JP2003319389A - Image data decoding apparatus and structure of image data - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a decoding apparatus capable of handling image encoded data in a unified way independently of the transmission system and to provide the structure of image data. <P>SOLUTION: A data input means 101 receives header data including N slice data being a part of an encoded image frame, a parameter required for particularizing a start position of the N slice data, and size information required for decoding the N slice data and data comprising connected slice data for connecting the N slice data, a slice separate means 102 refers to the parameter required for particularizing the start position in the header data to particularize the start position of the slice data included in the connected slice data received successively, a decoding means 103 decodes the data and a display means 104 displays the decoded data. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to MPEG-4 (M
moving Picture ExpertsGrou
The present invention relates to a decoding device for compression-coded image frame data such as p Phase 4).

[0002]

2. Description of the Related Art In 1999, MPEG-4 was standardized as an image coding system at a low bit rate. However, JVT (Joint Video), which is a joint working group of ISO / IEC and ITU-T, is used as an encoding method that realizes a compression efficiency higher than that of MPEG-4.
o Team), a new coding scheme is being standardized. In this encoding method, the grammar of the bit stream below the slice data is specified only, but the information higher than the slice data, that is, the information corresponding to the slice header or VOP header in the case of conventional MPEG-4, is not included. The policy is to depend on the transmission method, and is specified individually for each transmission method.

As a first example of the transmission method, the transmission method is M
PEG-2 transport stream (ISO / IE
C 13818-1), as shown in FIG.
It is considered that a slice header starting from a 32-bit slice start code is added to each slice, and a picture header preceding the slice header is added to the head slice of each image frame. Also, to prevent the same bit pattern as the start code from occurring below the slice data, as shown in FIG. 10, for example, when 16 bits of "0" continue in the slice data and below, 8 bits of "1" are forcibly inserted. And then MPE
It is assumed to be transmitted as a G-2 transport stream.

As a second example of the transmission method, the transmission method is R
TP (Realtime Transport Pro)
tocol, IETF / RFC1889),
As shown in FIG. 11, the RTP packet includes one slice data. As a third example of the transmission method,
The transmission method is MP4 file (ISO / IEC 1449
In the case of 6-1), as shown in FIG. 12, it is considered to specify the size information of each slice in the header portion normally arranged at the beginning of the MP4 file.

[0005]

As described above,
Since the information higher than the slice data is completely different for each transmission method, there is a problem that different processing is required for each transmission method at the time of decoding. For example, MPEG-
In 2, if 16 bits of '0' continue below the slice data, the process of deleting the following 8 bits is necessary.
This is not necessary for RTP and MP4. Also, when specifying one slice, MPEG-2 needs to search a slice start code, and MP4 needs to acquire size information of each slice described in the file header portion, but one packet is used in RTP. Since it is 1 slice, no special processing is required. In the case of MP4, since the size information of all slices is specified in the file header part, the size of the file header part increases, and there is also a problem that when playing back an MP4 file while downloading it, it takes a long time to start playing. Further, in the case of RTP, since one RTP packet includes only one slice, packet overhead increases and transmission efficiency decreases.

Another problem is that the processing becomes complicated when converting the transmission method. For example, when converting image data received by MPEG-2 to RTP and transmitting it, a sequence of '0's below the slice data is searched, and when 16 bits are consecutive, the following 8 bits are deleted and further sliced. It is necessary to delete the start code and make it 1RTP packet.

[0007]

In order to solve the above-mentioned problems, in the image data coding apparatus and the image data decoding apparatus of the present invention, information which is commonly used in each transmission system and which is higher than slice data is used. The header information and the slice data are encoded and output, or input and decoded, and are specifically configured by the following means.

The image data decoding apparatus according to claim 1 of the present invention is the number N of slice data (N is an integer of 1 or more) which is a part of an encoded image frame and the start of the slice data of N pieces. Input data including header data in which a parameter used to specify a position and side information used in decoding the N slice data are concatenated, and concatenated slice data in which N slice data are concatenated. Referring to the data input means and the parameter used for specifying the start position in the input header data, the slice separating means for specifying the start position of the slice data included in the input concatenated slice data, and the start position Decoding means for decoding the specified slice data by referring to side information used for decoding the header data; Display means for displaying the slice data, and comprising a.

The image data decoding apparatus according to claim 2 of the present invention is for identifying that the side information in the header information in claim 1 is side information relating to the first slice data in one image frame. It is intended to include information used for.

In the image data decoding apparatus according to claim 3 of the present invention, the side information in the header information in claim 1 is information used for specifying the display position of each slice data in one image frame. It is meant to be included.

An image data decoding apparatus according to a fourth aspect of the present invention forms one image frame when the side information in the header information according to the first aspect is at least the first slice data in one image frame. It includes information commonly referred to by a plurality of slice data.

The image data structure according to claim 5 of the present invention is the number N (N
Is an integer greater than or equal to 1), a parameter used to specify the start position of the N slice data, and side information used to decode the N slice data. , Concatenated slice data obtained by concatenating the N slice data.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION The image data decoding apparatus according to the present invention is a conventional MPEG-1, MPEG-2, MPEG-
4 does not require a start code for specifying the start position of a slice required for image coding, and MPEG-2
While inputting and referring to header information that can be commonly used in a transmission method, RTP, MP4 file, etc., it is possible to input a plurality of slice data and decode, reproduce and display.

(First Embodiment) A first embodiment of the image data decoding apparatus of the present invention will be described below with reference to FIGS. 1 to 8. FIG. 1 shows a functional block diagram of the image data decoding device. The image data decoding device includes data input means 1
01, slice separation means 102, slice data decoding means 103, and display means 104.

The data input means 101 inputs data having a structure as shown in FIGS. The data shown in FIG. 2 is obtained by concatenating slice data for one image frame,
Header data relating to these slice data is added. In this example, the five slice data forming one image frame are connected, but the number of slice data to be connected is arbitrary within the range of 1 or more. In addition,
On the right side of the figure, the starting position of each slice data is shown in bytes with the slice immediately after the header as the origin (zero). FIG. 3 is a schematic diagram showing the structure of header data. It holds information about slice data that follows the header data. First, there is a "number of entries" field that indicates the number of concatenated slice data. If five slice data are connected as shown in FIG. 2, the number of entries can be set to five. And each entry #
For 1 to # 5, the start position of each slice and the side information required when decoding each slice are described. The start position of the slice is set to 0 in the case of entry # 1. Further, in the example of FIG. 2, the fifth slice of the concatenated slice data starts from the 361th byte, so the slice start position in the entry # 5 of FIG. 3 is 361. In the present embodiment, the slice data immediately after the header data is used as the reference point, so it is set to zero, but another point, for example, the first header data
The byte may be used as the reference point, in which case the start position of the first slice is not 0, but a value that depends on the size of the header data. As the side information, as shown in FIG. 3, the number of the head macroblock of the slice can be used as an example. One slice consists of a plurality of macroblocks. The macroblock is data of width 16 and height 16 pixels in MPEG-4. Image size is 176 width
An image frame having pixels and a height of 144 pixels is composed of 99 macroblocks. The starting macroblock number in the first slice of the image frame is generally 0, but since the first macroblock may not be encoded, it may be other than 0. In the case of the first slice in each image frame, frame level information relating to the entire image frame is designated. As an example of frame level information,
For example, as the encoding mode of the image frame, it is possible to use information for identifying whether it is an intra-frame encoded frame or a frame using inter-screen correlation. Further, the frame level information may be specified not only in the first slice data in each image frame but also in any slice data. For example, as shown in FIGS. 4 and 5, when one image frame is divided into two slices on the one side and three slice data on the other side, the frame is included in the first slice of the separated header data. Level information may be specified. As a clue to determine the presence / absence of frame level information, a method of referring to the leading macroblock number or a method of arranging a flag for determining the presence / absence of frame level information before the frame level information and referencing this can be used. Is. In the former case, when the head macroblock number is zero, it can be determined that there is frame level information. Further, as a method for connecting slice data, as shown in FIG. 6, it is also possible to connect slice data of different image frames. Concatenated slice data # 3 in FIG. 6 is composed of slice data belonging to image frame # 1 and image frame # 2. In this case, the side information for the second slice data of the header data in the concatenated slice data # 3 includes the frame level information of the image frame # 2.

FIG. 7 shows a formatted grammar of input data. The input data is "compound_sli"
ce () ”. In FIG. 7A,“ compo ”
“Und_slice ()” first indicates the number of concatenated slices by the “number_of_slice” field of the 8-bit field, and then, for each slice, “slice_start_offs” of 15 bits.
The start position of the slice is indicated by "et", and 1-bit "
The "picture_header_flag" indicates whether or not the frame level information appears, and "picture
When e_header_flag "is 1, frame-level information" picture_header () "is arranged, followed by 8-bit" first_macr ".
The starting macroblock number is indicated by "block_number". After presentation of the header information, "number_of_slice" is set as concatenated slice data.
The slice data of the number indicated by "slice_data"
7 (b) is a modification of FIG. 7 (a). Instead of indicating the presence / absence of the frame level information by "picture_header_flag", "fir" is connected.
When st_macroblock_number "is zero, that is, when it is the first slice data of the image frame, the frame level information is arranged. FIG. 7C is an extension of FIG. 7A. First, 7-bit "video_sequence"
The e_index "field specifies information about the entire image sequence related to the concatenated slice data. The information about the entire image sequence includes image size, profile and level information, and specific examples include MPEG- Information corresponding to the video object layer in 4 and the sequence header in MPEG-2 is possible.Continued "group_of_pi"
"cture_flag" is "group_of_pic"
"true ()" appears, and if this flag is 1, "group_of_picture ()" is placed. This is information indicating that a plurality of image frames are grouped, and as a specific example, GOV (Group of Vop) in MPEG-4 and MPEG
-2 GOP (Group of Picture)
Information corresponding to e) is possible.

Further, although not shown in FIG. 7, it is possible to designate a flag indicating that random access is possible. Note that the field names shown in FIG. 7 are merely examples, and other field names may be used as long as they have the same meaning.

In order to decode the slice, the slice separating means 102 analyzes the header data part to specify the start position of the slice data, and decodes the slice, with respect to the data input means 101 which has input data having the above structure. It outputs to the slice decoding means 103 with the information required for. The slice decoding means 103 reads one slice of data from the data input means 101 based on the start position and size of the slice data designated by the slice separating means 102, and decodes the slice.

The display means 104 is a slice decoding means 103.
Display the slice data decoded by. As an example of the display timing, it is displayed immediately when the decoding of the slice data is completed, or it is displayed when the decoding of the slice for one frame is completed, or although not shown in the figure, an instruction from the outside is given. It is possible to display decoded slice data by a signal.

FIG. 8 shows PE which is an MPEG-2 transmission system.
An example (a) in which “compound_slice ()” is stored in the S packet, R which is also the RTP transmission method
An example (b) when storing in a TP packet and an example (c) when storing in an MP4 file which is an MP4 transmission method are shown. As shown in the figure, "compound" which is a common data structure in various transmission systems
_Slice () "can be used, there is no need to delete the start code in the MPEG-2 transmission system as described as the problem of the conventional example, and the overhead of 1RTP packet / slice data is eliminated in the RTP transmission system to reduce overhead. It is possible to reduce the size of the file header part in the MP4 file transmission method.

Since a common data structure is used regardless of the transmission system, conversion of the transmission system, for example, MPEG-2.
The conversion from the transmission system to the RTP transmission system can be easily performed. For example, remove the PES packet header and add R
By adding the TP header, the transmission system can be converted. This is, for example, MPEG for terrestrial or satellite broadcasting
-It is very effective when transmitting image data received by the -2 transmission method on the Internet by the RTP transmission method.

[0022]

According to the first aspect of the present invention, there is provided the image data decoding device, wherein the number N of slice data forming a part of an encoded image frame (N is an integer of 1 or more) and the N slice data. Data including header data obtained by concatenating the parameter used for specifying the start position of the slice information and side information used for decoding the N slice data, and concatenated slice data obtained by concatenating N slice data. A data input means for inputting, a slice separating means for identifying a start position of slice data included in the input concatenated slice data with reference to a parameter used for identifying the start position in the input header data, and a start Decoding means for decoding the slice data whose position is specified by referring to side information used for decoding the header data; To provide a display means for displaying the slice data, and without using a start code enables separating a plurality of slice data.

The image data decoding apparatus according to claim 2 of the present invention is characterized in that the side information in the header information in claim 1 includes information for identifying that it is related to the first slice data in one image frame. ,
It is possible to identify whether or not it is the first slice data in the image frame.

In the image data decoding apparatus according to claim 3 of the present invention, since the side information in the header information in claim 1 includes information indicating the display position of each slice data in one image frame, each slice is After decryption,
The display position can be reliably determined.

An image data decoding apparatus according to a fourth aspect of the present invention forms one image frame when the side information in the header information according to the first aspect is at least the first slice data in one image frame. Since it includes information commonly referred to by a plurality of slice data, it becomes possible to efficiently decode information commonly used between image frames.

The image data structure according to claim 5 of the present invention is the number N (N
Is an integer greater than or equal to 1), a parameter used to specify the start position of the N slice data, and side information used to decode the N slice data. , And a concatenated slice data in which the N slice data are concatenated, a common data structure can be realized in various transmission systems, and as a result, one transmission system to the other transmission system can be realized. In addition, the analysis part of the data structure can be made common in the image data decoding device that inputs and decodes the image data.

[Brief description of drawings]

FIG. 1 is a functional block diagram of an image data decoding device according to a first embodiment of the present invention.

FIG. 2 is a diagram for explaining input data of the image data decoding device according to the first embodiment of the present invention.

FIG. 3 is a diagram for explaining input data of the image data decoding device according to the first embodiment of the present invention.

FIG. 4 is a diagram for explaining input data of the image data decoding device according to the first embodiment of the present invention.

FIG. 5 is a diagram for explaining input data of the image data decoding device according to the first embodiment of the present invention.

FIG. 6 is a diagram for explaining input data of the image data decoding device according to the first embodiment of the present invention.

FIG. 7 is a grammar diagram for explaining input data of the image data decoding device according to the first embodiment of the present invention.

FIG. 8 is a data structure and MP according to the first embodiment of the present invention.
It is a figure for demonstrating the storage method in transmission systems, such as EG-2.

FIG. 9 is an explanatory diagram of a conventional example of an image data structure when transmitting by MPEG-2.

FIG. 10 is an explanatory diagram of a conventional example of an image data structure when transmitting by MPEG-2.

FIG. 11 is an explanatory diagram of a conventional example of a data structure when transmitting by RTP.

FIG. 12 is an explanatory diagram of a conventional example of a data structure when transmitting as an MP4 file.

[Explanation of symbols]

101 data input means 102 slice data separation means 103 slice decoding means 104 display means

Claims (5)

[Claims] 1. A number N of slice data (N is an integer of 1 or more) which is a part of an encoded image frame, a parameter used for specifying a start position of the slice data of N pieces, and the N pieces of slice data. Data input means for inputting data including header data concatenated with side information used for decoding slice data and concatenated slice data concatenated with N slice data, and the start in the inputted header data Referring to the parameter used to specify the position, the slice separating means for specifying the start position of the slice data included in the input concatenated slice data, and the slice data for specifying the start position are used for decoding the header data. Slice data decoding means for decoding by referring to the side information to be used, and a display for displaying the decoded slice data. An image data decoding device comprising: a stage. 2. The side information in the header information is
The image data decoding apparatus according to claim 1, further comprising information used for identifying that the side information is the side information regarding the first slice data in one image frame. 3. The side information in the header information is
The image data decoding apparatus according to claim 1, further comprising information used for specifying a display position of each slice data in one image frame. 4. The side information in the header information is
The image data decoding apparatus according to claim 1, wherein the image data decoding apparatus includes information commonly referred to by a plurality of slice data forming one image frame when it is at least the first slice data in one image frame. . 5. The number N of slice data forming image data (N is an integer of 1 or more), a parameter used for specifying a start position of the N slice data, and the N slices. An image data structure including header data in which side information used for data decoding is concatenated, and concatenated slice data in which the N slice data are concatenated.