JP5226089B2 - Image coding method converter - Google Patents

Description

  The present invention relates to an image coding method conversion apparatus that converts, for example, a JPEG video stream into an MPEG-2 video stream.

In recent years, a technique for compressing and encoding an image is generally used.
A representative example of this image encoding method is JPEG (Joint Photographic Experts Group) mainly used in digital cameras and the like.
There is also a method of encoding a moving image by continuously encoding JPEG. This method is sometimes called Motion JPEG or MJPEG.
Furthermore, there is a moving picture encoding method called MPEG-2 (Moving Picture Experts Group) adopted in DVD (Digital Verifiable Disk) -VIDEO.

As described above, there are various image encoding methods, and generally different image encoding methods are not compatible.
For this reason, in order for a user to support different image encoding methods, it is necessary to prepare an image decoding device corresponding to each image encoding method or to prepare an image encoding method conversion device.
For example, a commercially available DVD player can reproduce only the MPEG-2 system, and it is practically difficult to prepare an image decoding apparatus corresponding to each image encoding system.
On the other hand, if an image encoding method conversion device is prepared, it is not necessary to prepare an image decoding device corresponding to each image encoding method, but the time required for image conversion and the quality of the image after conversion are problematic. Become.

As the simplest conversion method in the image encoding method conversion apparatus, there is a method in which encoded data is once decoded and returned to image data, and then encoded again by a different encoding method.
However, this conversion method has the disadvantages that the amount of calculation processing is enormous and the image quality of the image tends to deteriorate.
Therefore, in order to suppress an increase in the amount of calculation processing and deterioration in image quality, an image coding method conversion device that converts an image coding method without completely decoding encoded data has been developed. The apparatus is disclosed in the following non-patent document 1.
However, in the image coding method conversion apparatus disclosed in Non-Patent Document 1 below, it is necessary to perform quantization and inverse quantization of encoded data.

An Application Level Video Gateway, Elan Amir, Steven McCanne, Hui Zhang, ACM Multimedia 95, 1995

  Since the conventional image coding method conversion apparatus is configured as described above, the image coding method is converted without completely decoding the encoded data. Can be prevented. However, since it is necessary to carry out quantization and inverse quantization of encoded data, there are problems such as limitations in reducing the amount of calculation processing and suppressing deterioration in image quality.

  The present invention has been made to solve the above-described problems, and provides an image coding method conversion apparatus that can greatly reduce the amount of arithmetic processing and can greatly suppress deterioration in image quality. For the purpose.

The image coding method conversion apparatus according to the present invention decodes a part of an encoded stream in which an image signal is encoded by the first encoding method, and generates header information and a first code from the encoded stream. A simple decoding means for extracting the quantization table and quantization frequency component of the encoding method, and a second encoding method for the quantization table and quantization frequency component of the first encoding method extracted by the simple decoding means The image data of the additional image is quantized using the quantization table and the data conversion means for converting to the quantization frequency component, and the quantization table of the second encoding method converted by the data conversion means, and after the quantization The image data adding means for adding the image data to the quantized frequency component of the second encoding method converted by the data converting means, and the header information extracted by the simple decoding means as the additional image The second information is changed according to the image width, the header information after the change, the quantization table of the second encoding method converted by the data conversion means, and the image data after the quantization is added by the image data adding means Coding stream creating means for creating a coded stream of the second coding method from the quantized frequency component of the coding method.
Further, the quantization table of the first encoding method extracted this time by the simple decoding means is compared with the quantization table of the first encoding method previously extracted by the simple decoding means, and further simplified decoding is performed. The quantization frequency component of the first encoding method output this time from the encoding means and the quantization frequency component of the first encoding method output last time from the simple decoding means are compared, and two comparison results are obtained. Provided with an encoding determination unit that determines whether or not to perform intra-frame encoding based on, and if the encoded stream creation unit receives a determination result indicating that intra-frame encoding is to be performed from the encoding determination unit, Intra-frame encoding is performed to create an encoded stream of the second encoding method, and if a determination result indicating that intra-frame encoding is not performed is received from the encoding determination means, intra-frame encoding is performed. Z , In which to output the encoded stream of the second coding scheme previously created.

According to this invention, a part of the encoded stream in which the image signal is encoded by the first encoding method is decoded, and the header information and the quantization table of the first encoding method are decoded from the encoded stream. And a simple decoding means for extracting the quantized frequency component, a quantization table for the first encoding method and a quantization frequency component extracted by the simple decoding means for the quantization table and the quantum for the second encoding method. Data conversion means for converting to a quantized frequency component, and image data of the additional image is quantized using the quantization table of the second encoding method converted by the data conversion means, and the image data after quantization is converted to data conversion means The image data adding means to be added to the quantized frequency component of the second encoding method converted by the above and the header information extracted by the simple decoding means are changed according to the image width of the additional image. Quantization frequency component of the second encoding method in which the header information after the change, the quantization table of the second encoding method converted by the data conversion means, and the image data after quantization by the image data adding means are added Since it is configured to include an encoded stream creating means for creating an encoded stream of the second encoding method from the above, it is possible to greatly reduce the amount of calculation processing and to greatly suppress deterioration in image quality There is an effect that can.
Further, the quantization table of the first encoding method extracted this time by the simple decoding means is compared with the quantization table of the first encoding method previously extracted by the simple decoding means, and further simplified decoding is performed. The quantization frequency component of the first encoding method output this time from the encoding means and the quantization frequency component of the first encoding method output last time from the simple decoding means are compared, and two comparison results are obtained. Provided with an encoding determination unit that determines whether or not to perform intra-frame encoding based on, and if the encoded stream creation unit receives a determination result indicating that intra-frame encoding is to be performed from the encoding determination unit, Intra-frame encoding is performed to create an encoded stream of the second encoding method, and if a determination result indicating that intra-frame encoding is not performed is received from the encoding determination means, intra-frame encoding is performed. Z , Since it is configured to output a coded stream of the second coding scheme previously created, there is an effect that can greatly reduce the amount of code.

It is a block diagram which shows the image coding system converter by Embodiment 1 of this invention. It is a block diagram which shows the image coding system converter by Embodiment 2 of this invention. It is explanatory drawing which shows the image data with which the image was added to the right end. It is explanatory drawing which shows the image data with which the image was added to the left end. It is explanatory drawing which shows the image data with which the image was added to the right end and the left end. It is a block diagram which shows the image coding system converter by Embodiment 3 of this invention. (A) represents the character “1” by 16 pixels × 32 lines (example drawn in two MBs in the vertical direction), and (b) represents the block unit comprising the character “1” consisting of a plurality of MBs. It is explanatory drawing represented by. It is explanatory drawing which shows an example of the additional image on which time information is displayed. It is explanatory drawing which shows an example of the image to which time information was added to the right end. It is a block diagram which shows the image coding system converter by Embodiment 4 of this invention. It is a block diagram which shows the image coding system converter by Embodiment 5 of this invention. It is a block diagram which shows the image coding system converter by Embodiment 6 of this invention. It is a flowchart which shows the processing content of the skip determination part 11 of the image coding system converter by Embodiment 6 of this invention. It is a block diagram which shows the image coding system converter by Embodiment 7 of this invention. It is a flowchart which shows the processing content of the MB type determination part 14 of the image coding system converter by Embodiment 7 of this invention. It is a block diagram which shows the image coding system converter by Embodiment 8 of this invention. It is explanatory drawing which shows a color difference format. It is explanatory drawing which shows an example of the image by which time information was added to the JPEG image conversion area.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing an image coding method conversion apparatus according to Embodiment 1 of the present invention. In FIG. 1, a JPEG simple decoding unit 1 encodes an image signal using JPEG (first coding method). A part of the JPEG stream (encoded stream) is decoded, header information, JPEG quantization table (quantization table of the first encoding method), and JPEG quantized DCT (Discrete Cosine Transform) coefficient from the JPEG stream A process of extracting (quantization frequency component of the first encoding method) is performed. The JPEG simple decoding unit 1 constitutes simple decoding means.

The data conversion unit 2 converts the JPEG quantization table and JPEG quantization DCT coefficients extracted by the JPEG simple decoding unit 1 into MPEG-2 quantization table parameters (quantization table of the second encoding method) and MPEG-2 quantization. A process of converting to a quantized DCT coefficient (quantized frequency component of the second encoding method) is performed. The data conversion unit 2 constitutes data conversion means.
The MPEG-2 stream creation unit 3 uses the MPEG-2 stream (from the header information extracted by the JPEG simple decoding unit 1, the MPEG-2 quantization table parameter converted by the data conversion unit 2, and the MPEG-2 quantization DCT coefficient ( A process of creating an encoded stream of the second encoding method is performed. The MPEG-2 stream creation unit 3 constitutes an encoded stream creation means.

Next, the operation will be described.
First, JPEG and MPEG-2 encoding will be briefly described.
JPEG performs DCT conversion of an image signal (pixel data) of an image, and quantizes DCT coefficients that are image signals of DCT conversion.
In JPEG, power can be concentrated in a low frequency region by DCT conversion of an image signal, and an image can be compressed by thinning out high frequency components by quantizing DCT coefficients.
In JPEG, the compression rate is further increased by performing run length coding, variable length coding, etc. of DCT coefficients.
On the other hand, MPEG-2 performs compression similar to JPEG, but is different from JPEG in that it is compressed in the time direction, and is different from JPEG in terms of encoding format.
In the first embodiment, all intra-frame encoding is performed without using inter-frame encoding in MPEG-2.
Hereinafter, the processing content of the image coding method conversion apparatus of FIG. 1 will be described in detail.

When a JPEG stream in which an image signal is encoded by JPEG is input, the JPEG simple decoding unit 1 decodes a part of the JPEG stream, and header information, a JPEG quantization table, and a JPEG quantization DCT from the JPEG stream. Extract coefficients.
That is, the JPEG simple decoding unit 1 extracts header information from the JPEG stream by decoding header information, Huffman code, and the like in the JPEG stream, and outputs the header information to the MPEG-2 stream creation unit 3. The JPEG quantization table and the JPEG quantization DCT coefficient are extracted from the JPEG stream, and the JPEG quantization table and the JPEG quantization DCT coefficient are output to the data conversion unit 2.

When the data conversion unit 2 receives the JPEG quantization table and the JPEG quantization DCT coefficient from the JPEG simple decoding unit 1, the data conversion unit 2 converts the JPEG quantization table into an MPEG-2 quantization table parameter, and the JPEG quantization DCT coefficient Are converted to MPEG-2 quantized DCT coefficients.
Hereinafter, a conversion algorithm in the data conversion unit 2 will be described.

ただし、ＪｐｅｇＤｅｑｕａｎｔＤＣＴ_i,jはＪＰＥＧ量子化ＤＣＴ係数が逆量子化されたＤＣＴ係数、ＪｐｅｇＱｕａｎｔＴａｂｌｅ_i,jはＪＰＥＧ量子化テーブル、ＪｐｅｇＱｕａｎｔＤＣＴ_i,jはＪＰＥＧ量子化ＤＣＴ係数である。
ｉ，ｊは８×８の各成分を表している。 First, the JPEG quantized DCT coefficient in JPEG encoding is inversely quantized by the following equation (1).

However, JpegQuantDCT _{i, j} is a DCT coefficient obtained by dequantizing the JPEG quantized DCT coefficient, JpegQuantTable _{i, j} is a JPEG quantization table, and JpegQuantDCT _{i, j} is a JPEG quantized DCT coefficient.
i and j represent 8 × 8 components.

ただし、ＭｐｅｇＤｅｑｕａｎｔＤＣＴ_i,jはＭＰＥＧ−２量子化ＤＣＴ係数が逆りょうしかされたＤＣＴ係数、ＭｐｅｇＱｕａｎｔＴａｂｌｅ_i,jはＭＰＥＧ−２量子化テーブル・パラメータ、ＭｐｅｇＱｕａｎｔＤＣＴ_i,jはＭＰＥＧ−２量子化ＤＣＴ係数、ＤｃＭｕｌｔはイントラＤＣ成分用量子化パラメータ、ＱｕａｎｔＳｃａｌｅは量子化スケールコードである。 On the other hand, in an intra MB (Macro Block) in MPEG-2 encoding, the MPEG-2 quantized DCT coefficient is composed of a DC (Direct Current) component with i = j = 0 and other AC (Alternating Current) components. Differently, the inverse quantization is performed by the following equation (2).

Where MpegQuantDCT _{i, j} is the DCT coefficient obtained by reversing the MPEG-2 quantized DCT coefficient, MpegQuantTable _{i, j} is the MPEG-2 quantization table parameter, MpegQuantDCT _{i, j} is the MPEG-2 quantized DCT coefficient, DcMult is an intra DC component quantization parameter, and QuantScale is a quantization scale code.

ただし、ＤｃＭｕｌｔはＭＰＥＧ−２の規定の範囲内の任意の値でよい。
式（３）〜（５）から分る通り、逆量子化及び量子化の処理を直流成分だけ行えばよく、演算量が削減されていることが理解される。 From the JPEG and MPEG-2 inverse quantization calculation formulas in the above equations (1) and (2), the intra MB may be encoded in MPEG-2 as shown below.

However, DcMult may be any value within the range specified in MPEG-2.
As can be seen from the equations (3) to (5), it is understood that the inverse quantization and the quantization process need only be performed on the DC component, and the amount of calculation is reduced.

When the JPEG simple decoding unit 1 extracts the header information and the data conversion unit 2 outputs the MPEG-2 quantization table parameter and the MPEG-2 quantization DCT coefficient, the MPEG-2 stream creation unit 3 An MPEG-2 stream is created from the MPEG-2 quantization table parameters and MPEG-2 quantized DCT coefficients.
In particular, MpegQuantTable _{i, j} may be encoded by being multiplexed on the MPEG-2 picture header portion, and DcMulti may be encoded when MPEG-2 VLC (Variable Length Coding) is performed.

As is apparent from the above, according to the first embodiment, a part of a JPEG stream in which an image signal is encoded by JPEG is decoded, and header information, a JPEG quantization table, and a JPEG quantum are decoded from the JPEG stream. JPEG simple decoding unit 1 for extracting the quantized DCT coefficients, and the JPEG quantization table and the JPEG quantized DCT coefficients extracted by the JPEG simple decoding unit 1 into MPEG-2 quantization table parameters and MPEG-2 quantized DCT coefficients A data conversion unit 2 for conversion, and MPEG-2 quantization table parameters and MPEG-2 quantization DCT coefficients converted by the data conversion unit 2 from the header information extracted by the JPEG simple decoding unit 1 and MPEG-2 Since it is configured to create a stream, the amount of computation processing can be greatly reduced. It is, the effect that it is possible to greatly suppress deterioration of image quality.
That is, according to the first embodiment, when the image encoding method is converted from JPEG to MPEG-2, the inverse quantization process and the quantization process associated with the conversion need only be performed once for one DCT block. The amount of computation can be greatly reduced, and image quality deterioration due to requantization processing can be suppressed.

  In the first embodiment, the image encoding method is converted from JPEG to MPEG-2. However, a similar conversion method may be used to convert from MPEG-2 I-picture to JPEG. it can. It can also be applied to conversion between other coding schemes using quantization and DCT.

Embodiment 2. FIG.
FIG. 2 is a block diagram showing an image coding method conversion apparatus according to Embodiment 2 of the present invention. In the figure, the same reference numerals as those in FIG.
The image data adding unit 4 uses the MPEG-2 quantization table parameter converted by the data converting unit 2 to add image data (here, the image data of the added image has already been converted to DCT coefficients). In addition to the MPEG-2 quantized DCT coefficient obtained by converting the quantized image data by the data converter 2, an additional MPEG-2 quantized DCT coefficient (MPEG-2 quantized DCT coefficient + (Quantized image data) is output to the MPEG-2 stream creation unit 5 and header change information indicating the image width of the additional image is output to the MPEG-2 stream creation unit 5. The image data adding unit 4 constitutes image data adding means.
The MPEG-2 stream creation unit 5 changes the header information extracted by the simple decoding unit 1 according to the header change information output from the image data addition unit 4 and is converted by the changed header information and the data conversion unit 2. The MPEG-2 stream is generated from the MPEG-2 quantization table parameter and the additional MPEG-2 quantized DCT coefficient output from the image data adding unit 4. The MPEG-2 stream creation unit 5 constitutes an encoded stream creation means.

In the second embodiment, a method for enlarging the width of a converted MPEG-2 image using image data of an additional image is shown.
MPEG-2 is used, for example, in a video encoding system in a DVD (Digital Versatile Disk) -VIDEO system.
In this DVD-VIDEO system, resolution is further limited as compared with MPEG-2. For example, the width is limited to 704 pixels × height 480 lines.
However, in general JPEG, the ratio of width to height is 4: 3. For example, it is recorded with a width of 640 pixels × a height of 480 lines.
By using the second embodiment, the width of the image can be enlarged at high speed.

Hereinafter, the processing content of the image coding method conversion apparatus of FIG. 2 will be described in detail.
When a JPEG stream in which an image signal is encoded with JPEG is input, the JPEG simple decoding unit 1 decodes a part of the JPEG stream and outputs header information and header information from the JPEG stream, as in the first embodiment. Extract the JPEG quantization table and JPEG quantization DCT coefficients.

  When the data conversion unit 2 receives the JPEG quantization table and the JPEG quantization DCT coefficient from the JPEG simple decoding unit 1, as in the first embodiment, the data conversion unit 2 converts the JPEG quantization table into an MPEG-2 quantization table parameter. And converting the JPEG quantized DCT coefficient into an MPEG-2 quantized DCT coefficient.

Upon receiving the MPEG-2 quantization table parameter from the data conversion unit 2, the image data adding unit 4 quantizes the image data of the additional image using the MPEG-2 quantization table parameter.
Further, the image data adding unit 4 adds the added MPEG-2 quantized DCT coefficient (MPEG-2 quantized DCT coefficient) in addition to the MPEG-2 quantized DCT coefficient obtained by converting the quantized image data by the data converting unit 2. Coefficient + quantized image data) is output to the MPEG-2 stream creation unit 5, and header change information indicating the image width of the additional image is output to the MPEG-2 stream creation unit 5.

Hereinafter, a method of creating an additional MPEG-2 quantized DCT coefficient will be described.
For example, when an additional image of 64 pixels × 480 lines is added to an image of 640 pixels × 480 lines to create an image of 704 pixels × 480 lines, as shown in FIG. 3, 640 pixels × 480 lines In addition to the form of adding an additional image of 64 pixels × 480 lines to the right end of the image, as shown in FIG. 4, the form of adding an additional image of 64 pixels × 480 lines to the left end of the image of 640 pixels × 480 lines, As shown in Fig. 5, an additional image of 32 pixels x 480 lines is added to the right end of an image of 640 pixels x 480 lines, and an additional image of 32 pixels x 480 lines is added to the left end of an image of 640 pixels x 480 lines The form to do is conceivable.
In any form, since the area of the image to be added is a multiple of 8, it can be considered in units of blocks.

The additional image may be image data of any image, such as an image painted with a single color or an image with some pattern or character, and may be image data of an image input by the user. May be.
Here, it is assumed that the image data of the additional image has already been converted into the DCT coefficient, but the image data adding unit 4 may convert the image data of the additional image into the DCT coefficient.
Therefore, it is assumed that image data of a predetermined image is used, or image data of an image input by a user before the start of conversion is used.

ただし、ＡｄｄＤＣＴ_i,jは追加画像の画像データ、ＡｄｄＱｕａｎｔＤＣＴ_i,jは量子化された追加画像の画像データである。 The image data addition unit 4 quantizes the image data of the additional image using the MPEG-2 quantization table parameter output from the data conversion unit 2 as shown in the following equation (6).

_However, AddDCT _{i, j} is the image data of the additional _image, AddQuantDCT _{i, j} is the image data of the additional image quantized.

As described above, the image data adding unit 4 adds the image data of the quantized additional image to the additional MPEG-2 quantized DCT coefficient (in addition to the MPEG-2 quantized DCT coefficient converted by the data converting unit 2). (MPEG-2 quantized DCT coefficient + quantized image data) is output to the MPEG-2 stream creation unit 5 and header change information indicating the image width of the additional image is output to the MPEG-2 stream creation unit 5.
If image data of a predetermined additional image is DCT-converted and held, only the quantization process needs to be performed, and an image can be added at high speed.

  When the MPEG-2 stream creation unit 5 receives header information from the JPEG simple decoding unit 1 and receives an MPEG-2 quantization table parameter, an additional MPEG-2 quantization DCT coefficient, and header change information from the data conversion unit 2, The information about the image width included in the header information is changed according to the header change information, and the MPEG-2 stream is changed from the changed header information, MPEG-2 quantization table parameters, and additional MPEG-2 quantization DCT coefficients. create.

  As apparent from the above, according to the second embodiment, the image data of the additional image is quantized using the MPEG-2 quantization table parameter converted by the data conversion unit 2, and the image data after quantization is quantized. In addition to the MPEG-2 quantized DCT coefficient converted by the data converting unit 2, an additional MPEG-2 quantized DCT coefficient (MPEG-2 quantized DCT coefficient + quantized image data) is created as an MPEG-2 stream. Since the header change information indicating the image width of the additional image is output to the MPEG-2 stream creation unit 5 as well as being output to the unit 5, the same effect as in the first embodiment can be obtained, and at the same time There is an effect that the image size can be changed. As a result, the degree of freedom in creating a DVD-VIDEO can be increased.

Embodiment 3 FIG.
6 is a block diagram showing an image coding system conversion apparatus according to Embodiment 3 of the present invention. In the figure, the same reference numerals as those in FIG.
The additional block image database 6 stores image data of an additional image corresponding to time information in advance (the image data of the additional image is image data in units of MPEG-2 MB (macroblock) and converted into DCT coefficients). Is a memory storing
When the time information is input, the additional image creation unit 7 refers to the additional block image database 6 to acquire image data of the additional image corresponding to the time information, and sends the image data of the additional image to the image data addition unit 4. Output.
The additional block image database 6 and the additional image creation unit 7 constitute additional image creation means.

Next, the operation will be described.
When the time information is input, the additional image creation unit 7 refers to the additional block image database 6 to acquire image data of the additional image corresponding to the time information, and the image data of the additional image is added to the image data addition unit 4. Output to.
This time information may be obtained from anywhere, for example, the current time may be input from a clock. Alternatively, the first JPEG image to be converted may be set to 0 and incremented according to JPEG input. Alternatively, the creation date and time of the input JPEG recorded electronically may be input.

For example, when time information of 21:00:14 on April 1, 2006 is obtained, the image data of an additional image necessary for displaying this time information from the additional block image database 6 is in MB units of MPEG-2 The DCT coefficients of the image data “2”, “0”, “6”, “4”, “1”, “year”, “month”, “day”, “hour”, and “minute” depicted in FIG.
Here, FIG. 7A shows the character “1” by 16 pixels × 32 lines (example drawn in two vertical MBs), and FIG. 7B shows the character “1”. It is explanatory drawing represented by the block unit which consists of several MB.
Since the additional block image database 6 stores MB-unit image data that has been converted into DCT coefficients, the additional image creation unit 7 converts the image data of the additional image into an image without performing DCT conversion on the image data of the additional image. The data can be output to the data adding unit 4.

  When receiving the image data of the additional image from the additional image creating unit 7, the image data adding unit 4 uses the MPEG-2 quantization table parameter converted by the data converting unit 2 as in the second embodiment. The image data of the additional image is quantized, and the quantized image data is added to the MPEG-2 quantized DCT coefficient converted by the data converting unit 2, and the additional MPEG-2 quantized DCT coefficient (MPEG-2 quantized DCT) Coefficient + quantized image data) is output to the MPEG-2 stream creation unit 5, and header change information indicating the image width of the additional image is output to the MPEG-2 stream creation unit 5.

In the third embodiment, as shown in FIG. 7A, an example is shown in which characters are drawn with two MBs in the vertical direction, but characters may be drawn with three or more MBs. Further, as shown in FIG. 7B, characters may be drawn not in units of MB but in units of blocks.
Moreover, characters other than numbers, such as characters and patterns, may be drawn.
Here, FIG. 8 is an explanatory diagram illustrating an example of an additional image in which time information is displayed, and FIG. 9 is an explanatory diagram illustrating an example of an image in which time information is added to the right end.

As is apparent from the above, according to the third embodiment, the additional block image database 6 is referred to acquire the image data of the additional image corresponding to the time information, and the image data of the additional image is added to the image data. Since it was comprised so that it might output to the part 4, there exists an effect which can add the image corresponding to time information.
As a result, the degree of freedom in creating a DVD-VIDEO can be increased, and the user can check time information or enjoy a pattern that is periodically replaced while viewing the created DVD-VIDEO. be able to.

Embodiment 4 FIG.
10 is a block diagram showing an image coding system conversion apparatus according to Embodiment 4 of the present invention. In the figure, the same reference numerals as those in FIG.
The additional block image database 8 is image data of an additional image corresponding to header information (information included in the header information in advance, for example, information such as the model number, manufacturer name, and time information of the camera used for photographing). (The image data of the additional image is image data in MB units of MPEG-2 and converted into DCT coefficients).
The additional image creation unit 9 refers to the additional block image database 8 to acquire image data of the additional image corresponding to the header information extracted by the simple decoding unit 1, and uses the image data of the additional image as the image data addition unit. 4 is output.
The additional block image database 8 and the additional image creation unit 9 constitute additional image creation means.

  In the third embodiment, the additional image creation unit 7 refers to the additional block image database 6 to acquire the image data of the additional image corresponding to the time information, and sends the image data of the additional image to the image data adding unit 4. Although what is output is shown, the additional image creation unit 9 may acquire the image data of the additional image corresponding to the header information and output the image data of the additional image to the image data adding unit 4.

That is, since the header information extracted by the simple decoding unit 1 includes various information such as the model number, manufacturer name, and time information of the camera used for shooting, it corresponds to the header information in advance. The image data of the additional image is recorded in the additional block image database 8.
Upon receiving the header information from the simple decoding unit 1, the additional image creation unit 9 refers to the additional block image database 8 and acquires image data of the additional image corresponding to the header information, and the image data of the additional image Is output to the image data adding unit 4.
As a result, the degree of freedom in creating a DVD-VIDEO can be increased, and the user can check the header information or enjoy a pattern that is periodically replaced while viewing the created DVD-VIDEO. There is an effect that can be.

In the third and fourth embodiments, the image data adding unit 4 adds the additional image to the additional region (see FIG. 9). However, as shown in FIG. 18, the image data adding unit 4 adds the additional image. May be added to the JPEG image conversion area.
FIG. 18 shows an example in which a part of the JPEG image conversion area is painted black and white time information is added.
Note that when adding an additional image to the JPEG image conversion area, for example, if the additional image of only the luminance component is added without particularly processing the color difference component, the processing speed can be increased.
When an additional image having only a luminance component is added to the JPEG image conversion area, the color difference component of the JPEG image remains, and the JPEG image can be seen through.

Embodiment 5 FIG.
FIG. 11 is a block diagram showing an image coding system conversion apparatus according to Embodiment 5 of the present invention. In the figure, the same reference numerals as those in FIG.
The DCT coefficient reduction unit 10 sets the high-order JPEG quantized DCT coefficient to “0” among the JPEG quantized DCT coefficients extracted by the simple decoding unit 1 and deletes the high-order JPEG quantized DCT coefficient To implement. The DCT coefficient reduction unit 10 constitutes a high frequency component deletion unit.
FIG. 11 shows an example in which the DCT coefficient reduction unit 10 is applied to the image coding method conversion apparatus in FIG. 1, but the DCT coefficient reduction is applied to the image coding method conversion apparatus in FIGS. 2, 6, and 10. Part 10 may be applied.

Next, the operation will be described.
In general, in an image, even if high frequency components are reduced, deterioration is often not conspicuous. Therefore, in the fifth embodiment, a high-order JPEG quantization DCT coefficient is set to “0” and high-order JPEG quantization is performed. By deleting the DCT coefficient, the code amount is reduced and the compression rate is increased.

That is, when the simple decoding unit 1 extracts the JPEG quantized DCT coefficient, the DCT coefficient reducing unit 10 sets the higher-order JPEG quantized DCT coefficient to “0” among the JPEG quantized DCT coefficients, Delete the JPEG quantized DCT coefficients.
Specifically, the number of JPEG quantized DCT coefficients that are not “0” is counted in order from the lower-order JPEG quantized DCT coefficients, and the number of JPEG quantized DCT coefficients that are not “0” exceeds the number specified in advance. At this stage, it is conceivable to set the subsequent higher-order JPEG quantized DCT coefficients to “0”.
In addition, when a JPEG quantized DCT coefficient having a higher order than a predesignated order is set to “0”, or when a JPEG quantized DCT coefficient of “0” continues for a predetermined number or more, A method of setting subsequent high-order JPEG quantized DCT coefficients to “0” is conceivable.
As a method for examining the two-dimensional JPEG quantized DCT coefficients in order from the lowest order, a method using the run-length encoded data used for JPEG encoding as it is can be used.
The DCT coefficient reduction unit 10 outputs the JPEG quantized DCT coefficient after the deletion process (the JPEG quantized DCT coefficient from which the higher-order JPEG quantized DCT coefficient has been deleted) to the data conversion unit 2.

  As is apparent from the above, according to the fifth embodiment, among the JPEG quantized DCT coefficients extracted by the simple decoding unit 1, the higher order JPEG quantized DCT coefficients are set to “0”, and higher order Since the configuration is such that the JPEG quantized DCT coefficients are deleted, the code amount can be reduced and the compression rate can be increased.

Embodiment 6 FIG.
FIG. 12 is a block diagram showing an image coding method conversion apparatus according to Embodiment 6 of the present invention. In the figure, the same reference numerals as those in FIG.
The skip determination unit 11 uses the JPEG quantization table extracted this time by the simple decoding unit 1 and the JPEG quantization table stored in the reference data buffer 12 (the JPEG quantization table previously extracted by the simple decoding unit 1). The JPEG quantized DCT coefficient extracted this time by the simple decoding unit 1 and the JPEG quantized DCT coefficient stored in the reference data buffer 12 (the JPEG quantized DCT coefficient previously extracted by the simple decoding unit 1) ) And a process of determining whether or not to perform intraframe coding based on the comparison result. However, when the skip determination unit 11 receives an intra-frame encoding determination flag indicating that it is necessary to perform intra-frame encoding from the MPEG-2 stream creation unit 13, the JPEG quantization table and the JPEG quantization DCT coefficient A determination result indicating that intra-frame encoding is to be performed is output without performing the comparison process.
The reference data buffer 12 is a memory that holds the JPEG quantization table and the JPEG quantized DCT coefficient extracted this time by the simple decoding unit 1 for the next comparison processing in the skip determination unit 11.
The skip determination unit 11 and the reference data buffer 12 constitute an encoding determination unit.

  The MPEG-2 stream creation unit 13 sets the intra-frame coding determination flag to “1” when the frame or MB to be encoded corresponds to the head of a GOP or slice in MPEG-2, and otherwise. If the determination result indicating that the intra-frame encoding is performed is received from the skip determination unit 11 while the intra-frame encoding determination flag is set to “0”, the intra-frame encoding is performed and the MPEG-2 stream If the determination result indicating that the intra-frame encoding is not performed is received from the skip determination unit 11, the process of outputting the previously generated MPEG-2 stream is performed without performing the intra-frame encoding. The MPEG-2 stream creation unit 13 constitutes an encoded stream creation means.

FIG. 12 shows an example in which the skip determination unit 11 and the MPEG-2 stream creation unit 13 are applied to the image coding method conversion apparatus of FIG. 1, but FIG. 2, FIG. 6, FIG. 10, and FIG. The skip determination unit 11 and the MPEG-2 stream creation unit 13 may be applied to the image coding method conversion apparatus.
FIG. 13 is a flowchart showing the processing contents of the skip determination unit 11 of the image coding method conversion apparatus according to the sixth embodiment of the present invention.

Next, the operation will be described.
When a JPEG stream in which an image signal is encoded with JPEG is input, the JPEG simple decoding unit 1 decodes a part of the JPEG stream and outputs header information and header information from the JPEG stream, as in the first embodiment. Extract the JPEG quantization table and JPEG quantization DCT coefficients.

When the JPEG simple decoding unit 1 extracts the header information from the JPEG stream, the MPEG-2 stream creation unit 13 selects “1” when the encoding target frame or MB hits the head of the GOP or the slice in MPEG-2. The intra-frame encoding determination flag is output to the skip determination unit 11, and when the encoding target frame or MB does not hit the head of the GOP or the slice in MPEG-2, the intra-frame encoding determination of “0” is performed. The flag is output to the skip determination unit 11.
In general, in a moving picture coding system that performs compression in the time direction, it is determined to forcibly perform intra-frame coding in order to prevent error propagation. In MPEG-2, this is the head of a GOP or the head of a slice.

When the simple decoding unit 1 extracts the JPEG quantization table and JPEG quantization DCT coefficient of the current frame, the skip determination unit 11 extracts the JPEG quantization table and JPEG quantization DCT coefficient of the current frame from the simple decoding unit 1. Is input (step ST1), and an intra-frame coding determination flag is input from the MPEG-2 stream creation unit 13 (step ST2).
If the intra-frame encoding determination flag input from the MPEG-2 stream creation unit 13 is “1” (step ST3), the skip determination unit 11 determines that it is necessary to perform intra-frame encoding, and the frame A determination result indicating that the inner encoding is to be performed is output to the MPEG-2 stream creation unit 13, and the JPEG quantization table and JPEG quantization DCT coefficient of the current frame are output to the data conversion unit 2 (step ST4). .
The skip determination unit 11 outputs the JPEG quantization table and JPEG quantization DCT coefficient of the current frame to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the frame of the previous frame). The JPEG quantization table and JPEG quantized DCT coefficient) are updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST9).

On the other hand, if the intra-frame coding determination flag input from the MPEG-2 stream creation unit 13 is “0” (step ST3), the reference data stored in the reference data buffer 12 (JPEG of the previous frame) The quantization table and JPEG quantization DCT coefficient) are input (step ST5).
The skip determination unit 11 compares the JPEG quantization table of the current frame with the JPEG quantization table of the previous frame, and also compares the JPEG quantization DCT coefficient of the current frame with the previous frame. The JPEG quantized DCT coefficient is compared (step ST6).

Hereinafter, the comparison process in the skip determination unit 11 will be specifically described.
When the skip determination unit 11 compares the JPEG quantized DCT coefficient of the current frame with the JPEG quantized DCT coefficient of the previous frame, for example, as shown in Equation (7) below, The JPEG quantization DCT coefficient is divided by the JPEG quantization table of the current frame, the JPEG quantization DCT coefficient of the previous frame is divided by the JPEG quantization table of the previous frame, and both division results It is determined whether or not the sum of differences is smaller than a threshold value.

ただし、Ｗ_i,jは成分毎の重み、ｃｕｒＱｕａｎｔＤＣＴ_i,jは現在のフレームのＪＰＥＧ量子化ＤＣＴ係数、ｒｅｆＱｕａｎｔＤＣＴ_i,jは１つ前のフレームのＪＰＥＧ量子化ＤＣＴ係数、ｃｕｒＱｕａｎｔＴａｂｌｅ_i,jは現在のフレームのＪＰＥＧ量子化テーブル、ｒｅｆＱｕａｎｔＤＣＴ_i,jは１つ前のフレームのＪＰＥＧ量子化テーブル、Ｔｈ_skip1はスキップ判定用の閾値である。

Where W _{i, j} is the weight for each component, curQuantDCT _{i, j} is the JPEG quantized DCT coefficient of the current frame, refQuantDCT _{i, j} is the JPEG quantized DCT coefficient of the previous frame, and curQuantTable _{i, j} is the current JPEG quantization table of the frame of the second frame, refQuantDCT _{i, j} is the JPEG quantization table of the previous frame, and Th _skip1 is a threshold value for skip determination.

If the above equation (7) is not satisfied and the left side of the equation (7) is equal to or greater than the threshold Th _skip1 (step ST7), the skip determination unit 11 determines the previous JPEG quantized DCT coefficient and the previous one. It is determined that the difference between the JPEG quantized DCT coefficient of the frame is large and it is necessary to perform intraframe encoding, and a determination result indicating that intraframe encoding is to be performed is output to the MPEG-2 stream creation unit 13. Then, the JPEG quantization table and JPEG quantized DCT coefficient of the current frame are output to the data conversion unit 2 (step ST4).
The skip determination unit 11 outputs the JPEG quantization table and JPEG quantization DCT coefficient of the current frame to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the frame of the previous frame). The JPEG quantization table and JPEG quantized DCT coefficient) are updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST9).

In addition, the skip determination unit 11 compares the JPEG quantization table of the current frame with the JPEG quantization table of the previous frame, and if both JPEG quantization tables are not the same and are different from each other. It is determined that the difference is large (step ST7), and a determination result indicating that intraframe encoding is to be performed is output to the MPEG-2 stream creation unit 13, and the JPEG quantization table and JPEG quantization DCT coefficient of the current frame are output. Is output to the data converter 2 (step ST4).
The skip determination unit 11 outputs the JPEG quantization table and JPEG quantization DCT coefficient of the current frame to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the frame of the previous frame). The JPEG quantization table and JPEG quantized DCT coefficient) are updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST9).

ただし、Ｔｈ_skip2はスキップ判定用の閾値である。 If the JPEG quantization table of the current frame is the same as the JPEG quantization table of the previous frame, the skip determination unit 11 determines the JPEG of the current frame as shown in the following equation (8), for example. It is determined whether or not the sum of differences between the quantization table and the JPEG quantization table of the previous frame is smaller than a threshold value.

However, Th _skip2 is a threshold value for skip determination.

If the above equation (8) does not hold and the left side of equation (8) is _greater than or equal to the threshold Th _skip2 (step ST7), the skip determination unit 11 determines the previous JPEG quantized DCT coefficient and the previous one. It is determined that the difference between the JPEG quantized DCT coefficient of the frame is large and it is necessary to perform intraframe encoding, and a determination result indicating that intraframe encoding is to be performed is output to the MPEG-2 stream creation unit 13. Then, the JPEG quantization table and JPEG quantized DCT coefficient of the current frame are output to the data conversion unit 2 (step ST4).
The skip determination unit 11 outputs the JPEG quantization table and JPEG quantization DCT coefficient of the current frame to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the frame of the previous frame). The JPEG quantization table and JPEG quantized DCT coefficient) are updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST9).

The skip determination unit 11 determines that it is not necessary to perform intraframe encoding when the above equation (7) is satisfied and equation (8) is satisfied, and indicates that the intraframe encoding is not performed. The determination result shown is output to the MPEG-2 stream creation unit 13 (step ST8). In this case, since the data conversion unit 2 does not need to perform encoding method conversion, the JPEG quantization table and JPEG quantization DCT coefficient of the current frame are not output to the data conversion unit 2.
The skip determination unit 11 outputs the JPEG quantization table and JPEG quantization DCT coefficient of the current frame to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the frame of the previous frame). The JPEG quantization table and JPEG quantized DCT coefficient) are updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST9).

When the data conversion unit 2 receives the JPEG quantization table and JPEG quantization DCT coefficient of the current frame from the skip determination unit 11, the data conversion unit 2 converts the JPEG quantization table into the MPEG-2 quantization in the same manner as in the first embodiment. Convert to table parameters and convert the JPEG quantized DCT coefficients to MPEG-2 quantized DCT coefficients.
However, in the sixth embodiment, a non-intra MB may be handled. In the case of a non-intra MB, the following formula (9) is used instead of the above formula (2), and the formula (4) Instead, the conversion is performed using the following equation (10).

When the MPEG-2 stream creation unit 13 receives the determination result indicating that the intra-frame coding is to be performed from the skip determination unit 11, the MPEG-2 stream creation unit 13 performs the intra-frame coding in the same manner as in the first embodiment, thereby An MPEG-2 stream is created from the header information output from the decoding unit 1, the MPEG-2 quantization table parameter output from the data conversion unit 2, and the MPEG-2 quantized DCT coefficient.
On the other hand, when receiving a determination result indicating that the intra-frame encoding is not performed from the skip determination unit 11, the previously generated MPEG-2 stream is output without performing the intra-frame encoding.

  As is apparent from the above, according to the sixth embodiment, the skip determination unit 11 determines whether or not to perform intra-frame encoding, and the MPEG-2 stream creation unit 13 receives the intra-frame encoding from the skip determination unit 11. If a determination result indicating that encoding is to be performed is received, intra-frame encoding is performed to create an MPEG-2 stream, and if a determination result indicating that intra-frame encoding is not to be performed is received from the skip determination unit 11 Since the configuration is such that the previously created MPEG-2 stream is output without performing intra-frame coding, there is an effect that the amount of codes can be greatly reduced. Further, by performing skip MB determination in the DCT coefficient region, it is possible to keep the amount of calculation low.

In the sixth embodiment, the conversion from JPEG to MPEG-2 is shown. However, it is also possible to convert from MPEG-2 I-picture to JPEG by the same method. Also, the present invention can be applied to conversion between other encoding methods using quantization or DCT.
In the sixth embodiment, since the converted format is MPEG-2, the unit for determining the skip is MB, but the determination unit may be changed depending on the format to be converted.
Further, it may be combined with the addition of the image in the second to fourth embodiments, and when a part of the area to be added is always the same image, it may be determined that the corresponding block is always skipped.

Embodiment 7 FIG.
FIG. 14 is a block diagram showing an image coding method conversion apparatus according to Embodiment 7 of the present invention. In the figure, the same reference numerals as those in FIG.
The MB type determination unit 14 compares the JPEG quantization table of the current frame extracted by the simple decoding unit 1 with the JPEG quantization table of the last frame to be encoded stored in the reference data buffer 15. In addition, the JPEG quantized DCT coefficient of the current frame extracted by the simple decoding unit 1 is compared with the JPEG quantized DCT coefficient of the last frame to be encoded stored in the reference data buffer 15. Based on these comparison results, a process for determining whether to perform intra-frame encoding, inter-frame encoding, or skip encoding is performed. However, when the MB type determination unit 14 receives an intra-frame encoding determination flag indicating that it is necessary to perform intra-frame encoding from the MPEG-2 stream creation unit 17, the JPEG quantization table and the JPEG quantization DCT coefficient A determination result indicating that intra-frame coding is to be performed is output without performing the comparison process.
The reference data buffer 15 is used for the next comparison process in the MB type determination unit 14, which is the frame that is the last target of encoding by the MB type determination unit 14 (intra-frame encoding or inter-frame encoding). Frame) and a JPEG quantization DCT coefficient.
The MB type determination unit 14 and the reference data buffer 15 constitute an encoding determination unit.

The data conversion unit 16 converts the JPEG quantized DCT coefficient or the differential JPEG quantized DCT coefficient into an MPEG-2 quantized DCT coefficient according to the determination result of the MB type determining unit 14, and converts the JPEG quantization table into an MPEG-2 quantized table. To convert to the conversion table parameter. The data converter 16 constitutes a data conversion means.
The MPEG-2 stream creation unit 17 sets the intra-frame coding determination flag to “1” when the frame or MB to be encoded corresponds to the head of a GOP or slice in MPEG-2, and otherwise. If the determination result indicating that intra-frame encoding is to be performed is received from the MB type determination unit 14 while the intra-frame encoding determination flag is set to “0”, intra-frame encoding is performed and MPEG-2 is performed. When a stream is generated and a determination result indicating that interframe encoding is to be performed is received from the MB type determination unit 14, an interframe encoding is performed to generate an MPEG-2 stream. If a determination result indicating that encoding is skipped is received, the previously created MPEG-2 stream is output without performing intra-frame coding or inter-frame coding. Process to implement that. The MPEG-2 stream creation unit 17 constitutes an encoded stream creation means.

FIG. 14 shows an example in which the MB type determination unit 14 and the MPEG-2 stream creation unit 17 are applied to the image coding method conversion apparatus of FIG. 1, but FIG. 2, FIG. 6, FIG. 10, and FIG. For example, the MB type determination unit 14 or the MPEG-2 stream creation unit 17 may be applied to the 11 image coding method conversion apparatuses.
FIG. 15 is a flowchart showing the processing contents of the MB type determination unit 14 of the image coding method conversion apparatus according to the seventh embodiment of the present invention.

Next, the operation will be described.
When a JPEG stream in which an image signal is encoded with JPEG is input, the JPEG simple decoding unit 1 decodes a part of the JPEG stream and outputs header information and header information from the JPEG stream, as in the first embodiment. Extract the JPEG quantization table and JPEG quantization DCT coefficients.

When the JPEG simple decoding unit 1 extracts the header information from the JPEG stream, the MPEG-2 stream creation unit 17 sets “1” when the encoding target frame or MB hits the head of the GOP or the slice in MPEG-2. The intra-frame encoding determination flag is output to the skip determination unit 11, and when the encoding target frame or MB does not hit the head of the GOP or the slice in MPEG-2, the intra-frame encoding determination of “0” is performed. The flag is output to the skip determination unit 11.
In general, in a moving picture coding system that performs compression in the time direction, it is determined to forcibly perform intra-frame coding in order to prevent error propagation. In MPEG-2, this is the head of a GOP or the head of a slice.

When the simple decoding unit 1 extracts the JPEG quantization table and JPEG quantization DCT coefficient of the current frame, the MB type determination unit 14 extracts the JPEG quantization table and JPEG quantization DCT of the current frame from the simple decoding unit 1. Coefficients are input (step ST11), and an intra-frame coding determination flag is input from the MPEG-2 stream creation unit 17 (step ST12).
If the intra-frame encoding determination flag input from the MPEG-2 stream creation unit 17 is “1” (step ST13), the MB type determination unit 14 determines that it is necessary to perform intra-frame encoding. The JPEG quantization table and JPEG quantized DCT coefficient of the current frame are output to the data converter 16 (step ST14).

  In addition, the MB type determination unit 14 outputs a determination result indicating that intra-frame encoding is to be performed to the MPEG-2 stream creation unit 17 and the data conversion unit 16 (step ST20), and JPEG quantization of the current frame The table and the JPEG quantized DCT coefficient are output to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the JPEG quantization table and the JPEG quantized DCT coefficient of the last frame to be encoded) ) Is updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST21).

If the intra-frame coding determination flag input from the MPEG-2 stream creation unit 13 is “0” (step ST13), the MB type determination unit 14 stores the reference data stored in the reference data buffer 12 (finally The JPEG quantization table and JPEG quantized DCT coefficient of the frame to be encoded are input (step ST15).
Then, the MB type determination unit 14 compares the JPEG quantization table of the current frame with the JPEG quantization table of the last frame to be encoded, and the JPEG quantization DCT coefficient of the current frame. Finally, the JPEG quantized DCT coefficient of the frame to be encoded is compared (step ST16).

That is, the MB type determination unit 14 performs the comparison processing of the JPEG quantization table and the JPEG quantization DCT coefficient in the same manner as the skip determination unit 11 of FIG.
For example, as shown below, the MB type determination unit 14 classifies the comparison results, and determines whether to perform intraframe encoding, interframe encoding, or skip encoding.
(1) When the difference between the JPEG quantization table or the JPEG quantized DCT coefficient is large → It is determined that intraframe coding is performed. (2) When the difference between the JPEG quantization table and the JPEG quantized DCT coefficient is small (JPEG quantization) (Including the case where the difference between the table or the JPEG quantized DCT coefficient is very small) → Judge that the inter-frame coding is performed (3) When the difference between the JPEG quantization table and the JPEG quantized DCT coefficient is very small → coding Decide to skip

If the MB type determination unit 14 determines that the difference between the JPEG quantization table or the JPEG quantized DCT coefficient is large (step ST17) and the intraframe encoding is to be performed, the JPEG quantization table and the JPEG quantized DCT coefficient of the current frame are determined. Is output to the data converter 16 (step ST14).
In addition, the MB type determination unit 14 outputs a determination result indicating that intra-frame encoding is to be performed to the MPEG-2 stream creation unit 17 and the data conversion unit 16 (step ST20), and JPEG quantization of the current frame The table and the JPEG quantized DCT coefficient are output to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the JPEG quantization table and the JPEG quantized DCT coefficient of the last frame to be encoded) ) Is updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST21).

When the MB type determination unit 14 determines that the difference between the JPEG quantization table and the JPEG quantization DCT coefficient is small (steps ST17 and ST18) and performs the interframe encoding, the data conversion unit converts the JPEG quantization table of the current frame. 16 is output.
Also, the difference between the JPEG quantized DCT coefficient of the current frame and the JPEG quantized DCT coefficient of the last frame to be encoded is obtained, and the difference is calculated as the JPEG quantized DCT coefficient (hereinafter referred to as “difference JPEG quantum”). Output to the data converter 16 (step ST19).
As described above, when the differential JPEG quantized DCT coefficient is output to the data conversion unit 16, the data conversion unit 16 does not search for a motion vector used for moving image encoding, and thus the amount of calculation is significantly reduced. Is done.

  In addition, the MB type determination unit 14 outputs a determination result indicating that interframe coding is to be performed to the MPEG-2 stream creation unit 17 and the data conversion unit 16 (step ST20), and JPEG quantization of the current frame. The table and the JPEG quantized DCT coefficient are output to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (the JPEG quantization table and the JPEG quantized DCT coefficient of the last frame to be encoded) ) Is updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST21).

When the difference between the JPEG quantization table and the JPEG quantized DCT coefficient is very small (steps ST17 and ST18) and the MB type determination unit 14 determines that encoding is to be skipped, the MB type determination unit 14 displays a determination result indicating that encoding is skipped as MPEG. -The stream is output to the stream creation unit 17 and the data conversion unit 16 (step ST20).
In this case, since the data conversion unit 16 does not need to perform encoding method conversion, the JPEG quantization table and JPEG quantization DCT coefficient of the current frame are not output to the data conversion unit 16.
In addition, the MB type determination unit 14 outputs the JPEG quantization table and the JPEG quantization DCT coefficient of the current frame to the reference data buffer 12, and the reference data stored in the reference data buffer 12 (finally encoded). The JPEG quantization table and JPEG quantized DCT coefficient of the target frame are updated to the JPEG quantization table and JPEG quantized DCT coefficient of the current frame (step ST21).

When the determination result of the MB type determination unit 14 is a determination result indicating that intra-frame encoding is to be performed, the data conversion unit 16 uses the same method as the data conversion unit 2 in FIG. Convert to MPEG-2 quantized DCT coefficients and convert JPEG quantization table to MPEG-2 quantization table parameters.
However, in the seventh embodiment, a non-intra MB may be handled. In the case of a non-intra MB, the following formula (9) is used instead of the above formula (2), and the formula (4) Instead, the conversion is performed using the following equation (10).

If the determination result of the MB type determination unit 14 is a determination result indicating that interframe coding is to be performed, the data conversion unit 16 converts the differential JPEG quantized DCT coefficient output from the MB type determination unit 14 into MPEG-2. Convert to quantized DCT coefficients and convert JPEG quantization table to MPEG-2 quantization table parameters.
However, when converting from JPEG to MPEG-2, no particular conversion process is performed. Only QuantScale = 16 in equation (2). However, in the case of other encoding methods, some conversion processing may need to be performed.
Note that, when the determination result of the MB type determination unit 14 is a determination result indicating that encoding is skipped, the data conversion unit 16 does not perform encoding method conversion processing as described above.

When the MPEG-2 stream creation unit 17 receives the determination result indicating that the intra-frame encoding is to be performed from the MB type determination unit 14, the MPEG-2 stream creation unit 17 performs the intra-frame encoding as in the first embodiment, An MPEG-2 stream is created from the header information output from the simple decoding unit 1 and the MPEG-2 quantization table parameters and MPEG-2 quantized DCT coefficients output from the data conversion unit 16.
When the MPEG-2 stream creation unit 17 receives the determination result indicating that the interframe coding is to be performed from the MB type determination unit 14, the MPEG-2 stream creation unit 17 performs the interframe coding and outputs the header output from the JPEG simple decoding unit 1. An MPEG-2 stream is created from the information, the MPEG-2 quantization table parameter output from the data converter 16 and the MPEG-2 quantized DCT coefficient.
When the MPEG-2 stream creation unit 17 receives the determination result indicating that the encoding is skipped from the MB type determination unit 14, the MPEG-2 stream creation unit 17 does not perform the intra-frame coding or the inter-frame coding, and performs the previously created MPEG-2 stream. Output a stream.

  As is apparent from the above, according to the seventh embodiment, the MB type determination unit 14 determines whether to perform intra-frame encoding, inter-frame encoding, or skip encoding, When the MPEG-2 stream creation unit 17 receives a determination result indicating that intra-frame encoding is to be performed from the MB type determination unit 14, the intra-frame encoding is performed to generate an MPEG-2 stream, and the MB type determination unit 14, if a determination result indicating that interframe encoding is to be performed is received, an interframe encoding is performed to generate an MPEG-2 stream, and a determination result indicating that encoding is skipped from the MB type determination unit 14 If so, it is configured to output the previously created MPEG-2 stream without performing intra-frame coding or inter-frame coding. An effect that can be reduced. In addition, the amount of calculation can be kept low by performing MB determination in the DCT coefficient region.

In the seventh embodiment, the conversion from JPEG to MPEG-2 is shown, but it is also possible to convert from MPEG-2 I-picture to JPEG by the same method. Also, the present invention can be applied to conversion between other encoding methods using quantization or DCT.
In the seventh embodiment, since the converted format is MPEG-2, the unit for determining the skip is MB, but the determination unit may be changed depending on the format to be converted.
Further, it may be combined with the addition of the image in the second to fourth embodiments, and when a part of the area to be added is always the same image, it may be determined that the corresponding block is always skipped.

Embodiment 8 FIG.
FIG. 16 is a block diagram showing an image coding system conversion apparatus according to Embodiment 8 of the present invention. In the figure, the same reference numerals as those in FIG.
The color difference format conversion unit 18 uses the JPEG quantization table extracted by the JPEG simple decoding unit 1 to dequantize the JPEG quantized DCT coefficient extracted by the JPEG simple decoding unit 1 to obtain a color difference component, and the color difference component And processing for converting the JPEG quantized DCT coefficient according to the information indicating the JPEG color difference format included in the header information extracted by the JPEG simple decoding unit 1 and the MPEG-2 color difference format. To do. The color difference format conversion unit 18 constitutes color difference conversion means.
FIG. 16 shows an example in which the color difference format conversion unit 18 is applied to the image coding method conversion apparatus in FIG. 1, but the images in FIGS. 2, 6, 10, 11, 12, and 14 are shown. The color difference format conversion unit 18 may be applied to the encoding method conversion apparatus.

The eighth embodiment relates to a method for performing a conversion process at high speed even when the color difference formats are different before and after the encoding system conversion.
In image compression, in order to increase the compression rate, an image is often compressed using the YCbCr color space. At this time, the CbCr component (color difference) component is often thinned out compared to the Y (luminance) component, and the thinning method differs depending on the encoding method. This is the difference in color difference format.
In the eighth embodiment, a high-speed conversion process is realized by mounting the color difference format conversion unit 18.

The processing contents of the color difference format conversion unit 18 will be described below.
Here, for convenience of explanation, it is assumed that the color difference format of JPEG is 4: 2: 2 format and the color difference format of MPEG-2 to be converted is 4: 2: 0 format.
As shown in FIG. 18, the 4: 2: 0 format is a format in which the CbCr signal (C in the figure) is one pixel for the Y signal for four pixels.
The 4: 2: 2 format is a format in which the CbCr signal (C in the figure) is equivalent to two pixels for the Y signal for four pixels.
However, the positional relationship between luminance and color difference differs depending on the encoding method, and may differ from the position shown in FIG.

Therefore, in terms of block units in image compression, two 8 × 8 DCT coefficients in 4: 2: 2 format may be converted to one of 8 × 8 DCT coefficients in 4: 2: 0 format. It will be.
When performing such conversion, the simplest method is to first decode an image, perform filtering from a 16 × 8 image signal to an 8 × 8 image signal, degenerate, and then perform DCT again. Although it is mentioned, there is a drawback that the calculation amount becomes large.
Therefore, in the eighth embodiment, the color difference format conversion unit 18 directly converts 16 × 8 quantized DCT coefficients into 8 × 8 quantized DCT coefficients to reduce the amount of calculation.

ただし、Ａ_M×NはＤＣＴ変換行列である。 First, the color difference format conversion unit 18 performs DCT conversion on the M × N image signal using the following conversion formula (11).

A _{M × N} is a DCT transformation matrix.

ただし、Ｙ_8×8u，Ｙ_8×8lは、変換後のＤＣＴ係数のそれぞれ上半分、下半分の成分である。 In Expression (11), when M = N = 8, two 8 × 8 image signals arranged vertically are respectively X _{8 × 8u} and X _{8 × 8l} , and each is DCT-transformed separately, then the following expression ( 12).

However, Y _{8 × 8u} and Y _{8 × 8l} are components of the upper half and the lower half of the DCT coefficients after conversion, respectively.

ここで、下記の式（１４）（１５）のようにおくと、式（１３）は、下記の式（１６）のようになる。

When the upper and lower components of the equation (12) are connected, the following deformation can be made.

Here, if it puts like the following formula | equation (14) (15), Formula (13) will become like the following formula | equation (16).

式（１７）のＹ_16×8が式（１６）のＺ_16×8から得られれば、直接変換することができることになる。
ここで、式（１６）を変形すると、下記の式（１８）のようになる。

Next, when normal DCT conversion is performed on a 16 × 8 image signal, if M = 16 and N = 8 in the above equation (11), the following equation (17) is obtained.

If Y _{16 × 8 in} equation (17) is obtained from Z _{16 × 8 in} equation (16), it can be directly converted.
Here, when the equation (16) is modified, the following equation (18) is obtained.

ただし、Ａ_16×16は１６×１６のＤＣＴ変換行列である。また、Ｂ_16×16 ^Tは式（１５）と８×８のＤＣＴ変換行列から求められることは明らかである。
この方法により、個別にＤＣＴ変換を行った２つの８×８のＤＣＴ変換係数Ｚ_16×8から１６×８のＤＣＴ変換係数Ｙ_16×8を算出することが可能であり、色差フォーマットの異なる画像ブロックから算出されたＤＣＴ係数を復号を行わずに直接変換することができる。
このようにして、色差フォーマット変換部１８により変換されたＪＰＥＧ量子化ＤＣＴ係数がデータ変換部２に出力される。 Expression (19) is obtained from Expression (17) and Expression (18).

A _{16 × 16} is a 16 × 16 DCT transformation matrix. It is obvious that B _{16 × 16} ^T can be obtained from the equation (15) and the 8 × 8 DCT transformation matrix.
By this method, it is possible to calculate _{16 × 8} DCT conversion coefficient Y _{16 × 8} from two 8 × 8 DCT conversion coefficients Z _{16 × 8} that have been individually DCT converted, and images having different color difference formats. The DCT coefficient calculated from the block can be directly converted without decoding.
In this way, the JPEG quantized DCT coefficient converted by the color difference format conversion unit 18 is output to the data conversion unit 2.

  As is apparent from the above, according to the eighth embodiment, information indicating the JPEG color difference format included in the header information extracted by the JPEG simple decoding unit 1 by the color difference format conversion unit 18, and MPEG-2 Since the JPEG quantized DCT coefficients are converted according to the color difference format, even if the color difference formats are different, there is an effect that high-speed conversion processing can be realized.

In the eighth embodiment, the conversion from JPEG to MPEG-2 has been described. However, the same method can be applied to conversion between other encoding methods having different color difference formats.
Even in the case of color difference format conversion other than the conversion from the 4: 2: 2 format to the 4: 2: 0 format, the conversion coefficient can be calculated from the method shown in this method.

  1 JPEG simple decoding unit (simple decoding unit), 2 data conversion unit (data conversion unit), 3 MPEG-2 stream creation unit (encoded stream creation unit), 4 image data addition unit (image data addition unit), 5 MPEG-2 stream creation unit (encoded stream creation unit), 6 additional block image database (additional image creation unit), 7 additional image creation unit (additional image creation unit), 8 additional block image database (additional image creation unit), 9 Additional image creation unit (additional image creation unit), 10 DCT coefficient reduction unit (high frequency component deletion unit), 11 Skip determination unit (encoding determination unit), 12 Reference data buffer (encoding determination unit), 13 MPEG-2 Stream creation unit (encoded stream creation unit), 14 MB type determination unit (encoding determination unit), 15 Irradiation data buffer (coding judgment means), 16 data converter (data conversion unit), 17 MPEG-2 stream creation unit (coded stream creating means), 18 chrominance format conversion unit (color difference conversion means).

Claims (7)

A part of the encoded stream in which the image signal is encoded by the first encoding method is decoded, and the header information, the quantization table and the quantization frequency component of the first encoding method are obtained from the encoded stream. Extracting simple decoding means, and converting the quantization table and quantization frequency component of the first encoding method extracted by the simple decoding means to the quantization table and quantization frequency component of the second encoding method The image data of the additional image is quantized using the data conversion means to be converted and the quantization table of the second encoding method converted by the data conversion means, and the quantized image data is converted by the data conversion means. The image data adding means to be added to the quantized frequency component of the second encoding method and the header information extracted by the simple decoding means are changed according to the image width of the additional image. Quantization of the second encoding method in which the header information after the change, the quantization table of the second encoding method converted by the data conversion unit, and the image data after quantization by the image data addition unit are added. Encoded stream creation means for creating an encoded stream of the second encoding method from the encoded frequency component ,
The quantization table of the first encoding method extracted this time by the simple decoding unit is compared with the quantization table of the first encoding method previously extracted by the simple decoding unit, and the simple The quantization frequency component of the first encoding method output this time from the decoding means and the quantization frequency component of the first encoding method output last time from the simple decoding means are compared. An encoding determination unit is provided for determining whether to perform intraframe encoding based on the comparison result,
When the coded stream creation means receives a determination result indicating that intra-frame coding is to be performed from the coding judgment means, it performs intra-frame coding and creates a coded stream of the second coding method. If the determination result indicating that intra-frame encoding is not performed is received from the encoding determination means, the previously generated encoded stream of the second encoding method is output without performing intra-frame encoding. An image coding method conversion apparatus characterized by the above. When the encoding determination unit receives a flag indicating that it is necessary to perform intra-frame encoding from the encoded stream creation unit, the encoding determination unit does not perform the comparison process of the quantization table and the quantization frequency component, but performs the intra-frame code. picture coding method conversion apparatus according to claim 1, wherein the outputting the determination result indicating that implementing reduction. A part of the encoded stream in which the image signal is encoded by the first encoding method is decoded, and the header information, the quantization table and the quantization frequency component of the first encoding method are obtained from the encoded stream. Extracting simple decoding means, and converting the quantization table and quantization frequency component of the first encoding method extracted by the simple decoding means to the quantization table and quantization frequency component of the second encoding method The image data of the additional image is quantized using the data conversion means to be converted and the quantization table of the second encoding method converted by the data conversion means, and the quantized image data is converted by the data conversion means. The image data adding means to be added to the quantized frequency component of the second encoding method and the header information extracted by the simple decoding means are changed according to the image width of the additional image. Quantization of the second encoding method in which the header information after the change, the quantization table of the second encoding method converted by the data conversion unit, and the image data after quantization by the image data addition unit are added. Encoded stream creation means for creating an encoded stream of the second encoding method from the encoded frequency component ,
The quantization table of the first encoding method extracted by the simple decoding means is compared with the quantization table of the first encoding method of the last frame to be encoded. The quantization frequency component of the first encoding method output from the decoding means is compared with the quantization frequency component of the first encoding method of the last frame to be encoded. Provided with an encoding determination means for determining whether to perform intra-frame encoding, inter-frame encoding, or skip encoding based on the comparison result,
When the coded stream creation means receives a determination result indicating that intra-frame coding is to be performed from the coding judgment means, it performs intra-frame coding and creates a coded stream of the second coding method. If the determination result indicating that inter-frame encoding is to be performed is received from the encoding determination means, inter-frame encoding is performed to generate an encoded stream of the second encoding method, and the encoding determination is performed. If a determination result indicating that encoding is skipped is received from the means, the previously generated encoded stream of the second encoding method is output without performing intraframe encoding and interframe encoding. An image coding method conversion apparatus. When the encoding determination unit receives a flag indicating that it is necessary to perform intra-frame encoding from the encoded stream creation unit, the encoding determination unit does not perform the comparison process of the quantization table and the quantization frequency component, but performs the intra-frame code. 4. The image coding method conversion apparatus according to claim 3 , wherein a determination result indicating that the conversion is performed is output. Create additional image by using the time information, the image encoding method of claim 1 or claim 3, wherein the providing the additional image generating means for outputting the image data of the additional image into image data addition means Conversion device. Create additional image using the header information extracted by the simple decoding means, according to claim 1 or claims, characterized in that a additional image generating means for outputting the image data of the additional image into image data addition means Item 4. The image coding method conversion device according to Item 3 . Color difference conversion means for converting the quantized frequency component of the first encoding method output from the simple decoding means according to the color difference format of the first encoding method and the color difference format of the second encoding method is provided. The image coding method conversion apparatus according to any one of claims 1 to 6 , wherein the image coding system conversion apparatus is any one of claims 1 to 6 .

Images