of an image,
Convert the image to PSF (Point-Spread Function: Point-Spread Functio)
There is a deconvolution process for restoring the original object by deconvolution in n).

FIG. 11 is an explanatory diagram showing an example of such a deconvolution process. In FIG. 11, 1 is a deconvolution processing means, 100 is an input image, 10
1 is an output image. The input image 100 is input to the deconvolution processing means 1, and an output image 101 is output as a processing result.

By performing such processing, it is possible to restore an object which is a photographing target, such as a photograph of the input image 100.

[0015]

However, in the conventional example shown in FIG. 7, there is a problem in that the oblique lines are smoothed to average the pixels, but the edges are unclear. FIG. 12 is an explanatory diagram showing an example of an enlarged edge portion before and after such interpolation processing.

The edge portion before the interpolation processing shown in FIG. 12A has a gradation characteristic as shown by "CL01" in FIG. 12, and the edge portion shown by "ED01" in FIG. Is switched sharply.

On the other hand, in the edge portion after the interpolation processing shown in FIG. 12B, as shown by "CL02" in FIG. 12, the gradation is averaged before and after the edge portion shown by "ED02" in FIG. Therefore, the gradation of the edge portion is smoothly switched, and the edge portion becomes unclear.

Further, in the sharpening process using the sharpening filter, there is a problem that the contrast is corrected, but the quantitativeness in the image is lost.

Further, in the case of the deconvolution processing shown in FIG. 11, there is a problem that the original object can be restored, but the resolution cannot be increased.

For this reason, in order to obtain a restored high-definition image, it is necessary to take a high-resolution input image with a high-precision camera and perform deconvolution processing on the input image. In this case, there are problems that the apparatus becomes expensive, the image tends to be dark, and the photographing conditions become severe. Therefore, an object of the present invention is to realize an image processing method and an image processing apparatus using the same, which can quantitatively generate high-definition images quantitatively at low cost.

[0021]

In order to achieve the above object, according to the first aspect of the present invention, in an image processing method for generating a high-definition image, a low-resolution input image is interpolated. By processing and refining the image, deconvolving the image with the point spread function and restoring the original object, it is possible to quantitatively and inexpensively generate a high-definition image. A possible image processing method can be realized.

According to a second aspect of the present invention, in the image processing method for generating a high-definition image, a low-resolution input image is subjected to interpolation processing to obtain a high-definition image, and the high-resolution image is converted into an original image. By restoring the original object by performing high-pass filtering corresponding to the point spread function, an image processing method capable of inexpensively and quantitatively generating a high-definition image can be realized.

According to a third aspect of the present invention, there is provided an image processing apparatus using an image processing method for generating a high-definition image. Provided with image processing means for deconvolution processing the restored image with a point spread function and restoring the original object, it is possible to realize an image processing apparatus capable of inexpensively and quantitatively generating a high-definition image. .

According to a fourth aspect of the present invention, the image processing apparatus includes a high-definition processing means for interpolating a low-resolution input image to obtain a high-definition image, and converting the high-resolution image into a point spread function. And deconvolution processing means for deconvolution processing to restore the original object, thereby realizing an image processing apparatus capable of inexpensively and quantitatively generating a high-definition image.

According to a fifth aspect of the present invention, in the image processing apparatus according to the fourth aspect, the deconvolution processing means is a blind deconvolution processing means, so that the point spread function of the optical system is reduced. There is no need to actually measure.

According to a sixth aspect of the present invention, there is provided an image processing apparatus using an image processing method for generating a high-definition image. Image processing means for restoring the original object by performing high-pass filtering corresponding to the point spread function of the original image on the original image enables quantitatively and inexpensively to generate high-definition images A simple image processing device can be realized.

According to a seventh aspect of the present invention, in the image processing apparatus according to the sixth aspect, the image processing apparatus comprises:
High-definition processing means for interpolating a low-resolution input image to obtain high definition; and converting the high-resolution image into a high-pass image corresponding to the point spread function of the original image. The digital processing means for restoring the original object by performing the filtering process can realize an image processing apparatus capable of quantitatively and inexpensively generating a high-definition image at low cost.

[0028]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail with reference to the drawings. FIG. 1 is a configuration block diagram showing an embodiment of an image processing apparatus using the image processing method according to the present invention. In FIG. 1, 5 is an image storage unit, 6 is a high definition processing unit,
7 is a deconvolution processing means, 102 is an input image, 103 is a high definition image, and 104 is an output image. 6 and 7 constitute the image processing means 50.

The input image 102 is input to the image storage means 5, and the output of the image storage means 5 is connected to the high definition processing means 6. High definition image 1 output from high definition processing means 6
03 is connected to the deconvolution processing means 7, and the deconvolution processing means 7 outputs an output image 104.

The operation of the embodiment shown in FIG. 1 will now be described with reference to FIGS.
This will be described with reference to FIGS. 3, 4, 5, and 6. FIG. FIG. 2 is a flowchart for explaining the operation of the image processing means, FIG. 3 is a characteristic curve diagram showing the relationship between the position of the original object and the gradation, and FIG. FIG. 5 is a characteristic curve diagram showing the relationship between the position and the gradation of the interpolation-processed high-definition image, and FIG. 6 is a characteristic curve diagram showing the relationship between the position and the gradation of the deconvolution-processed image. It is a curve figure.

At step S001 in FIG. 2, the image processing means 50 loads the input image 102 into the image storage means 5.

For example, an object having characteristics as indicated by "CL31" in FIG. 3 is referred to as "ST01", "ST0" in FIG.
2 "," ST03 "and" ST04 "are sampled at a low resolution at the positions indicated by" ST04 "and stored in the image storage means 5.

FIG. 4 shows an image sampled at a low resolution. For example, “ST” in FIG.
The data sampled at the positions indicated by "01" and "ST02" have gradations of "10", respectively, and are represented by "S" in FIG.
The data sampled at the positions indicated by T03 ”and“ ST04 ”have gradations of“ 110 ”, respectively.

Then, the input image 102 fetched in "S002" in FIG. 2 is subjected to interpolation processing to generate a high-definition image 103.

For example, the data of the gradation "10" and the gradation "110" in FIG. 5 are increased in resolution to double the resolution, and the sampling point "MP01" in FIG. 5 at "ST15" in FIG. The part shown in "" is removed and "ST1" in FIG.
The sampling point indicated by "4" is moved to the portion indicated by "MP02" in FIG.

By this processing, the resolution is doubled and the boundary between the data of the gradation "10" and the data of the gradation "110" is averaged.

That is, "ST11" and "ST1" in FIG.
The data sampled at the positions indicated by "2" and "ST13" have gray scales of "10", respectively, and "ST" in FIG.
The data sampled at the positions indicated by "14" and "ST15" have gradations of "60", respectively, and "S" in FIG.
The data sampled at the positions indicated by T16 "," ST17 "and" ST18 "have gradations of" 110 ", respectively.
become.

Finally, in "S003" in FIG. 2, the high-definition image 103 is further deconvoluted to restore the original object and output as an output image 104.

In the case of deconvolution processing, FIG.
Since the original characteristic indicated by “CL31” in the middle can be estimated from the characteristic of the point spread function, the characteristic is restored to, for example, the characteristic indicated by “CL41” in FIG. 6, and this characteristic is “CL31” in FIG.
And the characteristics shown in FIG. The resolution of the image at this time is twice the resolution of the input image 102.

As a result, by improving the resolution of an image captured at a low resolution by interpolation and then restoring the characteristics of the original object by deconvolution processing, a high-precision camera becomes unnecessary and the cost is reduced. Higher definition is achieved, and the contrast is corrected by deconvolution processing, so that quantitativeness is maintained.

That is, it is possible to realize an image processing method capable of quantitatively producing high-definition images at low cost and an image processing apparatus using the same.

In the embodiment shown in FIG. 1, the image processing means 50 is divided into a high-definition processing means 6 and a deconvolution processing means 7 for simplicity of description. A configuration in which two processes are sequentially performed in the image processing unit 50 may be employed.

In the embodiment shown in FIG. 1, the image storage means 5 is described for simplicity of explanation.
This is unnecessary when the storage means is provided in 0, and is not an essential component.

The deconvolution process is a process for restoring the original object using the above-mentioned PSF. However, even if the PSF is not actually measured, the original object can be input by inputting the lens NA or the like as a parameter. Blind deconvolution processing means capable of restoring the image may be used. In this case, there is no need to actually measure the PSF of the optical system.

In the deconvolution processing, the characteristics of the original object are restored by using the Fourier transform, but the same processing can be performed by the filter processing.
For example, high-pass digital filter means may be used.

In this case, the characteristic of "CL11" in FIG.
G (ω): ω is the spatial frequency and the original characteristic is “H (ω)”
Then, a digital filter having the characteristic of H (ω) = 1 / G (ω) (1) may be used.

Since the characteristic of "CL11" in FIG. 9 is the Fourier transform of the point spread function, "H (ω)" can be created if this point spread function is known.

[0048]

As is apparent from the above description,
According to the present invention, the following effects can be obtained. Claims 1, 2,
According to the third, fourth, sixth and seventh aspects of the present invention, a high-precision camera is provided by restoring the characteristics of an original object by deconvolution processing after increasing the resolution of an image captured at low resolution by interpolation processing. Unnecessary and cheaper,
High definition is achieved by interpolation processing, and contrast is corrected by deconvolution processing, so that quantitativeness is maintained. That is, it is possible to realize an image processing method capable of quantitatively generating a high-definition image inexpensively and an image processing apparatus using the same.

According to the fifth aspect of the present invention, by using the blind deconvolution processing means, there is no need to actually measure the PSF of the optical system.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram illustrating an embodiment of an image processing apparatus using an image processing method according to the present invention.

FIG. 2 is a flowchart illustrating an operation of an image processing unit.

FIG. 3 is a characteristic curve diagram showing a relationship between an original object position and gradation.

FIG. 4 is a characteristic curve diagram showing a relationship between a position of a low-resolution image and a gradation.

FIG. 5 is a characteristic curve diagram showing a relationship between a position and a gradation of a high-definition image subjected to an interpolation process.

FIG. 6 is a characteristic curve diagram showing a relationship between a position and a gradation of an image further subjected to deconvolution processing.

FIG. 7 is an explanatory diagram illustrating an example of a conventional image processing method.

FIG. 8 is an explanatory diagram showing an example in which a diagonal line portion before and after an interpolation process is enlarged.

FIG. 9 is an explanatory diagram illustrating an example of an optical transfer characteristic.

FIG. 10 is an explanatory diagram illustrating sharpening processing by a sharpening filter.

FIG. 11 is an explanatory diagram illustrating an example of a deconvolution process.

FIG. 12 is an explanatory diagram showing an example in which an edge portion before and after an interpolation process is enlarged.

[Explanation of symbols]

 1,7 Deconvolution processing means 2 Object 3 Lens 4 Camera 5 Image storage means 6 High definition processing means 50 Image processing means 100,102 Input image 101,104 Output image 103 High definition image

Claims (7)

[Claims] In an image processing method for generating a high-definition image, a low-resolution input image is subjected to interpolation processing to obtain high definition, and the high-resolution image is subjected to deconvolution processing using a point spread function. An image processing method for restoring an original object by using the method. 2. An image processing method for generating a high-definition image, comprising: interpolating a low-resolution input image to obtain a high-definition image; and converting the high-resolution image into a point spread function of an original image. An image processing method characterized by restoring an original object by corresponding high-pass filtering. 3. An image processing apparatus using an image processing method for generating a high-definition image, comprising: interpolating a low-resolution input image to obtain a high-definition image;
An image processing apparatus comprising image processing means for deconvolving the high definition image with a point spread function and restoring an original object. 4. An image processing apparatus comprising: a high-definition processing means for interpolating a low-resolution input image to obtain high definition; and performing a deconvolution process on the high-resolution image by a point spread function. 4. The image processing apparatus according to claim 3, further comprising a deconvolution processing unit for restoring an original object. 5. An image processing apparatus according to claim 4, wherein said deconvolution processing means is a blind deconvolution processing means. 6. An image processing apparatus using an image processing method for generating a high-definition image, comprising: interpolating a low-resolution input image to obtain a high-definition image;
An image processing apparatus comprising: image processing means for restoring an original object by performing a high-pass filter process on the high definition image corresponding to a point spread function of the original image. 7. An image processing apparatus, comprising: a high-definition processing means for interpolating a low-resolution input image to increase the definition; and converting the high-resolution image into an original image. 7. The image processing apparatus according to claim 6, further comprising digital filter means for restoring the original object by performing high-pass filtering corresponding to the point spread function of the image.