JP2001204027A - Image information converter and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image information converter that can realize conversion of interlaced scanning MPEG2 image compression information into progressive scanning MPEG4 image compression information at a low cost in real time. SOLUTION: The image information converter is provided with an MPEG2 image information decoding section 18 that conducts decoding by using all of eight-order discrete cosine transform coefficients in a vertical direction and using only low frequency fourth-order coefficient information in a horizontal direction, a scanning conversion section 19 that converts interlaced scanning into progressive scanning, and an MPEG4 image information coding section 20 that generates MPEG4 image compression information from a signal after scanning conversion, and these sections 18, 19 and 20 are connected in series to each other. Furthermore, a picture type discrimination section 17 that discriminates a picture type of frame data in the interlaced scanning MPEG2 image compression information, outputs only frame data with respect to I/P pictures on the basis of the result of discrimination, and aborts the frame data with respect to the B picture to convert the frame rate is provided in a pre- stage of the MPEG2 image information decoder 18.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an MPEG (Moving
Image information (bit stream) compressed by orthogonal transform such as discrete cosine transform and motion compensation, such as Picture Image Coding Experts Group), is transmitted via network media such as satellite broadcasting, cable television (TV), and the Internet. The present invention relates to an apparatus and method for converting image information suitable for receiving on a storage medium or processing on a storage medium such as an optical or magnetic disk.

[0002]

2. Description of the Related Art In recent years, image information is handled as digital information. At this time, orthogonal transformation such as discrete cosine transformation or the like is performed by utilizing the redundancy inherent in the image information for the purpose of efficiently transmitting and storing information. An apparatus conforming to the MPEG system for compressing the image information by motion compensation is becoming widespread in both information distribution at a broadcasting station and information reception in ordinary households.

In particular, MPEG2 (ISO / IEC 13
818-2) is defined as a general-purpose image coding method, and is used as a standard covering both interlaced scan images and progressive scan images, and standard resolution images and high-definition images. It is expected to be used in the future. By using this MPEG2 compression method, for example, a standard resolution interlaced scan image having 720 × 480 pixels is 4 to 8 Mbps and 1920 × 10
1 for high-resolution interlaced scan images with 88 pixels
By assigning a code amount (bit rate) of 8 to 22 Mbps, it is possible to realize a high compression rate and good image quality.

[0004] Even in digital broadcasting, which is expected to be widely used in the future, image information is transmitted by such a compression method. Standards include standard resolution images and high resolution images. Therefore, it is desired that the receiver has a function of decoding both of them. In particular, in order to construct an inexpensive receiver while coexisting the standard resolution image and the high resolution image, it is necessary to reduce the image quality of the high resolution image information to some extent while minimizing the image quality deterioration. Processing is required. Such a problem is considered to occur not only in transmission media such as digital broadcasting, but also in storage media such as optical disks and flash memories.

In order to solve such a problem, the present applicant has previously proposed an image information decoding apparatus as shown in FIG. The code buffer 10 of the device shown in FIG.
1. The basic operation principle of the compression information analysis unit 102, the variable length decoding unit 103, the inverse quantization unit 104, the motion compensation units 107 and 108, the video memory 109, the adder 106, etc. Is equivalent to that of

In FIG. 2, input compressed image information is input to a compressed information analyzing unit 102 via a code buffer 101. The compression information analysis unit 102 analyzes information necessary for decompression from the input image compression information, and converts the image compression information together with the information obtained by the analysis into the variable length decoding unit 10.
Send to 3. The variable length decoding unit 103 performs a variable length decoding process corresponding to the variable length coding process at the time of encoding. However, in the variable length decoding unit 103 of this example, only the necessary coefficients are decoded in the reduced inverse discrete cosine transform unit (4 × 8) 105 in the subsequent stage, and the EOB (End Of Bloc)
It is also conceivable that no processing is performed until k) is detected. FIG. 3 shows an example of processing when the input MPEG2 image compression information (bit stream) is a zigzag scan (FIG. 3A) or an alternate scan (FIG. 3B). Is shown. The numbers in FIG. 3 indicate the order of scanning. The data obtained by the variable-length decoding process in the variable-length decoding unit 103 is subjected to an inverse quantization process in an inverse quantization unit 104, and then sent to a reduced inverse discrete cosine transform unit 105.

The discrete cosine transform coefficients of the eighth order in the vertical and horizontal directions obtained by the inverse quantization process in the inverse quantization unit 104 are reduced inverse discrete cosine transform units (4 × 8) 105
Performs an inverse discrete cosine transform process.

When the macroblock is an intra macroblock, the reduced inverse discrete cosine transform unit (4 ×
8) The data after the inverse discrete cosine transform in 105 is directly stored in the video memory 109 via the adder 106. On the other hand, in the case of an inter macroblock, the motion compensation unit (field prediction) 107 or the motion compensation unit (frame prediction) 108 depends on whether the motion compensation mode is the field prediction mode or the frame prediction mode.
Performs interpolation processing of 1/4 pixel precision in the horizontal direction and 1/2 pixel precision in the vertical direction based on the reference data in the video memory 109, and predictive pixel data generated by the interpolation processing and inverse discrete cosine transform. Is the adder 1
06 and sent to the video memory 109.

The pixel values stored in the video memory 109 are converted by an image frame conversion unit 110 to an image frame size suitable for a display device (not shown) at the subsequent stage. The image signal from the image frame conversion unit 110 becomes the output decoded image signal of the image information decoding device in FIG.

The operation principle of the reduced inverse discrete cosine transform unit (4 × 8) 105 will be described below.

Reduced inverse discrete cosine transform unit (4 × 8) 10
5 performs an 8th-order inverse discrete cosine transform in the vertical direction, similarly to a normal MPEG2 image information decoding apparatus.
On the other hand, in the horizontal direction, among the 8th-order discrete cosine transform coefficients, four low-frequency coefficients are extracted and subjected to a 4th-order inverse discrete cosine transform. FIG. 4 shows an equation representing this processing.
That is, in the equation shown in FIG. 4, the reduced inverse discrete cosine transform unit (4 × 8) 105 includes a matrix iD ₈ for executing an eighth-order discrete cosine transform (DCT) in the vertical (V) direction, A discrete cosine transform coefficient matrix C inversely quantized by the quantization unit 104 and a matrix i for performing a low-order fourth-order discrete cosine transform in the horizontal (H) direction
An output pixel value (difference value) X is obtained by multiplication with D ₄ ^t / O ₄ . O ₄ is a 4 × 4 zero matrix. Such an operation can be realized in both the horizontal and vertical directions by, for example, a high-speed algorithm described below.

FIG. 5 shows an example of the high-speed algorithm, for example, the algorithm of Wang (reference: Zhon
g de Wang., "Fast Algorithms fot the Discrete W Tr
ansform and for the Discrete Fourier Thransform ",
A configuration for realizing a fourth-order inverse discrete cosine transform based on IEEE Tr. ASSP-32, No. 4, pp. 803-816, Aug. 1984) will be described.

In FIG. 5, a low-frequency four coefficient F (0)
To F (3), the adder 121 calculates the coefficients F (0) and F (0).
(2) is added, and the adder 122 adds a coefficient F (0)
Is subtracted by adding the inverted coefficient F (2). The output of the adder 121 is multiplied by a coefficient A (A = 1 / √2) by a multiplier 123 and then added to the adders 133 and 1.
34. The output of the adder 122 is the multiplier 1
After the coefficient A is multiplied by 24, the adders 131 and 13
Sent to 2.

On the other hand, in the adder 125, subtraction is performed by adding the inverted coefficient F (1) to the coefficient F (3).
After being multiplied by (D = C _3/8 ), it is sent to the adder 130, and is inverted and sent to the adder 129.

The coefficient F (3) is multiplied by a coefficient B (B = −C _1/8 + C _3/8 ) by the multiplier 126 and then sent to the adder 129, where the coefficient F (1) is , Is multiplied by a coefficient C (C = C _1/8 + C _3/8 ) by the multiplier 127 and then sent to the adder 130.

The adder 129 performs subtraction by adding the inverted output of the multiplier 128 to the output of the multiplier 126. The adder 130 adds the output of the multiplier 127 and the output of the multiplier 128. Is done. The output of the adder 129 is sent to the adder 131 and is inverted to be added to the adder 132.
Sent to The output of the adder 130 is
3 and inverted and sent to the adder 134.

The adder 131 adds the output of the multiplier 124 and the output of the adder 129, and the adder 132 performs subtraction by adding the output of the multiplier 124 and the inverted output of the adder 129. Will be The adder 133 adds the output of the multiplier 123 and the output of the adder 130, and the adder 134 adds the output of the multiplier 123 to the output of the adder 1.
A subtraction is performed by adding the inverted output of 30.

The output of the adder 133 becomes the coefficient f (0) after the fourth-order inverse discrete cosine transform.
Is the coefficient f (1), and the output of the adder 132 is the coefficient f
(2) The output of the adder 134 is a coefficient f (3).

As described above, according to the configuration shown in FIG. 5, a quadratic inverse discrete cosine transform is realized by nine adders and five multipliers. C _3/8 is _calculated by the following formula (1)
Is represented by

C _3/8 = cos (3π / 8) (1) Next, the operation principle of the motion compensation unit (field prediction) 107 and the motion compensation unit (frame prediction) 108 will be described.

These motion compensation units (field prediction) 10
7 and the operation principle of the motion compensation unit (frame prediction) 108 are the same as those of a normal MPEG2 image information decoding apparatus in the vertical direction. However, in the horizontal direction, it is necessary to perform an interpolation process in the reduced inverse discrete cosine transform unit (4 × 8) 105 because the resolution is thinned out to １ ／.
That is, first, a pixel equivalent to 精度 precision is created by a double interpolation filter such as a half-band filter, and a pixel equivalent to ４ precision is created by linear interpolation based on the created pixels. At this time, when a pixel value having the same phase as a pixel extracted from the video memory 109 is output as a predicted image by using a half-band filter, there is no need to perform a multiply-accumulate operation according to the number of taps. Operation is possible. In addition, when a half-band filter is used, it is not necessary to perform a product-sum operation according to the number of taps, so that high-speed operation is possible. When a half-band filter is used, its coefficient is x / 2
^Since it can be expressed in the form of ⁿ , high-speed execution is possible by realizing division necessary for filtering by shift operation. Alternatively, it is also conceivable to directly create pixels required for motion compensation by filtering with a quadruple interpolation.

As the actual processing, coefficients are prepared in advance so that the two-stage interpolation realized by the double interpolation filter and the linear interpolation as described above is performed at once.
The process is performed as if it were a stage interrogation. Also, only necessary pixel values are created according to the value of the motion vector in the input image compression information (bit stream).

When performing the double interpolation filtering, it may be necessary to refer to the outside of the picture frame in the video memory 109 depending on the value of the motion vector. In this case, as shown in FIG. 6 (A), the required number of taps is symmetrically folded around the end point (hereinafter, this is called mirror processing), or as shown in FIG. 6 (B), The pixels having the same value as the pixel value of the end point are handled as if they exist outside the image frame by the required number of taps (this is hereinafter referred to as hold processing).

Next, the operation principle of the image frame conversion unit 110 will be described.

For example, when the image frame of the input image compression information (bit stream) is 1920 × 1080 pixels, the image output from the video memory 109 is 960.
That is to say, × 1080 pixels. This is for example 720
When outputting to a display device of × 480 pixels (aspect ratio 16: 9), 3/4 in the horizontal direction and 4/9 in the vertical direction
Thinning processing is required. The image frame conversion unit 110 converts the image frame by performing such a thinning process. When the image compression information to be input is obtained by interlaced scanning, as shown in FIG.
The thinning process is divided into a first field and a second field.

Here, the above-mentioned MPEG2 is mainly intended for high-quality encoding suitable for broadcasting,
It does not support a coding amount (bit rate) lower than that of PEG1, that is, a coding method with a higher compression rate. On the other hand, with the spread of portable terminals in recent years, it is expected that the need for an encoding system with such a high compression rate will increase in the future, and in response to this, the MPEG4 encoding system has been standardized. M
Regarding the PEG4 image coding method,
In May, the standard was approved as an international standard as ISO / IEC 14496-2.

By the way, MPEG2 image compression information (bit stream) once encoded for digital broadcasting
There is a need to convert MPEG4 image compression information (bit stream) with a lower code amount (bit rate), which is more suitable for processing on a mobile terminal or the like.

As an image information conversion device (transcoder) for achieving the above object, "Field-to-Frame Transco
ding with Spatial and Temporal Downsampling ”(Sus
ie J. Wee, John G. Apostolopoulos, and Nick Feamste
r, ICIP '99) has proposed an apparatus as shown in FIG.

In FIG. 8, the data of each frame in the MPEG2 image compression information (bit stream) of the interlaced scanning to be input is first input to the picture type determination unit 111.

In the picture type discriminating section 111, the input data of each frame is related to an I picture (intra-coded picture) / P picture (forward predicted coded picture) or a B picture (bidirectional predicted coded picture). Then, only in the former case, information on the I / P picture is output to the subsequent MPEG2 image information decoding unit (I / P picture) 112.

The processing in the MPEG2 image information decoding unit (I / P picture) 112 is the same as that of a normal MPEG2 image information decoding device. However, since the data relating to the B picture is discarded by the picture type discrimination unit 111, the MPEG2 image information decoding unit (I / P picture)
The function at 112 is only required to be able to decode only I / P pictures. MPEG2 image information decoding unit (I
/ P picture) 112 is input to the thinning unit 113.

The thinning section 113 performs a half thinning process in the horizontal direction, leaves only one of the first and second fields in the vertical direction, and discards the other. , A progressively scanned image having a size of １ ／ of the input image information is generated. The progressive scan image generated by the thinning unit 113 is an MPEG4 image information encoding unit (I / P-VOP)
Enter 114.

The MPEG4 image information encoding unit (I / P
-VOP) 114 encodes the input signal of the sequentially scanned image to generate and output MPEG4 image compression information (bit stream).

At this time, the motion vector information in the input MPEG2 image compression information (bit stream) is mapped to a motion vector for the decimated image information in the motion vector synthesis unit 115, and the motion vector detection unit 116 A high-precision motion vector is detected based on the motion vector value synthesized by the motion vector synthesis unit 115.

In MPEG4, VOP (Vide
o Object Plane) represents an area composed of one or more macroblocks surrounding an object. The VOP area is classified into one of an I picture, a P picture, and a B picture according to a coding scheme. I-VOP (VOP of I picture)
Is an image (area) itself encoded (intra-encoded) without performing motion compensation. P-VO
P (VOP of P picture) is basically forward predictive coded based on an image (I or P-VOP) located earlier in time than itself. A B-VOP (a VOP of a B picture) is basically bi-directionally predictive coded based on two pictures (I or P-VOP) temporally located before and after itself.

[0036]

The image information converter shown in FIG. 8 can convert MPEG2 image compression information (bit stream) into MPEG4 image compression information (bit stream).

However, in the image information conversion apparatus shown in FIG. 8, the amount of arithmetic processing in the MPEG2 image information decoding unit (I / P picture) 112 increases, and the video memory (109) for storing image information. For example, the capacity is large, which hinders the construction of an inexpensive device when implemented by a dedicated LSI, and hinders real-time operation when implemented by a general-purpose processor, for example.

Accordingly, the present invention has been made in view of such circumstances, and is, for example, an interlaced scanning MPEG.
It is an object of the present invention to provide an image information conversion apparatus and method capable of converting two-image compression information (bit stream) into, for example, progressively scanned MPEG4 image compression information (bit stream) at low cost and in real time. I do.

[0039]

According to the present invention, there is provided an image information conversion apparatus for converting first image compression information of interlaced scanning, which has been compressed by orthogonal transformation processing and motion compensation processing, to a higher level than the first image compression information. An image information conversion device that converts the image data into second image compression information of progressive scanning, which becomes a compression ratio, and outputs the information;
Of the nth-order orthogonal transform coefficients in both the vertical and horizontal directions in the first image compression information, n in the vertical direction
The decoding process is performed using all of the following orthogonal transform coefficients, and in the horizontal direction, the decoding process is performed using only low-order m (m <n) -order orthogonal transform coefficients. First image information decoding means for performing decoding processing of the above, scan conversion means for converting interlaced scanning information output from the first image information decoding means into sequential scanning information, and information after the scan conversion To generate the second image compression information from the second
The above-mentioned problem is solved by having the image information encoding means.

In the image information conversion apparatus according to the present invention, the first image compression information is MPEG2 image compression information comprising 8th order discrete cosine transform coefficients in both the vertical and horizontal directions. The conversion means uses all of the 8th-order discrete cosine transform coefficients in the vertical direction and the 8th-order discrete cosine transform coefficients in the vertical and horizontal directions, and outputs only the information of the low-frequency fourth-order coefficients in the horizontal direction. The second image information encoding unit is an MPEG2 image information encoding unit that generates MPEG4 image compression information from the output after the scan conversion.

According to the image information conversion method of the present invention, the first image compression information of the interlaced scanning compressed by the orthogonal transformation processing and the motion compensation processing has a higher compression ratio than the first image compression information. This is an image information conversion method for converting the image data into progressively scanned second image compression information and outputting the converted image information. Of the n-th orthogonal transformation coefficients in the vertical and horizontal directions in the first image compression information, The decoding process is performed using all of the n-th order orthogonal transform coefficients, and the decoding process is performed using only the m-th (m <n) order orthogonal transform coefficients in the horizontal direction in the horizontal direction. By performing information decoding processing, converting the interlaced scanning information after decoding processing of the first image information into information of progressive scanning, and generating second image compression information from the information after scanning conversion, Solving the above problems .

In the image information conversion method of the present invention, the first image compression information is composed of 8th-order discrete cosine transform coefficients in both the vertical and horizontal directions, and is used for decoding the first image information. Used all of the 8th-order discrete cosine transform coefficients in the vertical direction and the 8th-order discrete cosine transform coefficients in the vertical and horizontal directions, and used only information of the low-frequency 4th-order coefficients in the horizontal direction. MPEG
(2) The decoding process is performed, and in the process of generating the second image compression information, the MPEG4 image compression information is generated from the output after the scan conversion.

In the image information conversion apparatus and method of the present invention, the encoding type of each frame data in the interlaced MPEG2 image compression information is determined, and based on the determination result, the intra-coded image / forward prediction is performed. The frame rate is converted by outputting only the frame data relating to the encoded image and discarding the frame data relating to the bidirectional predictive encoded image.

That is, according to the present invention, the input M
In the PEG2 image compression information (bit stream), I / O
The frame rate is converted by discarding the B-picture while leaving only the P-picture information, and the information on the I / P picture after the frame rate conversion is applied to the horizontal direction of the low-order 8th-order DCT coefficient among the DCT coefficients. 4
Partial decoding is performed using only the next-order coefficient information and all the eighth-order information in the vertical direction. After that, of the pixel values to be output after decoding the MPEG2 image information, only the data of the first field or the second field is left, and the rest is discarded to convert the data into sequential scanning. The output is encoded into MPEG4 image compression information (bit stream). Further, according to the present invention, based on a motion vector value in input image compression information (bit stream) detected at the time of decoding of MPEG2 image information,
Mapping is performed on a motion vector value for the image data after scan conversion, and a highly accurate motion vector is detected based on the motion vector value.

[0045]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a schematic configuration of an image information conversion apparatus according to an embodiment of the present invention.

In FIG. 1, the MPEG2 image compression information (bit stream) of the interlaced scanning to be input is
First, it is input to the picture type discriminating unit 17.

The picture type discriminating unit 17 determines whether the I / P
The information about the picture is output and sent to the MPEG2 image information decoding unit (I / P picture 4 × 8 down decoder) 18, but the information about the B picture is discarded. As a result, the frame rate is converted.

The MPEG2 image information decoding section (I / P picture 4 × 8 down decoder) 18 is equivalent to the apparatus shown in FIG. 2, but information relating to B pictures has already been discarded by the preceding picture type discriminating section 17. Therefore, the function only needs to be a function capable of performing decoding processing of only information on I / P pictures. Further, in the case of the present embodiment, the MPEG2 image information decoding unit (I / P picture 4 × 8 down decoder) 18 is, as in the image information decoding apparatus shown in FIG. Decoding processing using only coefficient information (DCT coefficients) is performed. Thereby, the MPE of FIG.
The capacity of the video memory required by the G2 image information decoding unit (I / P picture 4 × 8 down decoder) 18 is the same as the MPEG2 image information decoding unit (I / P
(P picture) 112 may be used, and the amount of processing required for the inverse discrete cosine transform may be １ ／.

The interlaced scan pixel data output from the MPEG2 image information decoder (I / P picture 4 × 8 down decoder) 18 is input to the scan converter 19. The scan converter 19 converts the interlaced scan pixel data into sequential scan pixel data by leaving only one of the first field and the second field and discarding the other, and outputs the converted data.

The pixel data of this progressive scanning is MPEG4
It is input to an image information encoding unit (I / P-VOP) 20.
The MPEG4 image information encoding unit (I / P-VOP) 2
0, the signal of the input progressively scanned image is encoded and MP
It generates and outputs EG4 image compression information (bit stream).

The MPEG2 image information decoding unit (I /
The motion vector information in the input MPEG2 image compression information (bit stream) detected by the P picture 4 × 8 down decoder 18 is input to the motion vector synthesizing unit 21 where the sequentially converted image after scan conversion is input. To the motion vector value at. Further, in the motion vector detecting unit 22, the motion vector synthesizing unit 2
High-precision motion detection is performed based on a motion vector value in a sequentially converted image after scan conversion which is an output of 1.

As described above, the image information conversion apparatus according to the present embodiment decodes interlaced MPEG2 image compression information using all of the eighth-order discrete cosine transform coefficients in the vertical direction. In the horizontal direction, decoding processing is performed using only the information of the low-order fourth-order coefficient, and after the decoding processing, conversion is performed from interlaced scanning to sequential scanning, and MPEG4 image compression information is generated from the output after the scan conversion. MP of interlaced scan input
Conversion of EG2 image compression information (bit stream) to progressively scanned MPEG4 image compression information (bit stream) can be realized at low cost and in real time.

In the above description, MPE is used as an input.
Using G2 image compression information (bit stream) as an example, output MPEG4 image compression information (bit stream)
However, the present invention is not limited to both input and output, and may be, for example, MPEG1 or H.264. H.263 or other image compression information (bit stream), and such image information can also be converted.

[0055]

As is apparent from the above description, in the present invention, decoding processing is performed using all of the nth-order orthogonal transform coefficients in the vertical direction, and m (m <N) A decoding process using only the next orthogonal transform coefficient is performed to decode the first image compression information, and then the information of the interlaced scanning is converted into information of the sequential scanning. By converting the MPEG2 image compression information (bit stream) of interlaced scanning into the MPEG4 image compression information (bit stream) of progressive scanning, a circuit using a smaller amount of processing and a smaller amount of video memory can be used. The configuration realizes inexpensive and real-time image information conversion.

[Brief description of the drawings]

FIG. 1 is a block circuit diagram illustrating a schematic configuration of an image information conversion device according to an embodiment of the present invention.

FIG. 2 shows an image information decoding apparatus (4 × 8 down decoder) that uses all information in the vertical direction but performs decoding using only fourth-order low-frequency information in the horizontal direction. FIG. 3 is a block circuit diagram illustrating a configuration example.

FIG. 3 is a diagram used to explain the operation principle of a variable length decoding unit of the device shown in FIG. 2;

FIG. 4 is a diagram used to explain the operation principle of the reduced inverse discrete cosine transform unit (4 × 8) 5 of the device shown in FIG. 2;

FIG. 5 is a diagram showing an example of a configuration for realizing a fourth-order inverse discrete cosine transform using the Wang high-speed algorithm in the device shown in FIG. 2;

FIG. 6 is a diagram used to explain mirror processing and hold processing in a motion compensation unit (field prediction) and a motion compensation unit (frame prediction) of the device shown in FIG. 2;

FIG. 7 is a diagram used to explain vertical processing in an image frame conversion unit of the apparatus shown in FIG. 2;

FIG. 8 shows MPEG2 image compression information (bit stream).
FIG. 1 is a block circuit diagram showing a conventional configuration example of an image information conversion device (transcoder) that receives MPEG as an input and outputs MPEG4 image compression information (bit stream).

[Explanation of symbols]

17 picture type discriminating section, 18 MPEG2 image information decoding section (I / P picture 4 × 8 down decoder), 19 scan conversion section, 20 MPEG4 image information encoding section (I / P-VOP), 21 motion vector synthesis section , 22 Motion vector detector

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Zoo Ewen 6-35, Kita-Shinagawa, Shinagawa-ku, Tokyo Inside Sony Corporation (72) Inventor Teruhiko Suzuki 6-35, Kita-Shinagawa, Shinagawa-ku, Tokyo Inside Sony Corporation (72) Inventor Yoichi Yagasaki 6-7-35 Kita Shinagawa, Shinagawa-ku, Tokyo Sony Corporation F-term (reference) 5C059 KK11 LA00 LB18 MA00 NN01 NN21 PP05 PP06 SS01 SS11 UA02 UA05 5J064 AA03 AA04 BA16 BB03 BC12 BD02

Claims (32)

[Claims] 1. An interlaced first image compression information compressed by an orthogonal transformation process and a motion compensation process is converted into a second image of a progressive scan having a higher compression ratio than the first image compression information. In the image information conversion device for converting into compressed information and outputting the compressed information, among the n-th orthogonal transform coefficients in the vertical and horizontal directions in the first image compressed information, n
The decoding process is performed using all of the following orthogonal transform coefficients, and in the horizontal direction, the decoding process is performed using only low-order m (m <n) -order orthogonal transform coefficients. First image information decoding means for performing decoding processing of the following; scanning conversion means for converting interlaced scanning information output from the first image information decoding means into sequential scanning information; and information after the scan conversion. And a second image information encoding unit for generating second image compression information from the image information conversion device. 2. The image compression information according to claim 1, wherein the first image compression information is an MPT comprising an 8th-order discrete cosine transform coefficient in both the vertical and horizontal directions.
EG2 image compression information, wherein the first image information decoding means uses all the 8th-order discrete cosine transform coefficients in the vertical direction among the 8th-order discrete cosine transform coefficients in both the vertical and horizontal directions, In the horizontal direction, M using only information of the low-order fourth-order coefficient is used.
MPEG2 image information decoding means for performing PEG2 decoding processing, wherein the second image information encoding means is MPEG4 image information encoding means for generating MPEG4 image compression information from the information after the scan conversion. The image information conversion device according to claim 1, wherein 3. The encoding type of each frame in the MPEG2 image compression information of the interlaced scanning is determined, and based on the determination result, only information on the intra-coded image / forward-predicted coded image is output, and bidirectional prediction is performed. 3. A picture type discriminating means for converting a frame rate by discarding information on an encoded image, and an output of said picture type discriminating means is input to said MPEG2 image information decoding means.
The image information conversion device described in the above. 4. The MPEG2 image information decoding means,
4. The image information conversion apparatus according to claim 3, further comprising a function of performing decoding processing of only the intra-coded image / forward prediction-coded image. 5. The MPEG2 image information decoding means comprises a variable length decoding means, wherein the variable length decoding means changes only the discrete cosine transform coefficients of the eighth order in the vertical direction and the fourth order in the low frequency in the horizontal direction. 3. The image information conversion device according to claim 2, wherein long decoding is performed. 6. The MPEG2 image information decoding means includes a reduced inverse discrete cosine transform means, and the reduced inverse discrete cosine transform means performs a process using all the octal discrete cosine transform coefficients in the vertical direction. 3. The image information conversion apparatus according to claim 2, wherein only the low-frequency fourth-order coefficient in the direction is extracted, and a fourth-order inverse discrete cosine transform is applied to the low-frequency fourth-order coefficient. 7. The image information conversion apparatus according to claim 6, wherein an inverse discrete cosine transform is executed in both the horizontal and vertical directions by a method based on a predetermined high-speed algorithm. 8. The MPEG2 image information decoding means includes a motion compensation means, and the motion compensation means has a 1/4 pixel precision in the horizontal direction according to a value of a motion vector in the input MPEG2 image compression information. 3. The image information conversion apparatus according to claim 2, wherein the interpolation processing is performed. 9. The motion compensation unit according to claim 1, wherein the interpolation is performed at a half pixel accuracy using a digital filter of a two-fold interpolation, and then the interpolation is performed at a quarter pixel accuracy by linear interpolation. 9. The image information conversion device according to claim 8, wherein: 10. The image information conversion apparatus according to claim 9, wherein said motion compensating means uses a half-band filter as the digital filter for the double interpolation. 11. The MPEG2 image information decoding means has a storage means for storing a pixel value, and the motion compensation means calculates in advance a coefficient equivalent to a series of interpolation processing, and inputs an MPEG2 image as an input. 10. The image coasting information conversion device according to claim 9, wherein a motion compensation process using the coefficient is performed on the pixel value extracted from the storage unit according to a value of a motion vector in the compression information. 12. When the motion compensation means needs a pixel value outside the image frame to perform double interpolation, the motion compensation means performs necessary processing according to the number of filter taps by mirror processing or hold processing. 10. The image information conversion apparatus according to claim 9, wherein the motion compensation processing is performed by virtually creating the pixel values of the image information. 13. The scanning conversion means according to claim 1, wherein one of the first field and the second field of the interlaced image information output from the MPEG2 image information decoding means is stored and the other is discarded. 3. The image information conversion apparatus according to claim 2, wherein conversion from interlaced scanning to sequential scanning is performed. 14. The MPEG4 image information encoding means has a function of encoding only an area composed of one or more macroblocks surrounding an object of an intra-coded image / a forward prediction-coded image. The image information conversion device according to claim 2, wherein: 15. A motion vector synthesizing means for synthesizing a motion vector value corresponding to pixel data after scan conversion based on motion vector information in MPEG2 image compression information to be input. The image information conversion device described in the above. 16. The image information conversion device according to claim 15, further comprising a motion vector detecting means for performing highly accurate motion vector detection based on the motion vector value generated by the motion vector synthesizing means. 17. An interlaced first image compression information compressed by an orthogonal transformation process and a motion compensation process is converted into a progressively scanned second image having a higher compression ratio than the first image compression information. In the image information conversion method of converting into compressed information and outputting the compressed information, among the n-th orthogonal transform coefficients in the vertical and horizontal directions in the first image compressed information, n
The decoding process is performed using all of the following orthogonal transform coefficients, and in the horizontal direction, the decoding process is performed using only low-order m (m <n) -order orthogonal transform coefficients. And converting the interlaced scanning information after the decoding processing of the first image information into the progressive scanning information, and generating second image compression information from the information after the scan conversion. Image information conversion method. 18. The first image compression information comprises an 8th-order discrete cosine transform coefficient in both the vertical and horizontal directions. When decoding the first image information, the first and second image compression information is 8 bits in both the vertical and horizontal directions. Among the next discrete cosine transform coefficients, MPEG2 decoding processing using all information of the eighth-order discrete cosine transform coefficients in the vertical direction and only information of the low-order fourth-order coefficients in the horizontal direction is performed. 18. The image information conversion method according to claim 17, wherein in the image compression information generation processing, MPEG4 image compression information is generated from the output after the scan conversion. 19. A coding type of each frame in the MPEG2 image compression information of the interlaced scanning is determined, and based on a result of the determination, only information on an intra-coded image / a forward prediction coded image is output, and bidirectional prediction is performed. 19. The image information conversion method according to claim 18, wherein the frame rate is converted by discarding information on the coded image, and the MPEG4 image information is generated from the output after the frame rate conversion. 20. The image information conversion method according to claim 19, wherein, at the time of decoding the MPEG2 image information, decoding processing is performed only on the intra-coded image / forward predicted coded image. 21. In the variable length decoding at the time of decoding the MPEG2 image information, only the discrete cosine transform coefficients of the eighth order in the vertical direction and the fourth order in the low frequency band in the horizontal direction are variable length decoded. An image information conversion method according to claim 18. 22. In the reduced inverse discrete cosine transform at the time of decoding the MPEG2 image information, processing is performed using all the 8th-order discrete cosine transform coefficients in the vertical direction, and only the low-frequency 4th-order coefficient in the horizontal direction is obtained. 2. A low-order fourth-order coefficient is subjected to a fourth-order inverse discrete cosine transform.
8. The image information conversion method according to item 8. 23. The image information conversion method according to claim 22, wherein said inverse discrete cosine transform is executed by a method based on a predetermined high-speed algorithm in both the horizontal and vertical directions. 24. In the motion compensation at the time of decoding the MPEG2 image information, an input MPE is input in the horizontal direction.
1/4 according to the value of the motion vector in the G2 image compression information
19. A pixel precision interpolation process is performed.
The described image information conversion method. 25. At the time of the motion compensation, first, a half-pixel precision interpolation is performed using a double interpolation digital filter, and then a quarter-pixel precision interpolation is performed by linear interpolation. The image information conversion method according to claim 24, characterized in that: 26. The image information conversion method according to claim 25, wherein at the time of said motion compensation, interpolation using a half-band filter is performed as a digital filter of said double interpolation. 27. In the decoding processing of the MPEG2 image compression information, an image value is stored. In the motion compensation, a coefficient equivalent to a series of interpolation processing is calculated in advance, and the MPEG2 image compression information to be input is calculated. 26. The image coasting information conversion method according to claim 25, wherein the stored pixel value is subjected to a motion compensation process using the coefficient according to the value of the motion vector in. 28. In the above motion compensation, when a pixel value existing outside an image frame is required to perform double interpolation, mirror processing or hold processing is used to perform the necessary processing in accordance with the number of filter taps. 26. The image information conversion method according to claim 25, wherein the motion compensation processing is performed by virtually creating pixel values for the minutes. 29. At the time of the scan conversion, one of a first field and a second field of interlaced image information which is an output after decoding of MPEG2 image information is stored, and the other is stored. 19. The image information conversion method according to claim 18, wherein conversion from interlaced scanning to sequential scanning is performed by discarding. 30. A function of encoding only an area composed of one or a plurality of macroblocks surrounding an object of an intra-coded image / a forward prediction-coded image when encoding the MPEG4 image information. 19. The image information conversion method according to claim 18, comprising: 31. The image information conversion method according to claim 18, wherein a motion vector value corresponding to the pixel data after scan conversion is synthesized based on the motion vector information in the input MPEG2 image compression information. . 32. The image information conversion method according to claim 31, wherein highly accurate motion vector detection is performed based on the synthesized motion vector value.