JP2001112002A - Digital moving picture decoder capable of image size conversion - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a digital moving picture decoder capable image size converting, which can change the size of the output image without causing rapid deterioration even in a part in fast motion with a simple configuration such that no low-pass filtering and interleaving processing are required after decoding. SOLUTION: The digital moving picture decoder is provided with a motion vector conversion section that converts the magnitude of a motion vector from motion vector information included a bit stream. The decoder uses only a DCT coefficient of low frequencies when the magnitude of the motion vector is small to conduct inverse DCT and uses a DCT coefficient of high frequencies in addition to the DCT coefficient of low frequencies to conduct inverse DCT when the magnitude of the motion vector is large.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital moving picture decoding apparatus, and more particularly, to a picture size conversion apparatus capable of changing the size of an output picture when decoding a coded signal containing a compressed digital moving picture. A possible digital video decoding device.

[0002]

2. Description of the Related Art In recent years, DVDs (Digital Ver.
With the spread of digital disks (satile disks) and the start of digital broadcasting, there is a demand for a function of decoding and reproducing digital moving images encoded on a personal computer. On the other hand, the resolution (the number of display pixels) of a personal computer varies widely depending on the graphic card used and its settings. In addition, personal computers
The reproduced image may be drawn on a window that can be scaled. Under such a background, a decoding device that can change the image size of a digital moving image when decoding a moving image is indispensable.

[0003] Here, ISO (International), which is one of the typical digital video coding methods, is used.
Organization for Standard
Recommendation) ISO / IEC13818-2
MPEG2 specified in the above will be described. MPEG2
Coding is a discrete cosine transform (Discc) of predictive coding.
Rete Cosin Transform,
DCT) is a hybrid coding method. As a general procedure, first, a macro block (1
Temporal information compression by motion compensation is performed on a luminance signal in units of (6 × 16 pixels). Next, the macro block is 8
It is subdivided into sub-blocks of × 8 pixels, and spatial information compression by DCT is performed. With this spatial information compression, D
CT coefficients are generated. Image format is 4: 2: 0
, The luminance signal block 4 per macroblock
And two color difference signal blocks.
After the DCT processing is performed, the entire code generation amount is controlled by controlling the quantization processing of the DCT coefficient. Finally,
Entropy coding based on Huffman codes is performed.
On the other hand, when decoding MPEG2, the procedure reverse to that of encoding is performed.

[0004] A conventional digital video decoding device capable of image size conversion will be described below. FIG. 7 is a block diagram showing a configuration example of a conventional digital video decoding device.
In FIG. 7, the digital video decoding device includes a Huffman decoding unit 201, an inverse quantization unit 202, an inverse DCT unit 20.
3, adder 204, accumulator 205, motion compensator 206,
A switch unit 207 and a DCT coefficient control unit 208 are provided. Next, the operation of the digital video decoding device having the above configuration will be described. First, the Huffman decoding unit 201
Is supplied with a bit stream of a Huffman-encoded video signal. The Huffman decoding unit 201 performs Huffman decoding on the input bit stream, and outputs DCT coefficients, a quantization table, and a motion vector to the inverse quantization unit 202. The inverse quantization unit 202 outputs the Huffman decoding unit 20
Of the DCT coefficients obtained in step 1, only the DCT coefficients on the low frequency side where desired frequency characteristics are obtained are multiplied by a constant using a quantization table, and the DCT coefficients on the high frequency side are deleted. Or replace with 0. The inverse DCT unit 203 uses the low-frequency DCT coefficient multiplied by a constant in the inverse quantization unit 202 and performs inverse DT so that an output of the number of pixels corresponding to a desired pixel size is obtained.
CT is performed and an image signal is output.

[0005] The image signal obtained by the inverse DCT section 203 is
In the case of an I-picture (Intra-Picture), it is input to an output terminal (not shown) and the storage unit 205 via the switch unit 207. On the other hand, the image signal obtained by the inverse DCT section 203 is a P-picture (Predictive).
Picture) or B picture (Bidirect)
ionicly predictive-Pictur
In the case of e), it is used as an addition signal of the addition unit 204. The addition unit 204 performs addition with the prediction signal generated by the motion compensation unit 206, and outputs an image signal to the output terminal and the storage unit 205 via the switch unit 207. The motion compensation unit 206 uses the image signal accumulated in the accumulation unit 205 and the motion vector decoded by the Huffman decoding unit 201,
Perform motion compensation and generate a prediction signal. The DCT coefficient control unit 208 controls the inverse quantization unit 202 and the inverse DCT unit 203 to obtain a low-frequency DCT that can obtain a frequency characteristic according to a desired pixel size among 8 × 8 DCT coefficients per block. Control is performed so that each process is performed using only the coefficient.

[0006]

However, in the above conventional configuration, the inverse quantization unit 202 and the inverse DCT unit 203
Since the decoding process is performed using only the low-frequency DCT coefficients, there is a problem that the image quality is deteriorated in a portion where the movement is severe.

Therefore, according to the present invention, the size of an output image is reduced by a simple configuration that does not cause a large image quality deterioration even in a portion where movement is intense and does not require a low-pass filtering process and a decimation process after a decoding process. It is an object of the present invention to provide a digital video decoding device capable of changing an image size, which can be changed.

[0008]

A first aspect of the present invention is a digital video decoding apparatus capable of converting the size of an image to be output, which has at least information of a picture type and a DCT coefficient. When a coded digital video bit stream is input, the input bit stream is Huffman-decoded,
A Huffman decoding unit for acquiring information including a DCT coefficient, a quantization table, and a motion vector; an image size setting unit for setting image size information when outputting an image based on an input bit stream; If the bit stream contains motion vector information,
A motion vector conversion unit configured to convert the size of the motion vector acquired by the Huffman decoding unit based on the image size information set in the image size setting unit; Inverse quantizing unit to be transformed, and DC inversely quantized by the inverse quantizing unit.
A DCT coefficient mask part for replacing a part of the T coefficient with 0, and a D coefficient part of which is replaced with 0 by the DCT coefficient mask unit.
An inverse DCT unit for performing an inverse DCT on a CT coefficient; a DCT coefficient control unit for controlling the DCT coefficient mask unit and the inverse DCT unit based on the size of the motion vector information and the image size information; A storage unit that stores an image including data inversely DCTed by the DCT unit; a motion vector converted by the motion vector conversion unit; and a motion compensation image generated by using the image stored in the storage unit. A motion compensation unit, an addition unit that adds the motion-compensated image generated by the motion compensation unit, and the data that has been subjected to the inverse DCT by the inverse DCT unit, and an image from the inverse DCT unit based on the picture type. ,
Or a switch unit that outputs an image from the addition unit, and the inverse DCT unit, when the magnitude of the motion vector is small, among the DCT coefficients acquired by the Huffman decoding unit, Inverse DCT is performed using only the coefficient, and when the magnitude of the motion vector is large, the inverse DCT is performed using the DCT coefficient on the high frequency side in addition to the DCT coefficient on the low frequency side.

According to a first aspect of the present invention, in a still image portion, a low-frequency DCT coefficient corresponding to a desired image size is selected.
Inverse DCT according to a desired image size is performed using CT coefficients. On the other hand, in the first invention, in the moving image part, in addition to the low-frequency DCT coefficient corresponding to the image size, the high-frequency DCT
Inverse DCT is performed using the coefficients. And the first invention is
The DCT coefficient is converted into pixel data, and the band is limited to the output image in the still image section. Thus, by outputting an arbitrary number of output pixels without impairing the motion component in the moving image portion, it is possible to change the output image size while suppressing the image quality deterioration of the moving image portion.

A second invention is dependent on the first invention,
The digital video decoding device further includes a coefficient counting unit that divides the DCT coefficient in the block into a plurality of regions, and counts the number of DCT coefficients having a value other than 0 existing in each region, wherein the inverse DCT When the magnitude of the motion vector is large, the DCT of the region where the count value is large in the coefficient counting unit is added to the DCT coefficient on the low frequency side.
Inverse DCT is performed using the coefficients.

In the second invention, the coefficient counting section counts the number of DCT coefficients having a value other than 0 for each area, thereby detecting to which area the motion component of the image is assigned. I do. Then, the inverse DCT unit uses the DCT coefficient of the high band of the region with a large count in addition to the DCT coefficient of the low band only in the region with a large count value. As described above, according to the second invention, the inverse DCT can be performed using the DCT coefficient of the high frequency band in the region where the motion component is large, but not in the entire region.
The calculation amount of T can be reduced.

A third invention is dependent on the first invention,
The digital video decoding device further includes a DCT coefficient absolute value sum calculation unit that divides a DCT coefficient in the block into a plurality of regions, and calculates a sum of absolute values of the DCT coefficients present in each region. When the magnitude of the motion vector is large, inverse DCT is performed using the DCT coefficient of the region where the output of the DCT coefficient absolute value sum calculation unit is large, in addition to the DCT coefficient on the low frequency side.

In the third aspect, the DCT coefficient absolute value sum calculation unit calculates the absolute value sum of the DCT coefficient for each area, thereby detecting which area the motion component of the image is assigned to. I do. Then, the inverse DCT unit uses the DCT coefficient of the high band of the region with a large count in addition to the DCT coefficient of the low band only in the region with a large absolute value sum. Thus, the third invention does not cover all areas,
Inverse D using a high-frequency DCT coefficient for an area with a large motion component
Since CT can be performed, the amount of inverse DCT computation can be reduced while suppressing image degradation.

According to a fourth aspect of the present invention, there is provided a recording medium storing a program for realizing, on a computer, an environment for converting the size of an image to be output, the program having at least information of a picture type and a DCT coefficient. And when a coded digital video bit stream is input, Huffman decoding is performed on the input bit stream to obtain information including a picture type, a DCT coefficient, a quantization table, and a motion vector. Processing, image size setting processing in which image size information is set when outputting an image based on an input bit stream, and when the input bit stream contains motion vector information, the image Based on the image size information set in the size setting process, obtained in the Huffman decoding process Motion vector conversion processing for converting the size of the input vector, inverse quantization processing for inversely quantizing DCT coefficients of the input bit stream, and a part of the DCT coefficients inversely quantized in the inverse quantization processing. A DCT coefficient masking process for replacing the DCT coefficient with 0, an inverse DCT process for performing an inverse DCT on the DCT coefficient partially replaced with 0 in the DCT coefficient masking process, and a magnitude of the motion vector obtained in the Huffman decoding process And a DCT coefficient control unit that controls DCT coefficients handled in the DCT coefficient mask processing and the inverse DCT processing based on the image size information set in the image size setting processing, and at least inverse DCT performed in the inverse DCT processing. Accumulation processing for accumulating the image including the data, the motion vector converted in the motion vector conversion processing, and the image accumulated in the accumulation processing. Preparative a motion compensation processing for generating a motion compensation image using a motion compensation image the generated by the motion compensation processing, inverse D in the inverse DCT processing
An addition process for adding the CT data and an image obtained by the inverse DCT process based on the picture type;
Or a switching process for outputting an image obtained by the addition process, wherein the inverse DCT process includes, when the magnitude of the motion vector is small, the low-frequency side of the DCT coefficients acquired by the Huffman decoding unit. Inverse DCT is performed using only the DCT coefficients of the above. When the magnitude of the motion vector is large, the inverse DCT is performed using the DCT coefficients of the high frequency side in addition to the DCT coefficients of the low frequency side.

A fifth invention is dependent on the fourth invention,
The recording medium on which the program is recorded is the DCT in the block.
Further comprising a coefficient counting process for dividing the coefficient into a plurality of regions and counting the number of DCT coefficients having a value other than 0 existing in each region, wherein the inverse DCT process is performed when the size of the motion vector is large. Inverse DCT is performed using the DCT coefficients in the region where the count value obtained by the coefficient counting process is large, in addition to the DCT coefficients on the low frequency side.

A sixth invention is dependent on the fourth invention,
The recording medium on which the program is recorded is the DCT in the block.
The coefficients are divided into a plurality of regions, and the DCT existing in each region
The method further includes a DCT coefficient absolute value sum calculation process for obtaining a sum of absolute values of the coefficients.
Inverse DCT is performed using the DCT coefficients in the region where the sum of absolute values obtained by the T-factor absolute value sum calculation processing is large.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) FIG. 1 is a block diagram showing a configuration of a digital video decoding device capable of image size conversion according to a first embodiment of the present invention. In FIG. 1, the digital video decoding apparatus includes a Huffman decoding unit 101, an inverse quantization unit 102, a DCT coefficient mask unit 103, an inverse DCT unit 104, a DCT coefficient control unit 105,
An accumulation unit 106, a motion compensation unit 107, an addition unit 108, a switch unit 109, a motion vector conversion unit 110, and an image size setting unit 111 are provided.

The Huffman decoding unit 101 receives a bit stream of a Huffman-encoded video signal. The Huffman decoding unit 101 performs Huffman decoding on the input bit stream to obtain information including a picture type, a DCT coefficient, a quantization table, and a motion vector. The inverse quantization unit 102 multiplies the DCT coefficient obtained by the Huffman decoding unit 101 by a constant using the quantization table obtained by the Huffman decoding unit 101.
The DCT coefficient mask unit 103 replaces a part of the DCT coefficient obtained by the inverse quantization unit 102 with 0. Inverse DCT unit 10
No. 4 performs inverse DCT using the DCT coefficient of which a part has been replaced by 0 in the DCT coefficient masking unit 103. The DCT coefficient control unit 105 selects a desired DCT coefficient according to the size of the motion vector and the preset size of the output image,
For the DCT coefficient mask unit 103, the DC to be replaced with 0
It designates a T coefficient and further controls a DCT coefficient used in inverse DCT section 104. The adder 108 is the inverse DCT 1
04 is a P or B picture, the motion compensation unit 107
And an output image is generated by adding the prediction signal generated in step (1).
The switch unit 109 switches the switch based on the picture type, and outputs the output of the inverse DCT unit 104 or the addition unit 1
08 is selected, and the selected output is output as an image signal.

The motion vector obtained by the Huffman decoding unit 101 is input to the motion vector conversion unit 110.
The motion vector conversion unit 110 converts the input motion vector into a motion vector corresponding to a preset output image size. The accumulation unit 106 accumulates the output image signal from the switch unit 109 for a predetermined time. The motion compensation unit 107 performs motion compensation based on the image signal stored in the storage unit 106 and the motion vector obtained from the motion vector conversion unit 110, generates a prediction signal, and outputs the prediction signal to the addition unit 108. . The image size setting unit 111
It outputs preset image size information to DCT coefficient control section 105 and motion vector conversion section 110.

Next, the operation of the digital video decoding apparatus configured as described above will be described with reference to FIGS. FIG. 2 is a schematic diagram showing a two-dimensional DCT coefficient arrangement of 8 × 8 size. Now, the output image size is set in advance to the original image size, and the horizontal and vertical sizes are set to 1 each.
/ 2 times. The DCT coefficient control unit 105 sets a threshold value for the size of the input motion vector, and when the image size is set to 倍 in the horizontal and vertical directions, if the size of the motion vector is less than the threshold value, For example, the DCT coefficient masking unit 103 is controlled so that only the DCT coefficients of the low-frequency 4 × 4 block area indicated by の in FIG. 2 are passed through, and the DCT coefficients of the other areas indicated by × are replaced with 0. On the other hand, when the magnitude of the motion vector is equal to or larger than the threshold, FIG.
The DCT coefficient masking unit 103 is controlled so as to pass all the 8 × 8 DCT coefficients in. Next, the inversely quantized DCT coefficient partially replaced with 0 is output to the inverse DCT section 104 based on the following equations (1) to (3).
T. If the picture type is an I picture,
The inverse DCT DCT coefficients are converted to pixel values. When the picture types are P and B pictures, the DCT coefficients subjected to inverse DCT are converted into pixel difference values.

(Equation 1)

(Equation 2)

(Equation 3) Here, f (j, k) in the above equation (1) is a transformed pixel value, and F (u, v) is a DCT coefficient before transformation. j and k indicate the horizontal and vertical positions of the pixel values in the 8 × 8 DCT block, and u and v indicate the horizontal and vertical positions of the DCT coefficients. When j and k increase the image size by, for example, 倍, j and k take values of 0, 2, 4, and 6, respectively. N in the above equation (1) is 1 of the block before conversion.
It means the number of DCT coefficients on the side. In the present embodiment, N
= 8. In this way, an image or a difference image of the size to be output is created using a part or all of the DCT coefficients constituting each block before conversion.

The motion vector converter 110 sets the size of the motion vector input from the Huffman decoding unit 101 to 1 in the horizontal and vertical directions by setting the output image size to 倍 each of the horizontal and vertical directions. / 2 times the motion compensation unit 107 and the DCT coefficient control unit 105
Output to The motion compensating unit 107 creates a prediction signal having a size １ ／ times as large in the horizontal and vertical directions based on the new motion vector converted by the motion vector converting unit 110, and outputs the signal to the adding unit 108. The pixel value converted in this way is output as an output image signal via the switch unit 109, and the pixel difference value is added to the prediction value generated by the motion compensation unit 107 in the addition unit 108. Are output as output image signals via the switch unit 109.

Here, the reason why the DCT coefficient is corrected by the DCT coefficient mask section as shown in FIG. 2 will be described. Normally, the inverse DCT uses 8 × 8 DCT coefficients per block to convert to 8 × 8 pixel values. In the present embodiment, in order to double the output image size by 水平 each in the horizontal and vertical directions, when performing inverse DCT, の 間 thinning is performed.
As shown in the above equation (1), the pixel value is converted into 4 × 4 pixel values.
At this time, if the DCT coefficient output from the inverse quantization unit 102 is used as it is, the image quality is impaired due to aliasing distortion. Therefore, the low frequency band 4 × represented by a circle in FIG.
By performing inverse DCT using only the DCT coefficients of the four regions, an output having characteristics similar to those obtained by filtering a frequency component equal to or less than half of the Nyquist frequency in each of the horizontal and vertical directions is obtained, and aliasing distortion is suppressed. I do. On the other hand, when a moving image is coded using DCT, the motion component tends to concentrate on the high frequency side of the DCT coefficient. Therefore, when decoding is performed using only the low frequency side of the DCT coefficient, the motion is intense. Image quality degradation occurs in some parts. Therefore, in a block in which the magnitude of the motion vector is small and the static portion or the motion can be determined to be small, as shown in FIG. 2A, only the low frequency DCT coefficient is used. On the other hand, in a block having a large motion vector and which can be determined as a moving image portion, as shown in FIG. 2B, inverse DCT is performed by using high-frequency DCT coefficients in addition to low-frequency DCT coefficients, and Suppress the image quality deterioration of the part.

As described above, according to the present embodiment, there is no significant deterioration in image quality even in a portion where movement is intense, and a simple configuration that does not require low-pass filtering and thinning-out after decoding. The image size of the output image can be changed. In the first embodiment described above, D
In the CT coefficient masking unit 103, when the magnitude of the motion vector is large, all the DCT coefficients are passed through. However, if the DCT coefficients of the high frequency band are used in addition to the low frequency 4 × 4 region, FIG. As shown in FIG.
The T coefficient may be passed through, and the right half DCT coefficient may be replaced with 0.

(Second Embodiment) FIG. 4 shows a second embodiment of the present invention.
It is a block diagram which shows the structure of the digital moving image decoding apparatus which can convert the image size concerning embodiment. The digital video decoding device in FIG. 4 differs from the digital video decoding device in FIG. 1 in that a coefficient counting unit 112 is further provided. Since other configurations are the same, in FIG. 4, components corresponding to those shown in FIG. 1 are denoted by the same reference numerals, and description thereof is omitted. The coefficient counting unit 112 divides the DCT coefficient in the block into a plurality of regions, and counts the number of DCT coefficients having a value other than 0 existing in each region. Further, coefficient counter section 112 outputs the number of DCT coefficients having a value other than 0 existing in each area to DCT coefficient control section 105.

Next, the operation of the digital video decoding apparatus configured as described above will be described with reference to FIGS. Now, it is assumed that the image size setting unit 111 has set the output image size to 水平 times the horizontal and vertical sizes of the original image size. Also, as shown in FIG. 5, it is assumed that the DCT coefficients in the block are divided into four 4 × 4 regions. The coefficient counter unit 112 counts the number of non-zero DCT coefficients present in the area shown in FIG. 5 and outputs each value to the DCT coefficient control unit 105. When the magnitude of the motion vector is smaller than the threshold, the DCT coefficient control unit 105 passes only the DCT coefficients of the region in FIG. 5 and replaces the DCT coefficients of the other regions with 0. On the other hand, if the magnitude of the motion vector is equal to or larger than the threshold,
In addition to the coefficients, the DCT coefficient masking unit 1 is configured to pass the DCT coefficients of the region having the largest count value from the coefficient counter unit 112 and replace the DCT coefficients of other regions with 0.
03 is controlled.

As described above, according to the second embodiment, a large image quality does not deteriorate even in a portion where movement is severe.
The image size of the output image can be changed with a simple configuration that does not require the low-pass filtering process and the decimation process after the decoding process.

(Third Embodiment) FIG. 6 shows a third embodiment of the present invention.
FIG. 10 is a block diagram illustrating a configuration of a digital video decoding device capable of image size conversion according to the embodiment. The digital video decoding device of FIG. 6 is different from that of FIG.
The difference is that a coefficient absolute value sum calculation unit 113 is further provided. Since other configurations are the same, in FIG.
Components corresponding to those in FIG. 1 are denoted by the same reference numerals, and description thereof will be omitted. DCT coefficient absolute value sum calculation unit 1
13 divides the DCT coefficient in the block into a plurality of regions, and obtains the sum of the absolute values of the DCT coefficients present in each region. The DCT coefficient absolute value sum calculation unit 113 calculates the DT coefficient in each region.
The absolute value sum of the CT coefficients is output to DCT coefficient control section 105.

Next, the operation of the digital video decoding apparatus configured as described above will be described with reference to FIGS. Now, it is assumed that the image size setting unit 111 has set the output image size to 水平 times the horizontal and vertical sizes of the original image size. Also, as shown in FIG. 5, it is assumed that the DCT coefficients in the block are divided into four 4 × 4 regions. The DCT coefficient absolute value sum calculation unit 113 calculates the sum of the absolute values of the DCT coefficients existing in the area shown in FIG. 5 and outputs the value to the DCT coefficient control unit 105. When the magnitude of the motion vector is smaller than the threshold, the DCT coefficient control unit 105 passes only the DCT coefficients of the region in FIG. 5 and replaces the DCT coefficients of the other regions with 0. On the other hand, when the magnitude of the motion vector is equal to or larger than the threshold, in addition to the DCT coefficients of the area in FIG. The DCT coefficient masking unit 103 is controlled so that the DCT coefficient of the

As described above, according to the present embodiment, a simple configuration that does not cause significant image quality deterioration even in a portion where movement is intense and that does not require low-pass filter processing and decimation processing after decoding processing. The image size of the output image can be changed.

In the first to third embodiments,
Although the image size was reduced by half for horizontal and vertical,
Other ratios of 1 to the power of 2 may be used. However, in this case, it is necessary to select the low-frequency DCT coefficients to be passed by the DCT coefficient masking unit 103 so as to obtain the optimal frequency characteristics. Further, in the second and third embodiments, the area of the DCT coefficient per block is divided into the four areas shown in FIG. 5, but other division methods may be used. In the second and third embodiments, when the magnitude of the motion vector is large, the low-frequency DCT
In addition to the coefficients, only the DCT coefficient of the area where the count value or the sum of absolute values in each area is the largest is passed. However, a threshold value is provided for the count value or the sum of absolute values, and the DCT coefficients of a plurality of areas exceeding the threshold value are passed. You may let it.

In the first to third embodiments, each component of the digital video decoding device capable of image size conversion has been described as being hardware, but the digital video decoding device capable of image size conversion has been described. It is also possible to replace all or a part of each of the components with software having the same functions as the corresponding functions of the hardware described above. In other words, all or some of the functions of each component of the digital video decoding device capable of converting the image size according to the first to third embodiments described above are implemented as a recording medium storing a program to be realized on a computer. Is also feasible.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a digital video decoding device capable of performing image size conversion according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram showing an arrangement of DCT coefficients used in the first embodiment.

FIG. 3 is a schematic diagram illustrating an arrangement of DCT coefficients used in the first embodiment.

FIG. 4 is a block diagram illustrating a configuration of a digital video decoding device capable of performing image size conversion according to a second embodiment of the present invention.

FIG. 5 is a schematic diagram showing DCT coefficient area division used in the second and third embodiments of the present invention.

FIG. 6 is a block diagram illustrating a configuration of a digital video decoding device capable of performing image size conversion according to a third embodiment of the present invention.

FIG. 7 is a block diagram showing a configuration of a conventional digital video decoding device capable of image size conversion.

[Explanation of symbols]

 Reference Signs List 101 Huffman decoding unit 102 Inverse quantization unit 103 DCT coefficient mask unit 104 Inverse DCT unit 105 DCT coefficient control unit 106 Storage unit 107 Motion compensation unit 108 Addition unit 109 Switch unit 110 Motion vector conversion unit 111 Image size setting unit 112 Coefficient counter Unit 113 DCT coefficient absolute value sum calculation unit

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C059 KK38 MA00 MA05 MA14 MA23 MC14 ME02 NN21 PP05 PP06 PP07 RC14 RC16 SS02 SS13 SS20 SS26 TA43 TA61 TB04 TC04 TC12 TC24 TC25 TD06 TD07 TD12 UA05 UA33 UA39 5J064 A16 BC02 BA08 BD03

Claims (6)

[Claims] 1. A digital video decoding apparatus capable of converting the size of an image to be output, comprising: a bit stream of an encoded digital video having at least information of a picture type and a DCT coefficient; Then, a Huffman decoding unit that performs Huffman decoding on the input bit stream and obtains information including a picture type, a DCT coefficient, a quantization table, and a motion vector, and outputs an image based on the input bit stream An image size setting unit in which the image size information is set, and when the input bit stream includes motion vector information, the Huffman based on the image size information set in the image size setting unit. A motion vector conversion unit for converting the size of the motion vector acquired by the decoding unit, and input bits An inverse quantization unit that inversely quantizes the DCT coefficient of the stream; a DCT coefficient mask unit that replaces a part of the DCT coefficient inversely quantized by the inverse quantization unit with 0; An inverse DCT unit for performing an inverse DCT on a DCT coefficient in which is replaced with 0, the DCT coefficient mask unit and the inverse DCT based on the magnitude of the motion vector information and the image size information.
A DCT coefficient control unit that controls the unit; a storage unit that stores at least an image including data that has been subjected to inverse DCT by the inverse DCT unit; a motion vector converted by the motion vector conversion unit; A motion compensation unit that generates a motion compensation image by using the obtained image, an addition unit that adds the motion compensated image generated by the motion compensation unit, and the data that has been inverse DCT by the inverse DCT unit, A switch unit that outputs an image from the inverse DCT unit or an image from the addition unit based on the picture type, wherein the inverse DCT unit performs the Huffman decoding when the magnitude of the motion vector is small. Performs inverse DCT using only the low-frequency side DCT coefficients among the DCT coefficients acquired by the section,
When the magnitude of the motion vector is large, the inverse DCT coefficient is used using the DCT coefficient on the high frequency side in addition to the DCT coefficient on the low frequency side.
Digital video decoding device capable of performing image size conversion. 2. The apparatus according to claim 1, further comprising: a coefficient counting unit that divides the DCT coefficient in the block into a plurality of regions and counts the number of DCT coefficients having a value other than 0 existing in each region. 2. The image according to claim 1, wherein when the magnitude of the motion vector is large, the inverse DCT is performed using the DCT coefficient in the area where the count value output from the coefficient counting unit is large, in addition to the DCT coefficient on the low frequency side. Digital video decoding device capable of size conversion. 3. A DCT coefficient absolute value sum calculation unit for dividing a DCT coefficient in a block into a plurality of regions and calculating a sum of absolute values of DCT coefficients present in each region, wherein the inverse DCT unit includes If the magnitude of the vector is large, the inverse DCT coefficient is calculated using the DCT coefficient in the area where the output of the DCT coefficient absolute value sum calculation unit is large, in addition to the DCT coefficient on the low frequency side.
The digital video decoding device capable of performing image size conversion according to claim 1, which performs CT. 4. A recording medium on which a program for realizing an environment for converting the size of an image to be output on a computer device is recorded, said medium having at least information of a picture type and a DCT coefficient and being encoded. Huffman decoding process of Huffman decoding the input bit stream and obtaining information including picture type, DCT coefficient, quantization table and motion vector, An image size setting process for setting image size information when outputting an image based on the input bit stream; and, when the input bit stream includes motion vector information, the image size setting process. Based on the set image size information, the motion vector obtained in the Huffman decoding process Motion vector transform processing for transforming the size of the input bit stream, inverse quantization processing for inversely quantizing the DCT coefficient of the input bit stream, and setting a part of the DCT coefficient inversely quantized in the inverse quantization processing to zero. DCT coefficient mask processing to be replaced, and D in which a part is replaced by 0 in the DCT coefficient mask processing.
An inverse DCT process for performing an inverse DCT of a CT coefficient; a magnitude of a motion vector acquired in the Huffman decoding process; and the DCT coefficient mask process and the image size information set in the image size setting process. A DCT coefficient control unit that controls a DCT coefficient handled in the inverse DCT processing, an accumulation processing that accumulates at least an image including data that has been subjected to the inverse DCT in the inverse DCT processing, and a motion vector converted in the motion vector conversion processing. A motion compensation process for generating a motion compensated image using the image accumulated in the accumulation process, and adding the motion compensated image generated in the motion compensation process and the data inverse DCT in the inverse DCT process And outputting the image obtained by the inverse DCT process or the image obtained by the addition process based on the picture type. And a switching processing, the inverse DCT processing, when the magnitude of the motion vector is small, among the DCT coefficients obtained by the Huffman decoding unit performs inverse DCT using only DCT coefficients of the low frequency side,
When the magnitude of the motion vector is large, the inverse DCT coefficient is used using the DCT coefficient on the high frequency side in addition to the DCT coefficient on the low frequency side.
A recording medium on which a program for performing T is recorded. 5. The apparatus according to claim 1, further comprising a coefficient counting process for dividing the DCT coefficient in the block into a plurality of regions and counting the number of DCT coefficients having a value other than 0 existing in each region. 5. The program according to claim 4, wherein when the magnitude of the motion vector is large, inverse DCT is performed using a DCT coefficient in a region where the count value obtained by the coefficient counting process is large, in addition to the DCT coefficient on the low frequency side. Recording medium on which is recorded. 6. A DCT coefficient absolute value sum calculation process for dividing a DCT coefficient in a block into a plurality of regions and obtaining a sum of absolute values of DCT coefficients present in each region, wherein the inverse DCT process includes The method according to claim 4, wherein when the magnitude of the vector is large, inverse DCT is performed using the DCT coefficient in a region where the absolute value sum obtained in the DCT coefficient absolute value calculation process is large, in addition to the DCT coefficient on the low frequency side. A recording medium on which the program described above is recorded.