WO2007043583A1

WO2007043583A1 - Image encoding device, image decoding device, and method thereof

Info

Publication number: WO2007043583A1
Application number: PCT/JP2006/320331
Authority: WO
Inventors: Chong Soon Lim; Sheng Mei Shen; Shinya Kadono
Original assignee: Matsushita Electric Industrial Co., Ltd.
Priority date: 2005-10-11
Filing date: 2006-10-11
Publication date: 2007-04-19

Abstract

There are provided an image encoding device, an image decoding device, and others capable of using a scan pattern specially set by a user. The image encoding device and the image decoding device employ a new method for adaptively changing the scan pattern of an orthogonal conversion coefficient. This new scan pattern is encoded in the encoded bit stream of a video sequence and can use an optimal scan pattern for each picture. By adding a small amount of header information, it is possible to select a scan pattern having a preferable compression efficiency of the entropy encoding, thereby increasing the video sequence compression efficiency.

Description

Specification

Image encoding apparatus, image decoding apparatus, and methods thereof

Technical field

TECHNICAL FIELD [0001] The present invention relates to an image encoding device and an image decoding device, and more particularly to control of a scanning method for converting a two-dimensional orthogonal transformation coefficient after orthogonal transformation of a moving image into one dimension. Background art

[0002] Conventionally, as a typical form for encoding interlaced pictures, there is a form performed in units of frame macroblocks or field macroblocks. In addition, in order to achieve the coding efficiency at that time, as shown in FIG. 1 (a), there are few cases where a frame macroblock and a field macroblock are mixed in the picture to be coded. Fig. 1 (b) is a diagram schematically showing the pixel configuration of the frame macroblock and the final macroblock.

[0003] Conventionally, a scan pattern defined in advance is used as a scanning method when the orthogonal transform coefficient after the orthogonal transform is made one-dimensionally in units of frame macroblocks or finered macroblocks. Fig. 2 (a) is an example of the scan pattern in the zigzag scan method used when encoding in the conventional 4 x 4 pixel frame macroblock unit, and Fig. 2 (b) shows the conventional 4 x 4 scan pattern. It is an example of a scan pattern in a field scan method used when encoding in a field macroblock unit of 4 pixels. Fig. 2 (c) is an example of a scan pattern in the zigzag scan method used when encoding in the conventional frame macroblock unit of 8 x 8 pixels. Fig. 2 (d) It is an example of a scan pattern in a field scan method used when encoding is performed in units of a conventional field macroblock of 8 × 8 pixels. Here, the upper left is a low-frequency component, the horizontal frequency increases toward the right, and the vertical frequency increases toward the bottom.

[0004] The above zigzag scanning method and the like are general scanning methods used in standards such as MPEG-1 Video, MPEG-2 Video, ITU-T H.264, and frequency conversion (MPEG-1, MPEG-2). Block coefficients that have undergone discrete cosine transform) This is a method for scanning in such a way that it is effectively compressed using a mouth pie code.

[0005] As shown in Figs. 2 (a) to (d) above, the transform coefficient of the block subjected to the two-dimensional orthogonal transform is scanned from the small coefficient coefficient position of the illustrated value to the large coefficient coefficient position. The coefficients are entropy coded in that order.

Non-Patent Literature 1: ITU-T Recommendation H.2b4, Advanced video coding for generic a udiovisual services, (03/2005)

Disclosure of the invention

Problems to be solved by the invention

However, depending on the video sequence or picture, a predetermined scan pattern such as a zigzag scan pattern may not be the best scan pattern for the picture. Depending on the content of the picture, the specially set scan pattern may be more efficient in code encoding than the predetermined scan pattern. In such a case, the code efficiency of these pictures can be increased by encoding a new scan pattern.

[0007] The present invention has been made in view of the above problems, and provides an image encoding device, an image decoding device, and the like that can also use a scan pattern specially set by a user. For the purpose.

Means for solving the problem

In order to solve the above problems, an image encoding device according to the present invention is an image encoding device that encodes a picture in units of macroblocks. A determination means for determining whether or not to code using at least two or more types of macroblocks, and when the determination means determines that encoding should be performed using the two or more macroblocks For each type of macroblock, information indicating one selected scanning method is encoded as header information, and at the same time, based on the selected scanning method. And an encoding means for encoding orthogonal transformation coefficients of a two-dimensional array of macroblocks corresponding to each scanning method in the order of the one-dimensional array.

[0009] Also, the image decoding apparatus according to the present invention provides a coded bitstream having header information. An image decoding apparatus that decodes a macroblock unit, decodes the header information of the code bitstream, and forms at least a picture from the decoded header information For two or more types of macroblocks, information indicating the scanning method for rearranging the orthogonal transformation coefficients of the two-dimensional array into a one-dimensional array is extracted, and the information indicating the scanning method is used to code in one-dimensional array order. Decoding means for decoding the orthogonal transform coefficient of the coded bit stream converted into a two-dimensional array is provided.

[0010] A novel point of the present invention is that an arbitrary new scan pattern can be downloaded to the image decoding apparatus by including it in the header information of the bitstream, so that an optimal scan pattern for each picture can be obtained. It is to be able to use it. As a result, a scan pattern with high compression efficiency of the entropy code can be selected by adding a small amount of header information, and image compression with higher code efficiency can be realized.

Note that the present invention can be realized as an image encoding method or an image decoding method in which a characteristic configuration unit in the image encoding apparatus or the image decoding apparatus is used as a step. It can also be realized as a program that causes a computer or the like to execute the steps. Needless to say, the program can be widely distributed via a recording medium such as a DVD or a transmission medium such as the Internet. Furthermore, the present invention can be realized as an integrated circuit including characteristic constituent means of the image encoding device or the image decoding device.

The invention's effect

[0012] According to the image encoding device and the image decoding device according to the present invention, it is possible to use a scan pattern specially set by the user, and it is possible to improve the encoding efficiency.

Brief Description of Drawings

[0013] FIG. 1 (a) is a diagram showing a state of a picture to be coded, in which a frame macroblock and a finered macroblock are mixed. Figure 1 (b) is a diagram schematically showing the pixel configuration of the frame macroblock and field macroblock.

[Fig. 2] Fig. 2 (a) shows the case where a conventional 4 x 4 pixel frame macroblock unit is coded. It is an example of the scan pattern in the zigzag scan system used. FIG. 2 (b) is an example of a scan pattern in the zigzag scan method used when encoding is performed in the field macroblock unit of the conventional 4 × 4 pixels. Fig. 2 (c) shows an example of a conventional zigzag scan method used for encoding in units of 8 x 8 pixel frame macroblocks. FIG. 2 (d) shows an example of a scan pattern in the field scan method used when coding in the conventional 8 × 8 pixel field macroblock unit.

FIG. 3 is a block diagram showing a functional configuration of an image encoding device according to the present invention.

FIG. 4 is a block diagram showing a functional configuration of an image decoding apparatus according to the present invention.

FIG. 5 is a flowchart showing a process for encoding a scan pattern in a sequence header in the image encoding device.

FIG. 6 is a flowchart showing a process of decoding a scan pattern in a sequence header in the image decoding apparatus.

FIG. 7 is a flowchart showing a process for encoding a scan pattern in a picture header in the image decoding apparatus.

FIG. 8 is a flowchart showing a process for decoding a scan pattern in a picture header in the image decoding apparatus.

FIG. 9 is a flowchart showing a process for encoding a scan pattern in a slice header in the image encoding apparatus.

FIG. 10 is a flowchart showing a process of decoding a scan pattern in a slice header in the image decoding apparatus.

[FIG. 11] FIG. 11 (a) is a diagram showing a stream structure of a code bit stream of a video sequence output from the image coding apparatus of the present invention. FIG. 11 (b) is a diagram showing a stream structure of a code bit stream of a picture in which the image code device capability of the present invention is also output. FIG. 11 (c) is a diagram showing the stream structure of the coded bit stream of the slice output from the image coding apparatus of the present invention.

FIG. 12 is a diagram showing the relationship between the first and second reference pictures and the target picture in the present invention. FIG. 13 (a) is a diagram showing an example of a scan pattern position and a reverse scan pattern position for a 4 × 4 block zigzag scan pattern. FIG. 13 (b) is a diagram showing a stream structure of an encoded scan pattern for 4 × 4 blocks output from the image encoding device of the present invention. FIG. 13 (c) is a diagram showing a stream structure of an encoded scan pattern for an 8 × 8 block in which the image encoding apparatus power of the present invention is also output.

FIG. 14 is a flowchart showing a process of selecting a predetermined scan pattern in the image encoding device and the image decoding device of the present invention.

FIG. 15 is a flowchart showing a process of selecting a memory strength scan pattern in the image encoding device and the image decoding device of the present invention.

FIG. 16 is a flowchart showing a process of storing a scan pattern in a memory unit in the image encoding device and the image decoding device of the present invention.

FIG. 17 is a flowchart showing a process of selecting a scan pattern corresponding to progressive, field, and block size from the memory unit in the image encoding device and the image decoding device of the present invention.

FIG. 18 is a flowchart showing a process of storing a scan pattern corresponding to progressive, field, and block size in the memory unit in the image encoding device and the image decoding device of the present invention.

FIG. 19 is a flowchart showing a process of selecting a progressive scan pattern from the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 1 of the present invention.

FIG. 20 is a flowchart showing a process of selecting a field scan pattern from the memory unit for the image coding apparatus and the image decoding apparatus according to Embodiment 1 of the present invention. is there.

FIG. 21 is a flowchart showing a process of storing a sequence progressive scan pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 1 of the present invention.

FIG. 22 shows a process of storing a sequence field scan pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 1 of the present invention. It is a flowchart to show.

FIG. 23 is a flowchart showing a process of storing a progressive scan pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 1 of the present invention.

FIG. 24 is a flowchart showing a process of storing a progressive scan pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 1 of the present invention.

FIG. 25 is a flowchart showing a process of selecting a progressive scan pattern from the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 2 of the present invention.

FIG. 26 is a flowchart showing a process of selecting a field scan pattern from the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 2 of the present invention. is there.

FIG. 27 is a flowchart showing a process of storing a sequence progressive scan pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 2 of the present invention.

FIG. 28 is a flowchart showing a process of storing a sequence field pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 2 of the present invention.

FIG. 29 is a flowchart showing a process of storing a progressive scan pattern in a memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 2 of the present invention.

FIG. 30 is a flowchart showing a process of storing a field scan pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 2 of the present invention.

FIG. 31 is a flowchart showing a process of selecting a field scan pattern from a memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 3 of the present invention. is there. FIG. 32 is a flowchart showing a process of storing a sequence field pattern in a memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 3 of the present invention.

FIG. 33 is a flowchart showing a process of selecting a field scan pattern from the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 4 of the present invention. is there.

FIG. 34 is a flowchart showing a process of storing a sequence field scan pattern in the memory unit in the image coding apparatus and the image decoding apparatus according to Embodiment 4 of the present invention. Explanation of symbols

100 image encoder

101 Encoded image signal

102 frame memory

103 Motion detector

104 Motion compensation unit

105 Orthogonal transformation unit

106 Quantizer

107 Variable length encoder

108 Inverse quantization

109 Inverse orthogonal transform

110 Bit rate controller

111 Differentiator

112 Adder

121 Code bitstream

130 Judgment part

135 Memory section

140 Overall control unit

141 Decoded image signal 200 Image decoding device

201 Variable length decoder

202 frame memory

208 Inverse quantization

209 Inverse orthogonal transform unit

212 Adder

235 Memory section

240 Overall control unit

BEST MODE FOR CARRYING OUT THE INVENTION

FIG. 3 is a block diagram showing a functional configuration of the image encoding device 100 according to the present invention. This image encoding device 100 is a header information for scanning suitable orthogonal transform coefficients corresponding to even when different types of macroblocks are mixed when encoding a picture. Is also a device that performs code coding, and includes a determination unit 130, an orthogonal transformation unit 105, a quantization unit 106, a variable length coding unit 107, a memory unit 135, an inverse quantization unit 108, an inverse orthogonal transformation unit 109, a frame memory 102, a motion detection unit 103, a motion compensation unit 104, a bit rate control unit 110, and the like. Note that the memory unit 135 performs Z-reading of information related to the scanning method according to an instruction from the variable-length code key unit 107.

[0016] Based on the characteristics of the input encoded image signal 101, the determination unit 130 determines the type of macroblock at the time of encoding, etc. (for example, the ability to mix a frame macroblock and a field macroblock in a picture) And whether the corresponding macroblock is a frame macroblock force field macroblock, etc.), and the result of the determination is converted into an orthogonal transform unit 105, a motion detection unit 103, an inverse orthogonal transform unit 109, The variable length coding unit 107 is notified.

[0017] The orthogonal transform unit 105 performs discrete cosine transform on the received encoded image signal 101 (image signal data) in units of macroblocks or in units of blocks constituting the macroblock.

(DCT: Discrete Cosine Transform) is performed, DCT coefficients are generated, and output to the quantization unit 106. Here, for the frame of Klntra coded picture, DCT operation is performed on the pixel values in the screen. For frames of P (Predictive coded) pictures, prediction errors based on forward prediction coding based on I pictures or P pictures located in the past in time! A DCT operation is performed on the difference pixel value. For B (Bidirectionally) picture frames, DCT calculation is performed on the prediction error pixel values based on bi-directional prediction codes based on I / P pictures located before and after in time! . The orthogonal transform unit 105 performs orthogonal transform corresponding to each frame or field based on the macroblock type (frame macroblock or field macroblock) determined by the determination unit 130.

[0018] Quantization section 106 quantizes the DCT coefficients input from orthogonal transform section 105 in the quantization step (or may be a quantization parameter) received from bit rate control section 110 for each macroblock. And output to the variable-length code unit 107 and the inverse quantization unit 108. The variable length code key 107 is a scan order based on the macro block type (frame macro block or field macro block) determined by the determination unit 130 for the quantized DCT coefficient input from the quantization unit 106. Is obtained from the memory unit 135, variable length coding and multiplexing of the DCT coefficients quantized in the scan order are performed, and a bit stream is output to an output buffer (not shown). In addition to the quantized DCT coefficients, the variable length code key unit 107 codes a motion vector, a quantization step, a macro block type, and the like to form a bit stream. The scan order based on the macroblock type is preliminarily stored in the memory unit 135 and is encoded as the first header information of the picture.

Inverse quantization section 108 performs an inverse quantization operation on the quantized DCT coefficient received from quantization section 106 and outputs the result to inverse orthogonal transform section 109. The inverse orthogonal transform unit 109 performs the inverse quantized DCT coefficient input from the dequantization unit 108 based on the macro block type (frame macro block or final red macro block) determined by the determination unit 130. Image signal data is reproduced by performing an inverse orthogonal transform operation corresponding to each frame or field and output to the frame memory 102.

The frame memory 102 adds and stores the decoded image signal data of the I picture or P picture and the motion compensation data generated by the motion compensation unit 104. The motion detection unit 103 detects a motion vector in units of frames or fields from the reference image stored in the frame memory 102, and the data representing the motion vector can be transmitted to the motion compensation unit 104. Output to variable length coding section 107.

[0021] The motion compensation unit 104 is based on the reference image stored in the frame memory 102 for encoding the P picture or the B picture and the data representing the motion vector input from the motion detection unit 103! /, Motion compensation data (reference image data) is generated. The bit rate control unit 110 receives the number of generated bits from the variable length code key unit 107, determines a quantization step based on the number of generated bits, and performs this quantization step with the quantization unit 106 and the variable length. The data is transmitted to the code key 107.

The overall control unit 140 is, for example, a microcomputer that includes a ROM, a RAM, and the like, and is a part that controls the entire image coding apparatus 100. The overall control unit 140 controls the processing timing of each unit based on the control signal and the like.

FIG. 4 is a block diagram showing a functional configuration of the image decoding apparatus 200 according to the present invention. The image decoding device 200 represents a scanning method corresponding to a plurality of types of macroblocks included in the code bitstream 121 output from the image coding device 100, at the time of coding. A device that decodes information and performs inverse scanning and decoding of orthogonal transform coefficients using information representing the decoded scanning method, variable length decoding section 201, memory section 235, motion compensation section 204, An adder 212, a frame memory 202, an inverse quantization unit 208, an inverse orthogonal transform unit 209, and the like are provided. Note that the memory unit 235 reads and writes information related to the scan method according to an instruction from the variable length decoding unit 201.

[0024] When the code bit stream 121 is input, the variable length decoding unit 201 receives the quantized DCT coefficient, the motion vector, and the code encoded from the input code bit stream 121. The quantization step and the macroblock type are separated and decoded, and output to the inverse quantization unit 208, motion compensation unit 204, and inverse orthogonal transform unit 209. Also, the scan order based on the macroblock type (frame macroblock or field macroblock) is decoded and stored in the memory unit 235. When decoding the quantized DCT coefficients, the scan order based on the decoded macroblock type is obtained from the memory unit 235, and the variable length of the DCT coefficients quantized in the scan order is obtained. Decrypt.

The motion compensation unit 204 is stored in the frame memory 202 using a macroblock type indicating a motion vector and a frame or a field output from the variable length decoding unit 201. Motion compensation data is generated from the decoded image.

[0026] Inverse quantization section 208 performs inverse quantization on the quantized DCT coefficient in the decoded quantization step to restore the DCT coefficient, and outputs the DCT coefficient to inverse orthogonal transform section 209. The inverse orthogonal transform unit 209 refers to the macroblock type indicating the frame or field, performs inverse frequency conversion from the DCT coefficient to the pixel difference value, and outputs the result to the adder 212. The addition unit 212 adds the pixel difference value and the predicted image output from the motion compensation unit 204 to obtain a decoded image. The decoded image is stored in the frame memory 202 when used for reference in subsequent inter-screen prediction. The decoded image is output to the outside as a decoded image signal 141.

The overall control unit 240 is, for example, a microcomputer that includes a ROM, a RAM, and the like, and is a part that controls the entire image decoding apparatus 200. The overall control unit 240 controls the processing timing of each unit based on the control signal and the like.

FIG. 5 is a flowchart showing a process of encoding a scan pattern in the sequence header 900 of the code bitstream 121 using the image coding apparatus 100 according to the present invention. Note that switching of frames and fields is performed in units of macroblocks. Since the scan sequence of force sequences is switched in units of blocks, it will be referred to as frame blocks and field blocks for convenience in the following description.

First, in step S 300, a progressive-sequence flag is encoded into the sequence header 900. Progressive—When the value of the sequence flag is “1”, this indicates that all the pictures in the video sequence are progressive, ie, frame pictures.

Next, in step S302, seq—adaptive—scanning—flag is encoded in the sequence header 900. seq—adaptive—scanning—flag indicates whether or not a plurality of picture powers in the video sequence are using a scan pattern different from the predetermined scan pattern (for example, when the flag value is “1”, the predetermined value is This flag is used to indicate the case where a scan pattern different from the specified scan pattern is used.

[0031] Next: 5304 [Sequentially, progressive-sequence scanning pattern]. progressive sequence scanning pattern This is information indicating the order of coefficient scan patterns within a frame block, and is encoded with the data structure shown in FIGS. In step S306, progress-sequence scanning pattern is stored in the memory unit 135 in the image encoding device 100 according to the present invention.

[0032] After this, if it is determined in step S310 that the value of progressive-sequence flag is not "1" (that is, if a picture in the video sequence includes a field picture), the image encoding device 100 encodes the field-sequence scanning pattern into the sequence header 900 (step S312). The field—sequence scanning pattern is information indicating the order of the scan patterns of all coefficients in one field block, and is encoded with the data structure shown in FIGS. 13 (a) to (c). Next, the field-sequence scanning pattern is stored in the memory unit 135 (step S314). Finally, in step S318, the image coding apparatus 100 continues coding according to the type of frame block or field block for the pictures in the sequence.

On the other hand, when it is determined in step S310 that the progressive-sequence flag is “1”, the image encoding device 100 encodes all the pictures in the sequence as frame blocks (step S318).

FIG. 6 is a flowchart showing a process of decoding a scan pattern in the sequence header 900 of the code bitstream 121 in the image decoding apparatus 200 according to the present invention.

First, in step 400, the progressive sequence flag in the sequence header 900 is decoded. Next, in step S402, the seq —adaptive—scanning—flag in the sequence header 900 is decoded. Next, in step S404, in sequence header 900

Further, the swap ； 40ο is stored in the memory unit 235 in the decoded progressive_sequence scanning pattern image decoding device 200.

Next, when it is determined in step 410 that the value of the progressive-sequence flag is not “1”, the image decoding apparatus 200 decodes the field-sequence scanning pattern from the sequence header 900 (step S412). This field sequence scanning pattern is In step S414, the data is stored in the memory unit 235. Finally, in step S418, the image decoding apparatus 200 decodes the pictures in the sequence according to the type of frame block or field block.

[0037] If it is determined in step S410 that! /, Progressive- sequence flag is "1"

The image decoding apparatus 200 decodes all the pictures in the sequence as frame blocks (step S418).

FIG. 7 is a flowchart showing a scan pattern encoding process in the picture header 904 of the code bitstream 121 output from the image encoding device 100 of the present invention.

[0039] First, in step S500, a progressive-frame flag is encoded in the picture header 904. When the progressive-frame flag is “1”, it indicates that the picture is a progressive, that is, a frame-coded picture. In this case, the field scanning pattern is not encoded in the picture header 904! /.

[0040] Next, in step S506, the value of seq-adaptive-scanning-flag is determined. This flag is a flag stored in the sequence header 900 as shown in FIG.

[0041] Further, in step S508, if the value of seq—adaptive—scanning—flag is determined to be “1”, pic—adaptive—scanning—one flag (“scan, pattern—update—flag”) Is encoded into the picture header 904 (step S510). This pic-adaptive-scanning-flag is a flag used to indicate whether a new scan pattern is encoded in the picture header 904 of the encoding target picture. If the pic-adaptive-sea nning-flag force is “l” (step S516: Yes), a new scan pattern is encoded in the picture header 904 in step S518.

Next, in step S520, a new scan pattern is stored in the memory unit 135 of the image encoding device 100. Further, in step S522, the image encoding device 100 selects a scan pattern from the memory unit 135, and encodes the slice of the target picture using the selected scan pattern (step S528).

If it is determined in step S516 that the pic-adaptive-scanning-flag is not “1”, the image coding apparatus 100 selects a scan pattern from the memory unit 135 ( In step S526) and step S528, the slice is encoded using the selected scan pattern.

If it is determined in step S508 that seq—adaptive—scanning—flag force ^ 1 ”, the image coding apparatus 100 selects a predetermined scan pattern. In step S514 and step S528, a slice of the target picture is encoded using the selected predetermined scan pattern. Here, examples of the predetermined scan order are shown in FIGS. 4 (a) to 4 (d).

FIG. 8 is a flowchart showing a decoding process of the scan pattern in the picture header 904 of the code bitstream 121 in the image decoding apparatus 200 according to the present invention.

[0046] First, in step S600, a progressive-frame flag is decoded from the picture header 904. When the progressive-frame flag value is “1”, the field scanning pattern is not encoded in the picture header 904! /.

[0047] Next, in step S606, the value of seq—adaptive—scanning—flag is determined. This flag is a flag stored in the sequence header 900 as shown in FIG.

[0048] Thereby, in step S608, when the value of seq—adaptive—scanning-flag is determined to be “1”, pic—adaptive—scanning-flag is decoded from picture header 904 ( Step S610). If the pic-adaptive-scanning-flag is determined to be “1” in step S614, the scan pattern is decoded from the picture header 904 (step S616). In step S618, the decoded scan pattern is stored in the memory unit 235 of the image decoding apparatus 200. Further, in step S620, the image decoding apparatus 200 selects a scan pattern from the memory unit 235, and in step S628, decodes the slice of the target picture using the selected scan pattern.

Note that if it is determined in step S614 that the pic-adaptive-scanning-flag is not “1”, the image decoding apparatus 200 selects a scan pattern from the memory unit 235 and selects the selected scan. Decode the slice using the pattern (step S628).

[0050] On the other hand, if it is determined in step S608 that the seq—adaptive—scanning—flag is not “1”, the image decoding apparatus 200 selects a predetermined scan pattern. ( In step S624), a slice of the current picture is decoded using the selected predetermined scan pattern (step S628). Examples of predetermined scan order are shown in FIGS. 4 (a;) to (d).

FIG. 9 is a flowchart showing a process of encoding a scan pattern into the slice header 908 of the code bitstream 121 in the image coding apparatus 100 according to the present invention.

First, in step S 700, the value of seq—adaptive—scanning—flag is determined.

Next, when the value of seq_adaptive_scanning_flag is “1” (step S702: Yes), slice—adaptive—scanning—flag is encoded into the slice header 908 in step S704. The slice-adaptive-scanning-flag is a flag used to indicate whether a new scan pattern is encoded in the slice header of the target picture. s slice—adaptive—scanning—If the flag is “1” (step S706: Yes), a new scan pattern is encoded in the slice header 908 (step S708).

Next, in step S710, the new scan pattern is stored in the memory unit 135 of the image encoding device 100. Further, in step S712, the image encoding apparatus 100 selects a scan pattern from the memory unit 135, and encodes the macroblock of the target picture using the selected scan pattern (step S720).

In step S 706, if “slice—adaptive—scanning—flag power ^ 1”, the image coding apparatus 100 selects a scan pattern from the memory unit 135 and selects the selected scan. The macroblock is encoded using the Yan pattern (step S720).

On the other hand, if the seq—adaptive—scanning—flag is not “1” in step S702 (step S702: No), the image coding apparatus 100 selects a predetermined scan pattern. (Step S716), and in Step 7S720, the macroblock of the target picture is decoded using the selected predetermined scan pattern. Examples of predetermined scan order are shown in Figures 4 (a) to (d).

FIG. 10 is a flowchart showing a scan pattern decoding process in the slice header 908 of the code bitstream 121 in the image decoding apparatus 200 according to the present invention. [0057] First, in step S800, the value of seq—adaptive—scanning—flag is determined. This flag is a flag stored in the sequence header 900 as shown in FIG. 11 (a).

When the value of seq—adaptive—scanning—flag is “1” (step S802: Yes), slice—adaptive—scanning—flag is decoded from the slice header 908 in step S804. slice—adaptive—scanning—when flag power S is “1” (step S806: Yes), a new scan pattern is decoded from slice header 908 (step S808).

Next, in step S810, the new scan pattern is stored in the memory unit 235 of the image decoding apparatus 200. Further, in step S812, the image decoding apparatus 200 selects a scan pattern from the memory unit 235, and in step S820, decodes the macroblock of the target slice using the selected scan pattern.

In step S806, if “slice—adaptive—scanning—flag power ^ 1” is not satisfied (step S806: No), the image decoding apparatus 200 selects a scan pattern from the memory unit 235. In step S816, the macro block of the target slice is decoded using the selected scan pattern.

On the other hand, if seq_adaptive_scanning_flag is not “1” in step S802 (step S802: No), image decoding apparatus 200 selects a predetermined scan pattern (step S814), In step S820, the macroblock of the target slice is decoded using the selected predetermined scan pattern. Examples of predetermined scan order are shown in Figures 4 (a) to (d).

FIGS. 11 (a) to 11 (c) are diagrams showing the stream structure of the code bit stream 121 output from the image coding apparatus 100 according to the present invention. As shown in this figure, the sequence header 900 in FIG. 10 (a) includes the seq—adaptive—scanning—flag, progressive scan (frame scan) order and field shared by 4 X 4 blocks and 8 X 8 blocks. Stores the scan order. The picture header 904 in FIG. 11 (b), the shared 4 X 4 blocks and 8 X 8 block pic adaptive scanning flagゝoff - Roguretsunpusu ³ r Yanokawa and field scan order is stored. The slice header 908 in Fig. 10 (c) contains a slice adaptive scanning flag shared by 4 X 4 blocks and 8 X 8 blocks. Yan order and field scan order are stored.

[0063] FIGS. 13 (a)-(c) show 4 X 4 blocks and 8 X in sequence header 900, picture header 904 and slice header 908 as shown in FIGS. 11 (a)-(c) above. FIG. 5 is a diagram illustrating a stream structure of a scan pattern for each of 8 blocks. As shown in Fig. 13 (b), the 4 X 4 block scan pattern consists of 16 scan pattern positions. Each scan pattern position is signed by 4 bits, and the position is numbered in the order of the reverse scan pattern. FIG. 13 (a) shows the relationship between the scan pattern position and the reverse scan pattern position. Figure 13 (c) shows a scan pattern for an 8 x 8 block with 64 scan pattern position forces. Each of these positions is signed with 6 bits and arranged in a reverse scan pattern.

[0064] FIG. 14 is a flowchart showing a process of selecting a medium scan pattern of a plurality of predetermined scan patterns in the image coding apparatus 100 and the image decoding apparatus 200 according to the present invention.

[0065] First, in step S1100, the value of progressive-sequence flag is determined. This flag is a flag encoded in the sequence header 900 as shown in FIG. 11 (a). In step S1102, if the progressive-sequence flag is “1” (step S1102: Yes), the progressive scan pattern is selected by the processes shown in steps S1116, S1120, SI 124, and SI 126.

[0066] In step S1102, the progressive-sequence flag is “1”! In the case of / (step S 1 102: No), the value of progressive-frame flag is determined (step S 1104). This flag is a flag encoded in the picture header 904 as shown in FIG. 11 (b) above.

[0067] In step S1106, in the case of progressive-frame flag force ^ 1 "(step SI 106:

Yes), Steps SI 116, S1120, S1124 and S1126 [Process shown] Progressive scan pattern is selected. On the other hand, when the progress ive-frame flag is not “1” in step S 1106 (step S 1106: No), the value of a advanced-pred-mode-disable flag is determined in step S 1108. This flag is a flag stored in the picture header 904, as shown in FIG. 11 (c). [0068] In step SI 110, the advanced-pred-mode-disable flag value is "1"! If / ヽ (step SI 110: No), the value of the macroblock mbp—field—flag is determined (step S1112). The mbp-field-flag is a flag in the macroblock header, and is used to identify whether the macroblock is a frame coding or a field code. mbp

“Field—flag force 0” (step S 1114: Yes), the progressive scan pattern is selected by steps S 1116, S 1120, S 112 4 and SI 126. On the other hand, in the case of mbp-field-flag force 1 "(step S1114: No), the field scan pattern is selected by the process shown in steps S1118, S1122, S1128, and S1130.

[0069] If the value of advanced-pred-mode-disable flag is "1" in step SI 110 (step SI 110: Yes), the steps in steps SI 112 and SI 114 are skipped. Steps S1118, S1122, S1128 and S1130 [shown here] The field scan pattern is selected.

[0070] The process of selecting a predetermined progressive scan pattern is determined as follows. In step S 1116, the macroblock conversion type is determined. As shown in step S1120, in the case of the conversion type force S8 X 8, in step S1126, a predetermined progressive scan pattern force memory unit 135 (or memory unit 235) force for 8 X 8 blocks is also selected. An example of a predetermined progressive scan pattern is a zigzag scan pattern as shown in Fig. 2 (c). If the conversion type is not 8 × 8 in step S1 120, the progressive scan pattern force for the 4 × 4 block S memory unit 135 (or memory unit 235) force is also selected in step S1124. . An example of the predetermined progressive scan pattern is a zigzag scan pattern as shown in FIG.

On the other hand, the process of selecting a predetermined field scan pattern is determined as follows. In step S1118, the conversion type of the macroblock is determined. If the macroblock conversion type is 8 X 8 (step S 1122: No), select the predetermined field scan pattern force memory unit 1 35 (or memory unit 235) for 8 X 8 block in step S 1130. Is done. An example of a predetermined field scan pattern is a field scan pattern as shown in Fig. 2 (d). Also, macroblock If the conversion type is 8 × 8! / (Step S 1122: Yes), step S 1128! /, The predetermined field scan pattern memory unit 135 for 4 × 4 block (or Memory section 235) force is also selected. An example of the predetermined field scan pattern is a field scan pattern as shown in FIG.

FIG. 15 is a flowchart showing a process of selecting a scan pattern from a plurality of scan patterns stored in advance in the image encoding device 100 and the image decoding device 200 of the present invention.

First, in step S 1300, the value of progressive-sequence flag is determined. pro gressive— When the sequence flag is “1” (step IS 1302: Yes), in step SI 316, from multiple progressive scan patterns stored in the memory unit 135 (or memory unit 235, the same shall apply hereinafter) One progressive scan pattern is selected.

On the other hand, when the progressive-sequence flag is not “1” (step IS 1302: No), the value of the progressive-frame flag is determined (step S I 304). As a result, when the progressive-frame flag is “1” (step S 1306: Yes), in step S 1316! /, 1 is selected from the plurality of progressive scan patterns stored in the memory unit 1 35. One progressive scan pattern is selected. When the progressive-frame flag is not “1” (step S 1306: No), the value of advanced-pred-mode-diable flag is determined in step S I 308. As a result, if the value of advanced—pred—mode—diable flag is not “1” (step S1310: No), the value of mb p—field—flag of the macroblock is set to step S1312. Is determined. mbp— field— When flag is “0” (step S 1314: Yes), in step S 1316, one progressive scan pattern is selected from a plurality of progressive scan patterns stored in the memory unit 135. . On the other hand, when the m bp-field-flag is “1” (step SI 314: No), in step SI 318, one field scan pattern is selected from the plurality of field scan patterns stored in the memory unit 135. Selected.

[0075] If the value of advanced-pred-mode-disable flag is "1" in step S1310 (step S1310: Yes), steps S1312 and S1314 are skipped, and in step S1318, the memory unit Multiple field scan patterns stored in 135 One field scan pattern is selected from the screen.

FIG. 16 is a flowchart showing a process for storing a scan pattern in the memory unit of the image coding apparatus 100 and the image decoding apparatus 200 according to the present invention.

First, in step S 1400, the value of progressive-sequence flag is determined. As a result, when the progressive_sequence flag is “1” (step S 1402: Yes), in step S1416, the progressive scan pattern is stored in the memory unit 135 (or the memory unit 235, the same shall apply hereinafter).

On the other hand, when the progressive-sequence flag is not “1” (step S 1402: No), the value of the prog ressive-frame flag is determined (step S 1404). As a result, when the progressive-frame flag is “1” (step S 1406: Yes), the progressive scan pattern is stored in the memory unit 135 in step S 1416! If the progressive-frame flag is not “1” (step S 1406: No), the advanced-pred-mode-diable flag value is determined (step S 1408). As a result, if the value of the advanced-pred-mode-diable flag is not “1” (step S1410: No), the progressive scan pattern is stored in the memory unit 135 in step S1416.

On the other hand, when the value of advanced-pred-mode-disable flag is “1” (step S1410: Yes), the field scan pattern is stored in the memory unit 135 in step S1414.

FIG. 17 shows a process of selecting a scan pattern corresponding to the size of each macroblock from a plurality of scan patterns stored in the memory units of the image encoding device 100 and the image decoding device 200 of the present invention. FIG. 14 is a flowchart (see FIGS. 14 and 15). In the flowchart shown in FIG. 17, when a scan pattern corresponding to the size of each macroblock is selected, the user can select a scan pattern that is specially set.

FIG. 18 is a flowchart showing a process of storing a scan pattern corresponding to the size of each macroblock in the memory units of the image encoding device 100 and the image decoding device 200 of the present invention (FIG. 14). And Figure 16).

In the flowchart shown in FIG. 18, the scan corresponding to the size of each macroblock When storing a pattern, it is possible to store a scan pattern specially set by the user.

[0083] (Embodiment 1)

FIG. 19 is a flowchart showing a process of selecting one progressive scan pattern from a plurality of progressive scan patterns stored in the memory unit of the image encoding device 100 and the image decoding device 200 of the present invention. .

First, in step S 1500, the picture type is determined. Next, if the picture type of the target picture is B picture (step S 1502: Yes), use the process defined in steps S1504, S1506, S1508, and S1510. Select the scan pattern of the previous decoded picture. In addition, if the picture type of the target picture is not a B picture in step S 1502 (step S 1502: No), the process defined in steps S1512, S 1514, S 1516 and S1518 is used. Select the scan pattern for the reference picture.

[0085] An example of a B picture is a picture defined by performing bi-directional prediction, as shown in Non-Reference B Picture I10O in FIG. In this example, the B picture decoded immediately before is a picture shown in Non-Reference B Picture 1008. When the target picture is Reference P Picture 1012 in FIG. 12, if the B picture is not a reference picture, the preceding reference picture is a P picture indicated by Reference Picture 1006 in FIG.

[0086] The process of selecting the scan pattern of the previous decoded picture is as follows. In step S1504, the macroblock conversion type is determined. If the conversion type is 8 X 8 in step S150 6 (step S 1506: Yes), in step S 1510, the scan pattern force memory section of the previous decoded picture for 8 X 8 block 135 (or memory unit 235, the same shall apply hereinafter). If the conversion type is not 8 × 8 (step S 1506: No), in step S 1508, the scan pattern force S of the previous decoding picture is selected from the S memory unit 135 for the 4 × 4 block.

The process for selecting the scan pattern of the immediately preceding reference picture is as follows.

In step S1514, the macroblock conversion type is determined. In step S1514, if the conversion type is 8 X 8 (step S1514: Yes), go to step S1518. And selected from the scan pattern force memory unit 135 of the previous reference picture for the 8 × 8 block. When the conversion type is not 8 × 8 (step S 1514: No), in step S 1516, the scan pattern force memory unit 135 of the immediately preceding reference picture is selected for the 4 × 4 block.

FIG. 20 is a flowchart showing a process for selecting one field scan pattern from among a plurality of field scan patterns stored in the memory units of the image encoding device 100 and the image decoding device 200 of the present invention. It is. This process is the same force as the progressive scan pattern selection process in FIG. 19 above. Scan patterns 1520 to 1526 stored in the memory part 135 (or memory part 235, the same shall apply hereinafter) in FIG. 19 above. The difference is that the scan patterns 162 0 to 1626 included in the memory in FIG. Scan patterns 1620 to 1626 are the field scan pattern of the reference picture immediately before the 4 X 4 block, the field scan pattern of the reference picture immediately before the 8 X 8 block, and the field of the decoding picture immediately before the 4 X 4 block, respectively. The scan pattern and the field scan pattern of the decoded picture immediately before the 8 × 8 block.

FIG. 21 is a flowchart showing a process of storing the progressive scan pattern of the sequence header in the memory unit 135 of the image encoding device 100 and the memory unit 235 of the image decoding device 200 according to the present invention. .

First, in step S 1700, the value of the scan pattern is determined. Next, in step S1702, a check is made to determine whether the video sequence performs adaptive conversion. If the sequence supports both 4 X 4 and 8 X 8 conversion, as shown in step S 1704, first in step S 1708, the new scan pattern memory section 135 for 8 X 8 block It is copied to scan pattern 1718 and scan pattern 1722, where the 8 × 8 block scan pattern for the previous reference picture and the previous decoding picture is stored respectively.

[0091] Next, in step S1710, the new scan pattern is copied to the scan pattern 1716 and scan pattern 1720 of the new scan pattern memory unit 135 for 4 X 4 blocks, where the immediately preceding reference picture and the immediately preceding decoded picture are respectively copied. 4 x 4 block scan pattern for Stored.

[0092] Furthermore, when the sequence only supports 4 × 4 conversion (step S1706: Yes), in step S1712, the scan pattern 1716 and scan pattern of the scan pattern force memory unit 135 for 4 × 4 blocks are scanned. 1720, where the 4 × 4 block scan pattern for the immediately preceding reference picture and the immediately preceding decoded picture is stored, respectively.

[0093] If the sequence corresponds only to 8 X 8 conversion (! / ヽ) (step S1706: No), in step S1714, the scan pattern of the 8X8 block scan pattern force memory unit 135 is scanned. Copied to 1718 and scan pattern 1722, where the 8 × 8 block scan pattern for the previous reference picture and the previous decoded picture, respectively, is stored.

FIG. 22 is a flowchart showing a process of storing the field scan pattern of the sequence header in the memory unit 135 of the image encoding device 100 and the memory unit 235 of the image decoding device 200 according to the present invention. . This process has the same force as the progressive scan pattern storage process in FIG. 21 except that the scan patterns 1716 to 1722 in FIG. 21 are replaced with the scan patterns 1816 to 1822, respectively. Scan patterns 1816 to 1822 are the field scan pattern of the reference picture immediately before the 4 X 4 block, the field scan pattern of the reference picture immediately before the 8 X 8 block, and the field scan of the decoded picture immediately before the 4 X 4 block, respectively. Pattern and field scan pattern of the decoded picture immediately before the 8 x 8 block

FIG. 23 shows the progressive scan pattern encoded into the picture and slice headers in the memory unit 135 of the image encoding device 100 and the memory unit 235 of the image decoding device 200 according to the present invention. It is a flowchart which shows the process to store.

First, in step S1900, the value of the scan pattern is determined. Next, in step S 1902, if the target picture is a P or I reference picture (step S 1902: Yes), the scan pattern block type is determined in step SI 904.

[0097] If the scan pattern block type is 4X4 (step S1906: Yes), In SI 908, they are stored as scan pattern 1920 and scan pattern 1924 in scan pattern value memory unit 135, where the previous reference picture and the previous 4 × 4 block scan pattern for the decoded picture are held, respectively. . On the other hand, when the scan pattern block type is 8 × 8 (step S 1906: No), the process proceeds to step S 1910! / And is stored as scan pattern 1922 and scan pattern 1 926 in the scan pattern value memory unit 135. Therefore, the 8 × 8 block scan pattern for the immediately preceding reference picture and the immediately preceding decoded picture is retained.

If the target picture is not a P or I picture (step S 1902: No), the scan pattern block type is determined in step S 1912. In step S1912, when the scan pattern block type is 4 × 4 (step S 1914: Yes), in step S1916, it is stored as the scan pattern 1924 in the scan pattern value memory unit 135, where the previous decoding is performed. 4 x 4 block scan pattern for digitized pictures is retained. In step S 1912, when the scan pattern block type is 8 × 8 (step S1914: No), in step S1918, the scan pattern value is stored as scan pattern 1926 in the memory unit 135, where the scan pattern value for the previous decoded picture is stored. An 8 x 8 block scan pattern is retained.

FIG. 24 is a flowchart showing a process of storing the field scan pattern encoded in the picture and slice header in the memory unit of the image encoding device 100 and the image decoding device 200 according to the present invention. . This process is the same as the progressive scan pattern storing process in FIG. 23, except that the scan patterns 1920 to 1926 in FIG. 23 are replaced with the scan patterns 2020 to 2026, respectively. Scan patterns 2020 to 2026 are the field scan pattern of the reference picture immediately before the 4 X 4 block, the field of the reference picture immediately before the 8 X 8 block, the scan pattern, and the field of the decoded picture immediately before the 4 X 4 block, respectively. Scan pattern and field scan pattern of decoded picture immediately before 8 X 8 block

[0100] (Embodiment 2)

FIG. 25 shows a memory unit and an image decoding device of the image encoding device according to the present embodiment. 5 is a flowchart showing a process of selecting one progressive scan pattern from a plurality of progressive scan patterns stored in the memory unit of FIG.

[0101] First, in step S2100, the picture type is determined. Next, when the picture type of the target picture is a B picture (step S2102: Yes), the temporal position of the target picture is determined in step S2104. Furthermore, if the target picture is the first reference picture [temporary [closer than this (step S2106: No)], the scan pattern of the second reference picture using the process defined in steps S2108, S2110, S2112 and S2114 is used. Select On the other hand, if the current picture is closer in time to the first reference picture (step S2106: Yes), the scan pattern of the first reference picture is selected using the process defined in steps S2116, S2120, S2122 and S2124. .

[0102] When the target picture is a B picture, for example, when Non-Reference B Picture 1004 in Fig. 12 is the target picture, the first reference picture for the B picture is Reference P Picture 1006 in Fig. 12. The second reference picture for the B picture is Reference I Picture 1000 in FIG. In general, the first reference picture is a reference picture temporally preceding the display time after the target B picture, and the second reference picture is a reference picture temporally subsequent to the target B picture. The first reference picture for P or I picture is the reference picture temporally before the target picture.

[0103] Furthermore, in step S2102, if the target picture is not a B picture (step S2102: No), the scan pattern of the first reference picture is selected using the process defined in steps S2116, S2120, S2122 and S2124. To do.

[0104] The process of selecting the scan pattern of the second reference picture is as follows. First, in step S2108, the macroblock conversion type is determined. If the conversion type is 8 × 8 (step S2108: Yes), the scan pattern force memory unit 135 of the second reference picture is selected for the 8 × 8 block in step S2114. If the conversion type is not 8 × 8 (step S2108: No), in step S2112 the scan pattern power of the immediately preceding decoded picture is selected for the 4 × 4 block S memory unit 2130 power

[0105] The process of selecting the scan pattern of the first reference picture is as follows. Ma In step S2116, the macroblock conversion type is determined. If the conversion type is 8 × 8 (step S2120: Yes), the scan pattern force S of the first reference picture is selected from the S memory unit 135 for the 8 × 8 block in step S2124. On the other hand, if the conversion type is not 8 X 8 (Step S2120: No), in step S2122, the scan pattern force immediately before the reference picture for 4 X 4 blocks S memory unit is selected from: L ₃ 5.

FIG. 26 shows a process of selecting one field scan pattern from a plurality of field scan patterns stored in the memory unit of the image encoding device and the memory unit of the image decoding device according to the present embodiment. It is a flowchart to show. This process is the same as the progressive scan pattern selection process in FIG. 25, except that the scan patterns 2126 to 2132 in FIG. 25 are replaced with scan patterns 2226 to 2232, respectively. The scan patterns 2226 to 2232 are respectively a 4 X 4 block first reference picture field scan pattern, an 8 X 8 block first reference picture field scan pattern, and a 4 X 4 block second reference picture field scan. Pattern and 8 x 8 block second reference picture field scan pattern.

FIG. 27 is a flowchart showing a process of storing the progressive scan pattern of the sequence header in the memory unit of the image coding apparatus and the memory unit of the image decoding apparatus according to the present embodiment. First, in step S2300, the value of the scan pattern is determined. Next, in step S2302, a check is made to determine if the video sequence undergoes adaptive conversion. If the sequence corresponds to both 4 X 4 and 8 X 8 conversions as shown in step S 2304 (step S 2304: Yes), first in step S2308, for the 8 X 8 block New scan patterns are copied as scan patterns 2318 and 2322 in the S memory section, where 8 X 8 block scan patterns for the first reference picture and the second reference picture are stored, respectively. Next, in step S2310, it is copied as scan patterns 2316 and 2330 in the new scan pattern memory section for 4 X 4 blocks, where the 4 X 4 block scan patterns for the first reference picture and the second reference picture respectively. Stored.

[0108] On the other hand, if the sequence only supports 4 X 4 conversion (step S2306: Yes) ), Copied in step S2312, as scan patterns 2316 and 2320 in the scan pattern force memory for 4 X 4 blocks, where the 4 X 4 block scan patterns for the first reference picture and the second reference picture are stored, respectively. Is done.

[0109] If the sequence only supports 8 X 8 conversion! / ヽ (step S2306: No), in step S2314, the scan patterns 2318 and 2322 in the scan pattern force memory unit for 8 X 8 blocks are used. Where 8 × 8 block scan patterns for the first and second reference pictures are stored, respectively.

FIG. 28 is a flowchart showing a process of storing the field scan pattern of the sequence header in the memory unit of the image encoding device and the memory unit of the image decoding device according to the present embodiment. This process is the same as the progressive scan pattern storage process in Fig. 27, except that scan patterns 2316 to 2322 in the memory unit in Fig. 27 are replaced with scan patterns 2416 to 2422 in the memory unit, respectively. . The scan patterns 2416 to 2422 in the memory part are respectively a field scan pattern of the first reference picture of 4 × 4 blocks, a field scan pattern of the first reference picture of 8 × 8 blocks, and a second reference picture of 4 × 4 blocks. Field scan pattern and 8 x 8 block second reference picture field scan pattern.

FIG. 29 shows the progressive scan pattern encoded in the picture and slice headers stored in the memory unit of the image encoding device and the memory unit of the image decoding device according to the present embodiment. It is a flowchart which shows the process to perform.

[0112] First, in step S2500, the value of the scan pattern is determined. At this time, if the target picture is a P or I reference picture (step S2502: Yes), the scan pattern block type is determined in step S2504. Further, in step S2506, when the scan pattern block type is 4 × 4 (step S2506: Yes), in step S2508, the scan pattern 2516 of the first reference picture in the memory unit is changed to the second reference picture in the memory unit. Copied to the scan pattern 2520. Then, the scan pattern value of the new 4 × 4 block is stored in the scan pattern of the first reference picture in the memory unit in step S2510. In step S2506, If the Yan pattern block type is 8 X 8 (step S2506: No), the scan pattern 2518 of the first reference picture in the memory unit is changed to the scan pattern 2522 of the second reference picture in the memory unit in step S2512. Copied. Then, in the new 8 × 8 block scan pattern value step S2514, it is stored in the scan pattern 2518 of the first reference picture in the memory unit.

[0113] FIG. 30 shows a process of storing the field scan pattern encoded in the picture and slice header in the memory unit of the image encoding device and the memory unit of the image decoding device according to the present embodiment. It is a flowchart. This process is the same as the progressive scan pattern storage process in Fig. 29 above.The scan patterns 2516 to 2522 in the memory unit in Fig. 25 are replaced by the scan patterns 2616 to 2622 in the memory unit, respectively. Is different. The scan patterns 2616 to 2622 in the memory section are respectively a field scan pattern of the first reference picture of 4 X 4 blocks, a field scan pattern of the first reference picture of 8 X 8 blocks, and a field of the second reference picture of 4 X 4 blocks. A scan pattern and a field scan pattern of the second reference picture of 8 x 8 blocks.

[0114] (Embodiment 3)

FIG. 31 shows a process of selecting one field scan pattern from a plurality of field scan patterns stored in the memory unit of the image encoding device and the memory unit of the image decoding device according to the present embodiment. It is a flowchart which shows another embodiment of this.

[0115] First, in step S2700, the picture type of the target picture is determined. At this time, if the target picture is a B picture (step S2702: Yes), a check is performed in step S2704 to determine whether the target macroblock includes only the top field sample. If the target macroblock does not include a top field sample (step S2704: No), the bottom field scan pattern of the previous decoded picture is selected using the process defined in steps S2706 to S2714. On the other hand, if the target macroblock includes a top field sample (step S2704: Yes), the process defined in steps S2710 to S2720 is used to execute the previous macroblock. Select the top field scan pattern of the decoded picture.

[0116] If the target picture is not a B picture in step S2702 (step S2702: No), a check is made in step S2722 to determine whether the target macroblock contains only top field samples. In step S2722, if the target macroblock contains a top field sample (step S2722: No), the bottom field of the previous reference picture using the process defined in steps S2732 to 2738 is used. Select a scan pattern. If the target macroblock includes a top field sample in step S2722 (step S2722: Yes), the top field scan pattern of the immediately preceding reference picture is selected using the process defined in steps S2724 to S2730.

[0117] The process of selecting the bottom field scan pattern of the previous decoded picture is as follows.

[0118] First, in step S2706, the conversion type of the macroblock is determined. If the conversion type is 8 × 8 (step S2708: Yes), the scan pattern force S of the 8 × 8 block is selected from the scan pattern 2756 in the memory unit in step S2714. On the other hand, if the conversion type is not 8 × 8 (step S2708: No), the scan pattern 2752 in the 4 × 4 block scan pattern force memory unit is selected in step S2712.

[0119] The process of selecting the top field scan pattern of the previous decoded picture is as follows. In step S2710, the macroblock conversion type is determined. If the conversion type is 8 × 8 (step S2716: Yes), a scan pattern of 8 × 8 blocks is selected from the scan pattern 2754 in the memory unit in step S2720. On the other hand, if the conversion type is not 8 × 8 (step S2716: No), a scan pattern of 4 × 4 blocks is selected from the memory unit 2750 in step S2718.

[0120] The process of selecting the bottom field scan pattern of the immediately preceding reference picture is as follows. In step S2732, the macroblock conversion type is determined. If the block conversion is 8 x 8 (step S2734: Yes), the scan pattern of 8 x 8 block is selected from the scan pattern 2748 in the memory in step S2738. Selected. On the other hand, if the conversion type is not 8 × 8 (step S2734: No), a scan pattern of 4 × 4 blocks is selected from the scan pattern 2744 in the memory unit in step S2736.

[0121] The process of selecting the top field scan pattern of the immediately preceding reference picture is as follows. In step S2724, the macroblock conversion type is determined. If the block conversion is 8 × 8 (step S2726: Yes), a scan pattern of 8 × 8 block is selected from the scan pattern 2746 in the memory unit in step S2728. On the other hand, when the block conversion is not 8 × 8 (step S2726: No), the scan pattern 2742 in the scan pattern force memory unit of 4 × 4 block is selected in step S2730.

FIG. 32 is a flowchart showing a process of storing the progressive scan pattern of the sequence header in the memory unit of the image coding apparatus and the memory unit of the image decoding apparatus according to the present embodiment.

[0123] First, in step S2800, the value of the scan pattern is determined. Next, in step S2802, a check is made to determine if the video sequence performs adaptive conversion. If the sequence supports both 4 X 4 and 8 X 8 conversion as shown in step S2804, first in step S2806, a new scan pattern force for 8 X 8 block ^ Scan patterns 2818, 2822, 28 26, and 2830, where the top field of the immediately preceding reference frame, the bottom field of the immediately preceding reference frame, the top field of the immediately preceding decoded frame, and the bottom field of the immediately preceding decoded frame, respectively. 8 x 8 block scan pattern for is stored.

[0124] Next, in step S2808, the data is copied to scan patterns 2816, 2820, 2824, and 2828 in the new scan pattern memory section for the 4 X 4 block, where the top field of the reference frame immediately before and the previous one respectively. The 4 × 4 block scan pattern for the bottom field of the reference frame, the top field of the immediately preceding decoded frame, and the bottom field of the immediately preceding decoded frame is stored.

[0125] As shown in step S2810, when the sequence only supports 4 X 4 conversion Are copied in step 2812 to scan patterns 2818, 2820, 2824 and 2828 in the new scan pattern force memory section for the 4 X 4 block, where the top field of the immediately preceding reference frame and the bottom field of the immediately preceding reference frame, respectively. Stores the 4 X 4 block scan pattern for the top field of the previous decoded frame and the bottom field of the previous decoded frame.

[0126] As shown in step S2810, if the sequence only supports 8 x 8 conversion, in step S2814, a new scan pattern for the 8 x 8 block is first created in the scan pattern in the memory card. 2818, 2822, 2826, and 2830, where the top field of the immediately preceding reference frame, the bottom field of the immediately preceding reference frame, the top field of the immediately preceding decoded frame, and the bottom field of the immediately preceding decoded frame, respectively. 8 x 8 block scan patterns for are stored.

FIG. 24 shows the field scan pattern encoded in the picture and slice header, the memory unit of the image encoding device according to the present embodiment and the memory unit of the restored image signal device. It is a flowchart which shows the process stored in. In the present embodiment, the field parity is stored in the field parity memory location corresponding to the new scan pattern force in the picture or slice header according to the target field picture nature.

For example, when the target picture is a top field picture, a new scan pattern is stored in the memory unit including the scan pattern for the top field of the immediately preceding decoded frame.

[Embodiment 4]

FIG. 33 shows a plurality of field scan patterns stored in the memory unit (not shown) of the image coding apparatus and the memory unit (not shown) of the image decoding apparatus according to the present embodiment. 6 is a flowchart showing another embodiment of a process for selecting one field scan pattern.

[0130] First, in step S2900, the picture type of the target picture is determined. Next, when the target picture is a B picture in step S2902 (step S2902: Yes), the temporal position of the target picture is determined in step S2904. If the target picture is not closer in time to the first reference picture in step S2906, In step S2908, a check is made to determine if the target macroblock contains only the top field sample. If the target macroblock does not include a top field sample (step S2908: No), the bottom field scan pattern of the second reference frame is selected using the process defined by steps S2918 to S2924. On the other hand, if the target macroblock contains a top field sample (step S2908: Yes), the top field scan pattern of the second reference frame is selected using the process defined in steps S2910 to S2916. .

[0131] If it is determined in step S2906 that the target picture is closer to the first reference picture in time (step S2906: Yes), or if it is not the target picture B picture in step S2902, (step S2902). : No), in step S2926! /, A check is made to determine if the target macroblock contains only the top field sample. If the target macroblock does not include a top field sample (step S2926: No), the bottom field scan pattern of the first reference frame is selected using the process defined by steps S2936 to S2942. On the other hand, if the target macroblock includes a top field sample (step S2926: Yes), the top field scan pattern of the first reference frame is selected using the process defined by steps S2928 to S2934.

[0132] The process of selecting the bottom field scan pattern of the second reference frame is as follows. First, in step S2918, the macroblock conversion type is determined. If the conversion type is 8 × 8 (step S1920: Yes), in step S2924, a scan pattern of 8 × 8 blocks is selected from the scan pattern 2958 in the memory unit.

On the other hand, when the conversion type is not 8 × 8 (step S1920: No), a scan pattern of 4 × 4 blocks is selected from the scan pattern 2954 in the memory unit in step S2922.

[0134] The process of selecting the top field scan pattern of the second reference frame is as follows. First, in step S2910, the macroblock conversion type is determined. Determined. If the block conversion is 8X8 (step S2912: Yes), the scan pattern of the 8X8 block is selected from the scan pattern 2956 in the memory unit in step S2916. On the other hand, if the block conversion is not 8X8 (step S2912: No), in step S2914, the scan pattern 2952 force in the memory unit is selected as the scan pattern of the 4X4 block.

[0135] Further, the process of selecting the bottom field scan pattern of the first reference frame is as follows. First, in step S2936, the macroblock conversion type is determined. If the conversion type is 8X8 (step S2938: Yes), in step 2942, the scan pattern of the 8X8 block and the scan pattern 2950 in the memory unit are also selected. On the other hand, if the conversion type is not 8X8 (step S2938: No), the scan pattern of 4X4 block is selected from the scan pattern 2946 in the memory unit in step S2940.

[0136] Further, the process of selecting the top field scan pattern of the first reference frame is as follows. First, in step S2928, the macroblock conversion type is determined. If the conversion type is 8X8 (step S2930: Yes), in step S2934, the scan pattern of the 8X8 block and the scan pattern 2948 in the memory unit are also selected. On the other hand, when the conversion type is not 8X8 (step S2930: No), in step S2932, the scan pattern of 4X4 block is selected from the scan pattern 2944 in the memory unit.

[0137] Fig. 34 stores the progressive scan pattern of the sequence header in the memory unit (not shown) of the image coding apparatus according to the present embodiment and the memory unit (not shown) of the image decoding device. It is a flowchart which shows a process.

[0138] First, in step S3000, the value of the scan pattern is determined. Next, in step S3002, a check is performed to determine if the video sequence performs adaptive conversion. If the sequence supports both 4X4 and 8X8 conversions as shown in step S3004 (step S3004: Yes), first in step S3006 a new 8 x 8 block Scan pattern force Copied to scan patterns 3020, 3022, 3028 and 3030 in the S memory, where the first reference frame 8x8 block scan patterns for the top field of the frame, the bottom field of the first reference frame, the top field of the second reference frame, and the bottom field of the second reference frame are stored.

[0139] Next, in step S3008, it is copied to scan patterns 3016, 3018, 3024, and 3026 in the new scan pattern memory section for the 4 X 4 block, where the top field and the first reference frame of the first reference frame respectively. Stores 4 x 4 block scan patterns for the bottom field of the 1 reference frame, the top field of the 2nd reference frame, and the bottom field of the 2nd reference frame.

[0140] If the sequence only supports 4 X 4 conversion (step S3010: Yes), in step S3012, the scan pattern 3016, 3018 in the new scan pattern force memory unit for 4 X 4 blocks 3024 and 3026 copies, where there are 4 X 4 block scans for the top field of the first reference frame, the bottom field of the first reference frame, the top field of the second reference frame and the bottom field of the second reference frame, respectively. A pattern is stored.

[0141] On the other hand, if the sequence only supports 8 × 8 conversion (step S3010: No), in step S3014, a new scan pattern for the 8 × 8 block is first scanned with the scan pattern 3020 in the memory unit. Copied to 3022, 3028 and 3030, where 8 X 8 for the top field of the first reference frame, the bottom field of the first reference frame, the top field of the second reference frame and the bottom field of the second reference frame, respectively Stores the block scan pattern.

FIG. 30 is a flowchart showing a process of storing the field scan pattern encoded in the picture and slice header in the memory unit of the picture encoding / decoding apparatus according to the present invention. In the present embodiment, a new scan pattern in the picture or slice header is stored in the memory location of the corresponding field parity depending on the parity of the target field picture. For example, when the target picture is a top field reference picture, a new scan pattern is stored in the memory unit including the scan pattern for the top field of the first reference frame.

It should be noted that in the above embodiment, a complete (that is, 16 in the case of a 4 × 4 pixel macroblock). (In the case of a macro block of 8 × 8 pixels, there are 64 scan orders.) Although a scan pattern is used, a part of the scan pattern may be omitted. In this case, scanning is performed according to a predetermined scanning method.

Industrial applicability

INDUSTRIAL APPLICABILITY The present invention can be used as an encoding device and a decoding device for multimedia data, and in particular, an image encoding device and an image decoding that perform scanning to make a two-dimensional orthogonal transform coefficient one-dimensional. Available to the device.

Claims

The scope of the claims

[1] An image encoding device for encoding a picture in units of macroblocks,

Determining means for determining whether or not to encode using a macroblock of at least two types when encoding a picture;

When the determination means determines that encoding is to be performed using the two or more macroblocks, information indicating one scan method selected from a plurality of scan methods for each type of macroblock As the header information, and based on the selected scanning method, the orthogonal transform coefficients of the two-dimensional array of macroblocks corresponding to each scanning method are encoded in the order of the one-dimensional array. Sign

An image encoding device comprising:

[2] An image decoding device that decodes a code bitstream having header information in units of macroblocks,

The header information of the code key stream is decoded, and from the decoded header information, orthogonal transform coefficients of a two-dimensional array are converted into a one-dimensional array for at least two types of macroblocks constituting a picture. Information indicating the scan method for rearrangement is extracted, and using the information indicating the scan method, the orthogonal transform coefficient of the code bitstream encoded in the one-dimensional array order is decoded and 2 Decoding means for converting to a dimensional array

An image decoding apparatus characterized by that.

[3] The header information includes information indicating the scan method corresponding to the frame macroblock, and information indicating the scan method corresponding to the field macroblock,

The determination means determines whether or not to use a frame macroblock and a field macroblock when coding a picture.

When the determination unit determines that the code should be encoded using the frame macroblock and the field macroblock, the decoding unit stores the frame macroblock or the field macroblock included in the header information. The pair corresponding to Extract information indicating the can method

The image decoding apparatus according to claim 2, wherein:

[4] The header information includes information indicating a scan method corresponding to the block size of the frame macroblock or the block size of the field macroblock,

The decoding means further comprises:

Extracts information indicating a scanning method corresponding to the block size included in the header information

The image decoding apparatus according to claim 2, wherein:

[5] The header information includes information indicating a prediction method used by the macroblock,

The decoding means further comprises:

Extracts information indicating a scanning method corresponding to the prediction method included in the header information

The image decoding apparatus according to claim 2, wherein:

[6] The header information includes information indicating a scan method corresponding to the block size of the frame macroblock or the block size of the field macroblock,

The decoding means further comprises:

The image decoding apparatus according to claim 2, wherein:

[7] The header information further includes information indicating that a part of the information indicating the scan method is omitted,

The decoding means further comprises:

It is determined whether a part of the information indicating the scanning method is omitted or not! If it is determined that the omitted power S is present, it is replaced with a predetermined scanning method.

The image decoding apparatus according to claim 2, wherein:

[8] An image encoding method for encoding a picture in units of macroblocks, and whether or not to encode using at least two or more types of macroblocks when encoding a picture. Judgment,

If it is determined that the two or more types of macroblocks should be used for coding, one of the multiple scanning methods is selected for each type of macroblock. Indicating the header information as header information, and, based on the selected scanning method, the one-dimensional array of orthogonal transformation coefficients of the two-dimensional array of macroblocks corresponding to each scanning method. Line up

An image encoding method characterized by the above.

[9] An image decoding method for decoding a code bitstream having header information in units of macroblocks,

The header information of the code key stream is decoded, and from the decoded header information, orthogonal transform coefficients of a two-dimensional array are converted into a one-dimensional array for at least two types of macroblocks constituting a picture. Extract information indicating the scan method for sorting, and use the information indicating the scan method to convert the orthogonal transformation coefficient of the one-dimensional array into a two-dimensional array.

An image decoding method characterized by the above.

[10] A program for causing a computer to execute a picture used in an image coding apparatus for coding a picture in units of macroblocks,

The program is

A determination step for determining whether or not to encode using a macroblock of at least two types when encoding a picture; and

When the determination means determines that encoding is to be performed using the two or more macroblocks, information indicating one scan method selected from a plurality of scan methods for each type of macroblock Is encoded as header information, and the orthogonal transform coefficients of the two-dimensional array of macroblocks corresponding to each scan method are encoded in the order of the one-dimensional array based on each of the selected scan methods. Encoding step and The program characterized by including.

[11] An image for decoding a coded bitstream having header information in units of macroblocks, a program for causing a computer to execute the program,

The header information of the code key stream is decoded, and from the decoded header information, orthogonal transform coefficients of a two-dimensional array are arranged in a one-dimensional arrangement order for at least two types of macroblocks constituting a picture. The information indicating the scan method for rearranging the encoded signal bitstream encoded in step (b) is extracted, and the orthogonal transform coefficient of the one-dimensional array is decoded into the two-dimensional array using the information indicating the scan method. Contains decryption steps to convert

A program characterized by that.

[12] An integrated circuit for encoding a picture in units of macroblocks,

Determination means for determining whether or not to encode using a macroblock of at least two or more types when encoding a picture;

When the determination means determines that encoding is to be performed using the two or more macroblocks, information indicating one scan method selected from a plurality of scan methods for each type of macroblock Is encoded as header information, and the orthogonal transform coefficients of the two-dimensional array of macroblocks corresponding to each scan method are encoded in the order of the one-dimensional array based on each of the selected scan methods. An integrated circuit comprising: encoding means.

[13] An integrated circuit that decodes a code bitstream having header information in units of macroblocks,

The header information of the code key stream is decoded, and from the decoded header information, orthogonal transform coefficients of a two-dimensional array are arranged in a one-dimensional arrangement order for at least two types of macroblocks constituting a picture. The information indicating the scan method for rearranging the encoded signal bitstream encoded in step (b) is extracted, and the orthogonal transform coefficient of the one-dimensional array is decoded into the two-dimensional array using the information indicating the scan method. Provide decryption means to convert An integrated circuit characterized by that.