JPH06225213A - Method and device or processing picture - Google Patents

Abstract

PURPOSE:To provide a picture processing method by which picture can easily be enlarged/ reduced at the time of reproducing the picture from picture information compressed by a DCT processing with respect to the enlargement/reduction of the picture. CONSTITUTION:The picture processing method is to obtain the reduced picture from compressed picture data by using discrete cosine transformation(DCT). A process for cutting down the high frequency component of a DCT coefficient displayed by a matrix system and obtaining a corrected DCT coefficient which is frequency analyzed by the DCT processing and which can be displayed by the matrix system of a desired dimension, and the process of executing an inverse DCT calculation for the corrected DCT coefficient and for obtaining picture data of the desired dimension are provided. The picture processing method for obtaining the enlarged picture from compressed picture data by using discrete cosine transformation(DCT) has a process for adding zero as the high frequency component to the DCT coefficient which is frequency-analyzed by the ECT processing and which can be displayed is frequency-analyzed by the ECT processing and which can be displayed by the matrix system and for obtaining the corrected DCT coefficient displayed by the matrix system of the desired dimension and a process for executing inverse DCT calculation for the corrected DCT coefficient and for obtaining picture data of the desired dimension.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to image information processing, and more particularly to image enlargement and reduction. In recent years, the discrete cosine transform (DCT) is often used as a compression technique of image information. The image data compressed by DCT is information obtained by frequency decomposition of the original image.

[0002]

2. Description of the Related Art When processing an image on a display by using a computer system such as a personal computer, it is sometimes desired to enlarge or reduce the image. For example, an image displayed on the full screen may be reduced in the window, or an image in the window may be enlarged in the full screen.

When reducing an image, the simplest method would be to delete image data at a constant rate. For example, if the screen is decomposed into a set of 2 × 2 pixels and only the upper left pixel is adopted, the image can be reduced by 1/2 (1/4 in area). However, the resulting image in this case is not always true to the original image. It is possible for a gray image to turn white or black.

In order to obtain an image more faithful to the original image, an average value is calculated in a certain minute area and the minute area is expressed by the average value. In this case, image information processing for processing the information of the original image is required.

When enlarging an image, the simplest method would be to simply enlarge the original image and make one pixel of information common to a plurality of pixels. However, in this case, the resulting image has a rough texture.

In order to obtain a finer enlarged image, the information of the original image is distributed in the screen of a desired dimension, and data is interpolated. In this case, image processing for creating interpolation data is required.

In recent years, the demand for handling high-resolution images has increased, and data compression techniques for image information of high-resolution images have been developed. FIG. 6 shows an image data compression technique using DCT.

As shown in FIG. 6A, the target screen 50 is divided into small subsections 51. The small section 51 has a size of 8 pixels × 8 pixels, for example. That is, the subdivision 51 constitutes a square matrix of 8 rows and 8 columns including 64 elements. The image information of the screen 50 is processed for each small section 51.

As shown in FIG. 6B, the image data 52 of the small section 51 is processed row by row by the forward DCT processor 53 to obtain a DCT coefficient (F) 54. This DCT
The coefficient 54 is a frequency analysis of image information. The DCT coefficient 54 is processed by the threshold processing device 55, and data below a certain value is truncated. next,
In order to reduce the non-zero data length, the normalization processing device 56 divides the data by a constant value to obtain shortened data 57.

The image data 57 thus obtained
, Both non-zero and zero are mixed, but the high frequency components are almost zero. For non-zero data, Huffman coding is performed and the data is further compressed. In addition, zero-length data is subjected to run-length coding, a block of zeros is treated as one piece of data, and Huffman coding is further performed.

When the original image is reproduced from the compressed data obtained in this way, the image data 57 is reproduced, and then the reverse process of the normalizing process is performed to obtain the forward DCT.
Reverse DCT processing, which is the reverse processing of the processing, is performed.

FIG. 6C is a forward direction D shown in FIG. 6B.
The content of CT processing is shown. The image data f is sandwiched between a transposed cosine coefficient matrix and a cosine coefficient matrix, and a DCT coefficient F is obtained by performing a matrix operation.

The forward DCT processing is further developed.
Then F = D^tfD = {(fD) ^tD}^tCan be expressed as
That is, whether the cosine coefficient matrix D is on the right side of the image data f
, Transpose the obtained matrix, and then again cosine
By multiplying and transposing the number matrix D, the DCT coefficient F
Can be obtained.

[0014]

As described above,
When trying to enlarge or reduce the image using conventional technology,
Information processing for that was necessary. When trying to obtain a high quality image, information processing was complicated.

An object of the present invention is to provide an image processing method capable of easily enlarging or reducing an image when reproducing the image from the image information compressed by the DCT process.

[0016]

An image processing method of the present invention is an image processing method for obtaining a reduced image from image data compressed by using discrete cosine transform (DCT), and frequency decomposition is performed by the DCT process. A step of truncating the high frequency components of the DCT coefficient that can be displayed in a matrix form to obtain a modified DCT coefficient that can be displayed in a matrix form of a desired dimension; and performing an inverse DCT operation on the modified DCT coefficient to obtain image data of the desired dimension. And a step of obtaining

The image processing method of the present invention is an image processing method for obtaining an enlarged image from image data compressed using discrete cosine transform (DCT).
DC that is frequency decomposed by processing and can be displayed in matrix format
A step of adding 0 as a high frequency component to the T coefficient to obtain a modified DCT coefficient that can be displayed in a matrix format of a desired dimension; and a step of performing an inverse DCT operation on the modified DCT coefficient to obtain image data of the desired dimension. Have.

[0018]

The DCT coefficient obtained by subjecting the image information to the DCT processing is a frequency decomposition of the image information. When reducing the image, the high frequency component of the DCT coefficient is cut off and the inverse D
If the image data is obtained by CT calculation, a reduced image in which the low-frequency component important for the image is maintained as it is can be obtained.

When the image is enlarged, 0 is added as a high frequency component to the DCT coefficient and the image data is obtained by the inverse DCT operation, whereby an enlarged image retaining the characteristics of the original image can be obtained. In this case, appropriate interpolation is automatically performed.

[0020]

1 is a block diagram showing an image reducing / enlarging device according to an embodiment of the present invention. The original image compressed data 1 obtained by performing the DCT process on the original image is decoded by Huffman coding or run length coding by the decoder 7 to obtain the DCT coefficient 54. DCT coefficient 54
Is information obtained by frequency analysis of the original image.

When the image is reduced or enlarged, the DCT coefficient is 5
4 is supplied to the matrix conversion circuit 8. The matrix conversion circuit 8 is
The size value 5 of the original image and the size value 3 of the output image are input. When the size value 5 of the original image is larger than the size value 3 of the output image, the high frequency component of the input DCT coefficient in the matrix format is cut off and converted into a matrix of small dimensions, and the inverse DC
It is supplied to the T circuit 10.

The size value 3 of the output image is input to the transformation matrix supply circuit 9, and the transformation matrix matching the size value 3 of the output image is supplied from the transformation matrix supply circuit 9 to the inverse DCT circuit 10. The inverse DCT circuit 10 uses the supplied transform matrix to inverse DCT the DCT coefficient supplied from the matrix transform circuit 8.
T calculation is performed to reproduce the image information.

When the size value 5 of the original image is smaller than the size value 3 of the output image, the matrix conversion circuit 8 outputs D in matrix form.
0 is added to the CT coefficient as a high-frequency component, and a DCT coefficient in a matrix format that matches the size value 3 of the output image is supplied to the inverse DCT circuit 10. The conversion matrix supply circuit 9 supplies a conversion matrix adapted to the size value 3 of the output image. The inverse DCT circuit 10 performs an inverse DCT operation according to the size value 3 of the output image.

The operations performed by the matrix conversion circuit 8 and the inverse DCT circuit 10 will be described in more detail below. Figure 2
Indicates conversion of DCT coefficients when reducing an image. FIG. 2A shows the structure of DCT coefficients when the DCT processing is performed by a standard 8 × 8 unit. When the original image has 8 × 8 units, the DCT coefficient can also be displayed in an 8 × 8 matrix. The DCT coefficient F is obtained by frequency-analyzing the information of the original image. F _{00 in} the upper left indicates a DC component in the horizontal and vertical directions, and F _i0 directed in the horizontal direction is
In the horizontal direction, the information gradually becomes high-frequency components.

Further, the vertical direction F _0j becomes information indicating a high frequency component gradually in the vertical direction. The lower frequency component is more important for the image information, and the higher frequency component is more noise-like information.

Now, consider the case where an image is reduced to 1/2 (1/4 in area). In this case, the low frequency component 4 × 4 unit is extracted from the 8 × 8 DCT coefficient. Figure 2
2B is a DCT coefficient of a 4 × 4 reduced image obtained by extracting the low-frequency component from the DCT coefficient of the original image of FIG. 2A. The difference between the DCT coefficient in FIG. 2A and the DCT coefficient in FIG. 2B is the presence or absence of a high frequency component.

From the DCT coefficient shown in FIG.
When the DCT coefficient shown in (B) is obtained, the low frequency component important as the information of the original image is retained as it is, and the image information in which only the high frequency component is cut off is obtained.

FIG. 3 shows a cosine coefficient matrix used for reproducing the information of the original image from the DCT coefficient. Figure 3
FIG. 3A shows a 8 × 8 unit cosine coefficient matrix, and FIG. 3B shows a 4 × 4 unit cosine coefficient matrix.
When the 8 × 8 unit DCT coefficient of FIG. 2A is reproduced with the same size, the 8 × 8 unit cosine coefficient matrix shown in FIG. 3A is used.

In order to obtain a reduced image, when the matrix conversion circuit 8 obtains the DCT coefficient of 4 × 4 units of FIG. 2B, the 4 × 4 unit cosine coefficient matrix shown in FIG. 3B is obtained. Used.

If the inverse DCT operation is performed by the inverse DCT circuit 10 by using the cosine coefficient matrix of 4 × 4 units shown in FIG. 3B for the DCT coefficient of the reduced image shown in FIG. 2B, 4 × 4 is obtained. Image information is obtained. Since this image information stores the low-frequency component of the original image as it is, it becomes a reduced image that is faithful to the original image.

FIG. 4 is a diagram for explaining the DCT process when enlarging an image. The original image is 4 × in FIG.
The DCT coefficient in the case of an image of 4 units is shown. Consider a case where an image of 8 × 8 units is reproduced from the DCT coefficient of 4 × 4 units.

4 × 4 unit DCT shown in FIG.
When an image of 8 × 8 units is reproduced from the coefficients, as shown in FIG. 4B, 0 is added as a high frequency component in the matrix conversion circuit 8 to create a DCT coefficient of 8 × 8 units. This process can be performed extremely easily because only 0 is added according to the size of the output image.

If the inverse DCT operation is performed by the inverse DCT circuit 10 using the cosine coefficient matrix of the 8 × 8 unit shown in FIG. 3A, the expanded DCT coefficient shown in FIG. 4B is 8 × 8. Image information of the unit is obtained.

FIG. 5 shows 0 for a DCT coefficient of a predetermined dimension.
FIG. 6 is a diagram for explaining the effect of addition of the above on the obtained image information. FIG. 5A shows a simplified matrix operation performed in the inverse DCT operation.

The matrix at the center of the left side of FIG. 5A is the matrix of DCT coefficients. In this matrix of DCT coefficients, 0 is added as a high frequency component according to the size of the enlarged output image. The matrices on both sides of the matrix of DCT coefficients are the cosine coefficient matrix and its transposed matrix, but here, for convenience of explanation, it is assumed that all elements are 1 for simplicity.

When this matrix calculation is performed, as shown in the figure, appropriate image information also appears in the portion where 0 is added as a DCT coefficient. This is because the addition of 0 to the DCT coefficient is the addition of 0 as a high frequency component, and the signal obtained as image information is the expanded low frequency component of the original DCT coefficient. is there.

FIG. 5B is an equation more accurately representing the inverse DCT operation. Image data 52 is obtained by multiplying the matrix of DCT coefficients 54 by the cosine coefficient matrix from the left and the transposed cosine coefficient matrix from the right.

If 0 is added to the DCT coefficient 54 and the transposed cosine coefficient matrix and transposed cosine matrix having the expanded dimension are used, the expanded image data 52 is obtained, and the low frequency which the DCT coefficient 54 has at that time is obtained. The ingredients are stored as is.

In this way, when the image information is the image information compressed by the DCT processing, the size of the matrix of DCT coefficients is changed so that only the inverse DCT operation for reproducing the image information is performed. Reduced and enlarged image information can be obtained.

In the configuration of FIG. 1, the transformation matrix supply circuit 9 is used to receive the size value 3 of the output image and create a cosine coefficient matrix and a transposed cosine coefficient matrix of a predetermined size. The transform matrix supply circuit 9 may be a circuit that stores a cosine coefficient matrix of a possible size or one that calculates a cosine coefficient matrix by calculation. When the size value 5 of the original image and the size value 3 of the output image are the same, the matrix conversion circuit 8 supplies the input DCT coefficient to the inverse DCT circuit 10 as it is.

Although the case of converting a square matrix into a square matrix has been described, the enlargement and reduction of an image are not limited to isotropic ones. An image having an arbitrary aspect ratio can be reproduced by freely setting the horizontal magnification and the vertical magnification and creating the DCT coefficient and the cosine transform matrix according to the shape of the output image. It is also possible to enlarge one direction and reduce the other direction.

The present invention has been described above with reference to the embodiments.
The present invention is not limited to these. For example,
It will be apparent to those skilled in the art that various changes, improvements, combinations and the like can be made.

[0043]

As described above, according to the present invention,
When an image is reproduced from the compressed image data obtained by the DCT process, the image is enlarged by an extremely simple process,
Reduction can be performed. The reproduced image thus obtained is one in which the features of the original image are efficiently saved.

[Brief description of drawings]

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

FIG. 2 is a conceptual diagram for explaining processing of DCT coefficients when reducing an image.

FIG. 3 is a conceptual diagram showing an example of a cosine coefficient matrix used for DCT calculation.

FIG. 4 is a conceptual diagram for explaining processing of DCT coefficients when enlarging an image.

FIG. 5 is a conceptual diagram for explaining the intention of adding 0 to the DCT coefficient when enlarging an image.

FIG. 6 is a conceptual diagram for explaining DCT processing.

[Explanation of symbols]

 1 Original Image Compressed Data 3 Output Image Size Value 5 Original Image Size Value 7 Decoder 8 Matrix Transformation Circuit 9 Transformation Matrix Supply Circuit 10 Inverse DCT Circuit 11 Output Image Data 50 Screen 51 8 × 8 Unit (Small Section) 52 Images Data 53 Forward DCT circuit 54 DCT coefficient 55 Threshold processing circuit 56 Normalize processing circuit 57 Compressed data

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Office reference number FI technical display location H04N 7/133 Z

Claims (6)

[Claims] 1. An image processing method for obtaining a reduced image from image data compressed by using a discrete cosine transform (DCT), which is a high frequency component of a DCT coefficient which is frequency decomposed by the DCT process and can be displayed in a matrix format. To obtain a modified DCT coefficient that can be displayed in a matrix format with a desired dimension, and an inverse DCT operation on the modified DCT coefficient to obtain image data with the desired dimension. 2. An image processing method for obtaining an enlarged image from image data compressed by using the discrete cosine transform (DCT), the high frequency component being a DCT coefficient which is frequency decomposed by the DCT process and can be displayed in a matrix format. 0 to add a modified DCT coefficient that can be displayed in a matrix format having a desired dimension, and an inverse DCT operation is performed on the modified DCT coefficient to obtain image data of the desired dimension. . 3. Further, receiving a signal representing a desired dimension,
The image processing method according to claim 1, further comprising the step of supplying a conversion matrix of that dimension. 4. An image processing apparatus for obtaining a reduced image from image data compressed using discrete cosine transform (DCT), the high frequency component of a DCT coefficient which is frequency decomposed by DCT processing and can be displayed in a matrix format. (8) means for obtaining modified DCT coefficients that can be displayed in a matrix format with a desired dimension
An image processing apparatus comprising: and a means (10) for performing an inverse DCT operation on the modified DCT coefficient to obtain image data of the desired dimension. 5. An image processing apparatus for obtaining an enlarged image from image data compressed by using discrete cosine transform (DCT), the high frequency component being a DCT coefficient which is frequency decomposed by DCT processing and can be displayed in a matrix format. Means for obtaining a modified DCT coefficient that can be displayed in a matrix of a desired dimension by adding 0 as
An image processing apparatus comprising: and a means (10) for performing an inverse DCT operation on the modified DCT coefficient to obtain image data of the desired dimension. 6. Further, receiving a signal representing a desired dimension,
An image processing apparatus according to claim 4 or 5, further comprising means (9) for supplying a conversion matrix of that dimension.