JP2007020002A - Motion picture encoder - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a motion picture encoder capable of concealing decoding errors of coding data belonging to a character multiplex area while maintaining transmission efficiency of motion picture encoding data and image quality in decoding successful. <P>SOLUTION: The motion picture encoder 1 which converts, quantizes motion picture data representing a motion picture having the character multiplex area and performs variable length coding to the motion picture data has: a variable length coding part 6 which performs variable length coding to the motion picture data performs motion picture data belonging to a redundant area overlapped with the character multiplex area along outer edges of the character multiplex area more redundantly than in the case of motion picture data belonging to areas other than the redundant area and codes motion picture data belonging to each normal slice by unit of normal slice by dividing the motion picture like a belt; and a redundant slice setting part 4 which sets a redundant slice overlapped with the character multiplex area along the outer edge of the character multiplex area, wherein the variable length coding part 6 further codes motion picture data belonging to the redundant slice. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a moving image encoding apparatus that encodes moving image data, and more particularly, to a moving image encoding apparatus that converts and quantizes moving image data representing a moving image having a character multiplex region and performs variable length encoding. .

  In recent years, moving picture coding apparatuses are widely used in videophone systems, video distribution systems (streaming), and the like.

  FIG. 17 is a block diagram showing a configuration of a conventional moving picture coding apparatus 90. As shown in FIG. The moving image encoding device 90 includes a conversion unit 97. The converting unit 97 converts the input moving image data into moving image data in the frequency domain by DCT or the like and outputs the moving image data to the quantizing unit 98. The quantization unit 98 quantizes the converted moving image data based on the quantization parameter set by the rate control unit 99 and outputs the quantized moving image data to the variable length coding unit 96. The rate control unit 99 sets a normal slice by dividing the moving image into strips, and changes the quantization parameter value. The variable length coding unit 96 performs variable length coding on the quantized moving image data belonging to each normal slice in units of normal slices obtained by dividing the moving image into strips, and outputs the encoded bit stream.

  Even if an error occurs in one normal slice, it does not affect the decoding of other normal slices (here, it means variable-length decoding), so the normal slice is used, for example, for the purpose of enhancing error tolerance. used.

  In an application that assumes an environment with many errors, error resilience is enhanced by increasing the number of normal slice divisions. On the other hand, in an application that assumes an environment with few errors (or no errors), the overhead caused by normal slice division is reduced by reducing the number of normal slice divisions.

  When the generated code amount is predicted to exceed the target bit rate, the quantization parameter value is increased to suppress the generated code amount. Conversely, if the generated code amount is predicted to be significantly below the target bit rate, the quantization parameter value is decreased in order to increase the generated code amount.

  When an error is detected in a certain frame or slice during decoding by the decoding device, concealment processing such as replacing the error part with another image is performed. For example, when an error is detected in a certain frame, a method of concealing the error part using a frame decoded immediately before is disclosed (for example, see Patent Document 1).

  In addition, as a technique for preventing image quality degradation on the encoding device side, when encoding a video signal containing a lot of high-frequency components, such as a caption scene, between intra-frame encoded images and inter-frame encoded images. A technique for preventing the image quality from being deteriorated due to the difference in expression of the high-frequency component is disclosed (for example, see Patent Document 2).

  Furthermore, various error tolerances such as resynchronization marker (Resync Marker), data partitioning (Data Partitioning), RVLC, etc. are provided in order to enhance tolerance against transmission errors that occur in the transmission path between the encoding device and the decoding device. A tool has been introduced (see Non-Patent Document 1, for example).

Further, a redundant slice, an arbitrary slice order, and the like have been proposed. Redundant slices are not always encoded, but when moving image data belonging to a macroblock included in a normal slice cannot be correctly decoded due to a transmission error or the like, a moving image belonging to a macroblock included in the normal slice The error can be concealed by replacing the data with moving image data belonging to the macroblock included in the redundant slice set for the entire normal slice.
Japanese Patent Laid-Open No. 9-93589 (published April 4, 1997) Japanese Patent Laid-Open No. 10-32816 (published February 3, 1998) ISO / IEC 14496-2 Draft ISO / IEC 14496-10 Draft

  However, in the above conventional decoding device or encoding device, an error is mixed in the transmission path in the encoded data of the character multiplex area where an object such as a caption scene represented by a video signal containing many high-frequency components is displayed. Thus, when a decoding error occurs, there is a problem that proper concealment cannot be performed. This will be specifically described below.

  For example, in the configuration of Patent Document 1 described above, in the concealment process of the character multiplex area in which a decoding error has occurred, if the character string of the macro block of the character multiplex area in the frame decoded immediately before is not a corresponding character string, There is a problem that a correct character string cannot be displayed.

  In addition, since a video signal representing a character or the like displayed in the character multiplex area contains many high-frequency components, there is a problem that the character multiplex area in which a decoding error has occurred is difficult to conceal with peripheral pixels of the current frame.

  Patent Document 2 only discloses a technique for preventing the image quality from being deteriorated due to the difference between the intra-frame encoded image and the inter-frame encoded image. It does not describe concealing a decoding error caused by this, and does not show recognition of the problem to be solved by the present invention.

  Further, in the configuration in which the decoding error of the character multiplex area is concealed by the redundant slice as described above, since the redundant slice is set in the entire normal slice including the character multiplex area, the transmission code amount is increased and the transmission path having a narrow band is formed. Then, there is a problem that transmission efficiency deteriorates. Further, since the transmission code amount is increased by the redundant slice as described above, there is a problem that the code amount allocated to the normal slice is decreased and the image quality is deteriorated.

  The present invention has been made in view of the above-described conventional problems, and an object of the present invention is to perform encoding belonging to a character multiplex region while maintaining good transmission efficiency of moving image encoded data and image quality at the time of decoding. An object of the present invention is to provide a moving picture coding apparatus capable of concealing a data decoding error.

  In order to solve the above problems, a moving image encoding apparatus according to the present invention is a moving image encoding apparatus that converts and quantizes moving image data representing a moving image having a character multiplex region, and performs variable length encoding. The moving image data belonging to the redundant area overlapping the character multiplexed area along the outer edge of the character multiplexed area is more variable-length encoded than the moving image data belonging to an area other than the redundant area. .

  According to the above configuration, the moving image data belonging to the redundant area that overlaps the character multiplexed area along the outer edge of the character multiplexed area is more variable-length encoded than the moving image data belonging to the area other than the redundant area. For this reason, overhead (additional processing) due to the redundancy of variable-length coding is reduced compared to the conventional configuration in which the entire moving image data of the normal slice including the character multiplex area is redundantly variable-length coded. can do. As a result, the transmission code amount of moving image encoded data can be reduced, and the code amount allocated to the normal slice can be secured. Therefore, it is possible to conceal the decoding error of the encoded data belonging to the character multiplex area while maintaining good transmission efficiency of moving image encoded data and image quality at the time of decoding.

  In the moving image encoding apparatus of the present invention, variable length encoding means for encoding moving image data belonging to each normal slice in units of normal slices obtained by dividing the moving image into strips, along the outer edge of the character multiplex region And redundant slice setting means for setting redundant slices overlapping the character multiplex region, and the variable length encoding means preferably further encodes moving image data belonging to the redundant slice.

  According to this configuration, the redundant slice that overlaps the character multiplex area along the outer edge of the character multiplex area is set by the redundant slice setting means, and the moving image data belonging to the redundant slice is further encoded. For this reason, it is possible to encode moving image data in a character multiplex area constituting a part of a normal slice more redundantly than moving image data belonging to an area other than the character multiplex area. Compared to the conventional configuration in which the entire slice is encoded as a redundant slice, overhead (additional processing) due to the redundant slice can be reduced. As a result, the transmission efficiency can be improved by reducing the transmission code amount of the moving picture coding apparatus. Further, it is possible to provide a good image quality by securing the amount of code allocated to the normal slice.

  In the moving image encoding device of the present invention, the moving image encoding device further includes character multiplexed region detection means for detecting the character multiplexed region from the moving image and generating character multiplexed region information, and the redundant slice setting means includes the character multiplexed region detection It is preferable to set the redundant slice based on the character multi-region information generated by the means.

  According to this configuration, when the character multiplex area is not known in advance, the character multiplex area information can be obtained by detecting the character multiplex area from the moving image, and the redundant slice is obtained based on the character multiplex area information. Can be set.

  In the moving image encoding apparatus of the present invention, transmission means for transmitting the moving image data belonging to the normal slice encoded by the variable length encoding means to a decoding device, and moving image data belonging to the normal slice Receiving means for receiving a decoding error notification of moving image data belonging to the character multiplex region, and the variable length encoding means, when the decoding error notification is received by the receiving means, to the redundant slice It is preferable that the moving image data belonging thereto is encoded, and the transmission unit transmits the moving image data belonging to the redundant slice encoded by the variable length encoding unit to the decoding device.

  According to this configuration, the moving image data in the redundant slice is encoded when a decoding error is notified. For this reason, when no decoding error is notified, the overhead due to the redundant slice can be further reduced by configuring so that the moving image data in the redundant slice is not encoded. Further, since a larger amount of code can be allocated to the normal slice, it is possible to provide a better image quality.

  In the moving image encoding device of the present invention, the variable length encoding means encodes moving image data belonging to the normal slice having the redundant slice before moving image data belonging to the normal slice not having the redundant slice. The transmission means encodes the moving picture data belonging to the normal slice having the redundant slice encoded by the variable length coding means before the moving picture data belonging to the normal slice not having the redundant slice. It is preferable to transmit to the decoding device.

  According to this configuration, it is possible to determine at an earlier point in time whether or not to receive a decoding error notification of moving image data belonging to the character multiplex region and whether to transmit moving image data belonging to a redundant slice. The actual transmission path has a limited bandwidth of, for example, 64 kbps. When a redundant slice is not transmitted without receiving a decoding error notification in a certain frame, the code amount of the redundant slice is encoded in the next frame. Therefore, it is possible to improve the image quality as a whole.

  In the moving picture coding apparatus of the present invention, the redundant slice setting means sets a slice group that is an area obtained by dividing the moving picture so as to include the character multiplexed area, and the redundant slice is determined based on the slice group. It is preferable to set.

  According to this configuration, even when the character multiplex area has a vertically long shape and when the character multiplex area has a complicated shape, an appropriate shape according to the shape of the character multiplex area. By setting this slice group, it is possible to easily set one redundant slice along the outer edge of the character multiplex area.

  The moving image encoding apparatus of the present invention further comprises object encoding means for encoding moving image data belonging to a video object for the entire moving image, and the object encoding means targets the character multiplexed region. It is preferable to further encode the moving image data belonging to the character multi-region video object.

  According to this configuration, the moving image data belonging to the character multiplex area video object for the character multiplex area is further encoded. For this reason, moving image data belonging to the character multiplex region can be encoded more redundantly than moving image data belonging to regions other than the character multiplex region, and therefore, the entire normal slice including the character multiplex region is made a redundant slice. Compared to the conventional configuration of encoding, overhead due to redundant encoding of moving image data in a character multiplex region can be reduced.

  In addition, a redundant slice composed of a plurality of macroblocks is applied to a rectangular character multiplex region. However, since a video object uses each object constituting the screen as an encoding unit, The present invention can be applied not only to a character multiplex region but also to an arbitrarily shaped character multiplex region.

  As described above, the moving image encoding apparatus according to the present invention allows moving image data belonging to a redundant area that overlaps the character multiplexed area along the outer edge of the character multiplexed area to be different from moving image data belonging to an area other than the redundant area. Redundant variable-length encoding is performed.

  For this reason, compared to the conventional configuration in which the entire moving image data of the normal slice including the character multiplex area is redundantly variable-length encoded, overhead due to the redundancy of variable-length encoding can be reduced. In addition, the transmission code amount of moving image encoded data can be reduced, and the code amount allocated to the normal slice can be secured.

  Therefore, it is possible to provide a moving picture coding apparatus capable of concealing a decoding error of coded data belonging to a character multiplex area while maintaining good transmission efficiency of moving picture coded data and image quality at the time of decoding. There is an effect.

  In this specification, a “slice” refers to an area having a plurality of macroblocks (blocks of 16 × 16 pixels) and serving as a unit of encoding processing, and normally reaches the right end of a picture (screen). Continuing to the left end one line down. “Normal slice” refers to a slice obtained by dividing one picture (screen) in GOP (Group of Pictures). A “redundant slice” refers to a slice for encoding a part or all of a picture (screen) separately from a normal slice.

(Embodiment 1)
An embodiment of the present invention will be described with reference to FIGS. 1 to 11 as follows. The moving image encoding apparatus 1 includes a character multiple region detection unit 2. The character multiple region detection unit 2 detects the character multiple region based on moving image data representing a moving image based on the YUV 4: 2: 0 format, for example, and generates character multiple region information 3. 3 is output to the redundant slice setting unit 4 together with the moving image data.

  The redundant slice setting unit 4 sets a redundant slice that overlaps the character multiplex area along the outer edge of the character multiplex area represented by the character multiplex area information 3, and generates redundant slice position information 5 that specifies the position of the redundant slice. At the same time, the moving image data received from the character multiple region detection unit 2 is output to the conversion unit 7 in units of macroblocks.

  The conversion unit 7 converts the moving image data into frequency-domain moving image data by DCT or the like and supplies the converted data to the quantization unit 8. The moving image encoding apparatus 1 includes a rate control unit 9. The rate control unit 9 sets a quantization parameter for the quantization unit 8 and sets a normal slice obtained by dividing the moving image represented by the moving image data in a band shape.

  The quantization unit 8 quantizes the converted moving image data according to the quantization parameter set by the rate control unit 9 and supplies the quantized moving image data to the variable length coding unit 6. The variable length encoding unit 6 encodes moving image data belonging to each normal slice in units of normal slices set by the rate control unit and provides the encoded bit stream to the transmission unit 10, while the redundant slice setting unit 4 The moving image data belonging to the redundant slice specified by the generated redundant slice position information 5 is further encoded and provided to the transmission unit 10 as an encoded bit stream.

  The transmission unit 10 transmits the moving image data belonging to each normal slice encoded by the variable length encoding unit 6 to the decoding device 17 and further transmits the moving image data belonging to the redundant slice. The moving image encoding apparatus 1 is provided with a receiving unit 11. The receiving unit 11 receives a decoding error notification of moving image data belonging to the character multiplex area from the decoding device 17.

  The operation of the moving picture encoding apparatus 1 configured as described above will be described. FIG. 2 is a flowchart showing the operation of the moving image encoding apparatus 1. First, the character multiple region detection unit 2 detects a character multiple region based on moving image data representing a moving image (step S21).

  FIG. 3A is a diagram showing the moving image 12 input to the character multi-region detection unit 2, and FIG. 3B is a diagram showing the character multi-region information 3 generated by the character multi-region detection unit 2. .

  For example, the character multiplex area detection unit 2 converts the input moving image 12 into frequency domain data, and detects a macroblock 13 containing a lot of high frequency components as a character multiplex area. The broken line in the figure indicates the boundary of the macro block 13 composed of 16 × 16 pixels. Here, the character multiplex area information is an attribute value set for each macro block 13, and is “1” (characters are included in the macro block 13) or “0” (characters are not included in the macro block 13). Take. The character multiple region detection unit 2 determines whether or not a character is included in each macroblock 13 and sets “1” or “0” as the attribute value of the macroblock 13. Then, the attribute values of all the macroblocks 13 are collectively output to the redundant slice setting unit 4 as the character multiplex area information 3. In the example shown in FIGS. 3A and 3B, the attribute value of the macro block 13 including the character string “Aiueo Kakikuke” at the bottom of the screen is “1”, and the attribute values of the other macro blocks 13 are “0”. It becomes.

  Next, the redundant slice setting unit 4 sets a redundant slice that overlaps the character multiplex area along the outer edge of the character multiplex area specified by the character multiplex area information 3 (FIG. 2: step S22), and the position of this redundant slice. Redundant slice position information 5 for identifying is generated.

  FIG. 4A is a diagram showing the moving image 12 input to the redundant slice setting unit 4, and FIG. 4B is a diagram for explaining the redundant slice 15 set by the redundant slice setting unit 4. .

  An arrow 14 shown in FIG. 4A indicates the order in which the macroblocks 13 constituting the moving image 12 are encoded. As indicated by the arrow 14, in the first embodiment, encoding is performed in the raster scan order from the macroblock 13 at the upper left of the moving image 12. The redundant slice setting unit 4 outputs the moving image data in units of macroblocks 13 to the conversion unit 7 in accordance with the encoding order of the macroblocks 13 indicated by the arrow 14, and the character multiplexed region output from the character multiplexed region detection unit 2 Based on the information 3 (FIG. 3B) and the coding order of the macroblock 13 indicated by the arrow 14 (FIG. 4A), the redundant slice 15 is configured by only the character multiplex area. The start macroblock and the end macroblock of the redundant slice in the multi-region are determined.

  Specifically, the character multiplex area information 3 in FIG. 3B is scanned in accordance with the encoding order indicated by the arrow 14 in FIG. 4A, and the macroblock 13 satisfying a predetermined condition is selected as the start macro of the redundant slice. Block or end macroblock. Then, the macro block number of the start macro block and the macro block number of the end macro block are notified as the redundant slice position information 5 to the variable length coding unit 6. Here, the macroblock number is a number assigned with 0, 1, 2,... In the order of raster scanning from the upper left macroblock of the moving image 12, as shown in FIG. In the example shown in the first embodiment, the encoding order of the macroblock 13 matches the order of the macroblock numbers, but the encoding order of the macroblock 13 is the macro as in the case of the arbitrary slice order described later. The order of block numbers may be different.

  FIG. 5 is a flowchart showing the operation of the redundant slice setting unit 4 and shows a processing flow showing a redundant slice setting method in the character multiplex area. In FIG. 5, i indicates the scanning order of the macroblock 13, MB (i) indicates the i-th scanned macroblock 13, and attr (X) indicates the Xth-scanned macroblock 13 attribute value ( Character multi-region information).

  First, an initial scan macroblock (MB (0)) is set (i = 0) (step S51). Next, the attribute value (attr (MB (i−1))) = “0” of the macro block (MB (i−1)) to be scanned i−1 and “ith scan” It is determined whether or not the attribute value (attr (MB (i))) = “1” of the macro block (MB (i)) to be executed (step S52). If true (YES in step S52), MB (i) is set as the start macroblock of the redundant slice (step S53).

  The attribute value (attr (MB (i−1))) = “0” of the “i−1th scanned macroblock (MB (i−1))” and the “ith scanned macroblock (MB (i)) attribute value (attr (MB (i))) is not “1” (NO in step S52), or MB (i) is set as the start macroblock of the redundant slice (step In S53), the attribute value (attr (MB (i))) = “1” of the i-th scanned macroblock (MB (i)) and “i + 1th scanned macroblock ( It is determined whether or not the attribute value (attr (MB (i + 1))) = “0” of MB (i + 1)) (step S54). If true (YES in step S54), MB (i) is set as the end macroblock of the redundant slice (step S55).

  “Attribute value (attr (MB (i))) =“ 1 ”of the i-th scanned macroblock (MB (i))” and “i + 1th scanned macroblock (MB (i + 1))” If the attribute value (attr (MB (i + 1))) is not "0" (NO in step S54), or if MB (i) is set as the end macroblock of the redundant slice (step S55), Is set (i ++) (step S56). Then, it is determined whether MB (i) is the last macroblock of the moving image (step S57). If MB (i) is the last macroblock of the moving image (YES in step S57), the process ends. If not (NO in step S57), the process returns to step S52 to continue the process.

  When the redundant slice setting method is applied to the character multiplex area information 3 shown in FIG. 3B, the macroblock number 79 becomes the start macroblock of the redundant slice 15 as shown in FIG. 85 macroblocks become the end macroblock of the redundant slice 15, and seven macroblocks numbered 79, 80, 81, 82, 83, 84, and 85 become one redundant slice 15. The start macroblock number = 79 and the end macroblock number = 85 of the redundant slice 15 are notified to the variable length coding unit 6 as redundant slice position information 5 (FIG. 1).

  In this way, when moving image data in a character multiplex area is encoded as a redundant slice, the redundant slice is set so as to overlap the character multiplex area along the outer edge of the character multiplex area. For this reason, the moving image data to be encoded by the redundant slice can be limited to only the moving image data of the macro block belonging to the character multiplex region, and therefore the overhead due to the redundant slice can be minimized.

  FIG. 6 is a diagram showing the normal slices 16a, 16b, 16c, 16d, 16e, 16f, and 16g and the redundant slice 15 that are units of coding by the variable length coding unit 6. The rate control unit 9 sets normal slices 16a, 16b, 16c, 16d, 16e, 16f, and 16g obtained by dividing the moving image 12 into strips. The normal slices 16a, 16b, 16c, 16d, 16e, 16f, and 16g are encoded by the variable length encoding unit 6 in this order from the upper left of the moving image 12 in the raster scan order. The redundant slice 15 is set by the redundant slice setting unit 4 in a part of the normal slice 16e. The variable length encoding unit 6 performs variable length encoding on the moving image data quantized by the quantization unit 8 in units of normal slices set by the rate control unit 9, and provides the encoded bit stream to the transmission unit 10 ( FIG. 2: Step S23).

  Data necessary for encoding in units of normal slices by the variable length encoding unit 6 includes a start code located at the head of the normal slice, a macroblock number at the head of the normal slice, and a normal slice type (I slice, P slice, B slice etc.). Data required for encoding processing in units of macroblocks in encoding processing in units of normal slices includes macroblock types (intra coding mode and inter coding mode), motion vectors (in the case of inter coding), and each block. Quantized data and the like. Details are described in Non-Patent Document 1 and Non-Patent Document 2.

  Then, it is determined whether or not the redundant slice 15 is set in the variable-length encoded normal slice (FIG. 2: step S24). If it is determined that the redundant slice 15 has been set (YES in step S24), the variable length coding unit 6 performs variable length coding on the moving image data belonging to the redundant slice 15 as a coded bit stream, and the transmission unit 10 (FIG. 2: Step S25).

  When it is determined that the redundant slice 15 is not set (NO in step S24), or when the redundant slice 15 is variable-length encoded (step S25), the codes of all normal slices in the moving image (one frame) It is determined whether or not conversion is completed (frame end) (step S26). If it is not the frame end (NO in step S26), the process returns to step S23. If it is the frame end (YES in step S26), the transmission unit 10 uses the variable-length encoding unit 6 to perform variable-length encoding of the moving image data and the variable-length encoding unit 6 to change the variable length. The encoded moving image data belonging to the redundant slice is transmitted to the decoding device 17 (step S27).

  Then, it is determined whether or not all frames have been encoded (sequence end) (step S28). If it is determined that it is not the sequence end (NO in step S28), the process returns to step S21. If it is determined that the sequence is ended, the entire process is terminated.

  Although FIG. 2 illustrates an example in which one redundant slice is always included in one normal slice, the present invention is not limited to this. One redundant slice may be arranged so as to straddle a plurality of normal slices. In this case, it is assumed that the same redundant slice is not duplicated multiple times and encoded in step S25.

  FIG. 7 is a flowchart showing a modified example of the operation of the moving image encoding apparatus 1. In the flowchart of FIG. 2, the same components as those described above are denoted by the same reference numerals, and detailed description thereof is omitted.

  As described above with reference to FIG. 2, the redundant slice set by the redundant slice setting unit 4 is not unconditionally encoded, but only when a decoding error is notified by an external means such as the decoding device 17. Video data belonging to a redundant slice may be encoded.

  First, a character multiplex area is detected by the character multiplex area detection unit 2 (step S21), and a redundant slice is set by the redundant slice setting unit 4 (step S22). Then, the variable length encoding unit 6 performs variable length encoding on the moving image data in units of normal slices (step S23). Next, it is determined whether or not it is a frame end (step S26). If it is not a frame end (NO in step S26), the process returns to step S23. If it is the frame end (YES in step S26), one frame of moving image data encoded in units of normal slices is transmitted to the decoding device (step S71).

  Thereafter, it is determined whether or not a decoding error notification of moving image data belonging to the character multiplex area among moving image data belonging to the normal slice transmitted to the decoding device has been received by the receiving unit 11 (step S72).

  When it is determined that a decoding error notification has been received (YES in step S72), the variable length encoding unit 6 encodes moving image data belonging to the redundant slice set by the redundant slice setting unit 4, and transmits the transmission unit 10 Transmits the moving image data belonging to the redundant slice encoded by the variable length encoding unit 6 to the decoding device 17 (step S73).

  Then, it is determined whether or not it is a sequence end (step S28). If it is determined that it is not a sequence end (NO in step S28), the process returns to step S21, and if it is determined that it is a sequence end (in step S28). YES), the entire process is terminated.

  According to this configuration, when no decoding error occurs, moving image data belonging to a redundant slice is not encoded, so a larger amount of code is allocated to moving image data encoded in units of normal slices. be able to. For this reason, the image quality of the moving image decoded by the decoding device can be improved.

  FIG. 8A is a diagram illustrating another example of a moving image input to the redundant slice setting unit 4, and FIG. 8B illustrates another example of the redundant slice set by the redundant slice setting unit 4. It is a figure for doing. FIG. 9 is a diagram for explaining normal slices encoded in an arbitrary order. As in FIG. 6, normal slices 16a, 16b, 16c, 16d, 16e, 16f, and 16g obtained by dividing the moving image 12a into strips are arranged in the raster scan order from the upper left of the moving image 12a. It is encoded and transmitted from a normal slice. In the example shown in FIG. 9, the redundant slice 15a is set as the normal slice 16e, and the redundant slice 15b is set as the normal slice 16b. First, the normal slices 16e and 16b are encoded and transmitted in this order, and then the normal slices 16a, 16c, 16d, 16f, and 16g are encoded and transmitted in this order. That is, the moving image data belonging to the normal slices 16e and 16b each having the redundant slices 15a and 15b is encoded and transmitted before the moving image data belonging to the normal slice not having the redundant slices 15a and 15b. When a slice is encoded, the first macroblock number of the slice is encoded in the slice header, so that even if a normal slice is encoded and transmitted in an arbitrary order, it can be correctly decoded.

  FIGS. 8A and 8B show an example of a moving image 12a in which two character multiplexed areas are detected. The normal slice 16b in which the redundant slice 15b including the character multiplex area for displaying the character string “Kakikakeko” is set and the character string “Aiueo” using the arbitrary slice order that encodes the normal slice according to the arbitrary order. The normal slice 16e in which the redundant slice 15a including the character multiplex area to be displayed is set is encoded and transmitted before the other normal slices.

  FIG. 10 is a flowchart showing another modified example of the operation of the moving picture encoding apparatus 1, and FIG. 11 is a diagram showing transmission / reception timing between the moving picture encoding apparatus 1 and the decoding apparatus 17.

  First, a character multiplex area is detected by the character multiplex area detection unit 2 (step S21), and redundant slices 15a and 15b are set by the redundant slice setting unit 4 (step S22).

  Then, the normal slice 16e including the redundant slice 15a corresponding to the character multiplex area and the normal slice 16b including the redundant slice 15b corresponding to the other character multiplex area are variable before the other normal slices. Encoding is performed by the long encoding unit 6 to generate encoded data 18 and 19, and the encoded data 18 and 19 are transmitted to the decoding device 17 by the transmission unit 10 (step S101).

  Next, it is determined whether or not the reception unit 11 has received the decoding result signals 20 and 21 including the decoding error notification of the moving image data belonging to the character multiplex region of the normal slice 16e or the normal slice 16b from the decoding device 17 (Step S1). S102). When it is determined that the decoding error notification has been received (YES in step S102), the variable length coding unit converts the moving image data belonging to the redundant slice 15a set to the normal slice 16e or the redundant slice 15b set to the normal slice 16b. 6 is generated, and the encoded data 22 and 23 are transmitted to the decoding device 17 by the transmission unit 10 (step S103).

  When it is determined that a decoding error notification is not received (NO in step S102), or when moving image data belonging to the redundant slice 15a or the redundant slice 15b is encoded and transmitted (step S103), the remaining normal slices 16a and 16c are transmitted. The moving image data belonging to 16d, 16f, and 16g is encoded and transmitted to the decoding device 17 (step S104).

  Next, it is determined whether or not it is the frame end (step S26). If it is not the frame end (NO in step S26), the process returns to step S101. If it is a frame end (YES in step S26), it is determined whether or not it is a sequence end (step S28). If it is determined that it is not the sequence end (NO in step S28), the process returns to step S21. If it is determined that the sequence is ended (YES in step S28), the entire process is terminated.

  In this way, by decoding and encoding the moving image data belonging to the character multiplex region by encoding and transmitting the normal slice including the character multiplex region before the normal slice not including the character multiplex region using the arbitrary slice order. Whether or not an error notification is received and whether or not moving image data belonging to the redundant slice is transmitted to the decoding device can be determined at an earlier time point. For this reason, when a redundant slice is not transmitted without receiving a decoding error notification in a certain frame, the code amount of the redundant slice can be allocated to encoding of the next frame, and the overall image quality is improved. be able to.

  In addition, this invention is not limited to said embodiment, A various change is possible within the scope of the present invention. For example, in the above-described embodiment, the example in which the character multiplex area detection unit 2 detects the character multiplex area has been described, but the present invention is not limited to this. In a movie, news subtitle, or the like, if the position to be displayed is known in advance, the attribute value of the macroblock at the predetermined position may be set to “1” without performing the detection process. Alternatively, the character string detection processing may be directly input from the outside to the redundant slice setting unit 4 without performing the character string detection process in the moving image encoding device.

  Alternatively, assuming that the same image data is input for several frames in a region where characters are multiplexed, a macroblock having a small difference from the previous frame may be detected as a character multiplexed region.

  In the first embodiment, an example in which one redundant slice 15 is set in one character multiplex area has been described. However, one redundant slice 15 is further divided into a plurality of redundant slices and set. Also good. If the number of divisions is increased by dividing into a plurality of redundant slices, error tolerance can be further enhanced.

  Further, for example, when the character multiplex area detection unit 2 determines that the macro block number 82 in FIG. 4B does not include a character string, that is, when the macro block attribute value of the number 82 is “0”. In this example, one redundant slice is composed of macroblocks numbered 79, 80, and 81, and one redundant slice is composed of macroblocks numbered 83, 84, and 85. However, the present invention is not limited to this. Not. In order to avoid this, the processing in step S52 and step S54 in FIG. 5 may be performed as follows.

  That is, whether or not “i-Nth to i−1th scanned macroblock attribute value =“ 0 ”” and “ith scanned macroblock attribute value =“ 1 ””. (N is an integer of 2 or more). Thus, the process of step S52 is changed, and “the attribute value of the i-th scanned macroblock =“ 1 ”” and “the attribute value of the macroblock scanned from the (i + 1) th to i + Nth =“ It is determined whether it is 0 "" (N is an integer of 2 or more). As described above, the process of step S54 is changed.

  As a result, it is possible to prevent the overhead from being increased by dividing the character multiplex area into a plurality of unintended redundant slices.

  In addition, when the encoding method does not support redundant slices according to the standard, or when redundant slices cannot be used because they are supported by the standard but cannot be used in operation, etc. With regard to, all the macro blocks of the normal slice including the character multiplex area may be intra-encoded (a mode in which encoding is performed without referring to an already-encoded image).

(Embodiment 2)
The following will describe another embodiment of the present invention with reference to FIGS. The schematic configuration of the moving picture coding apparatus according to the second embodiment is as shown in FIG. 1. Since the configuration other than the redundant slice setting unit 4 is the same as that of the first embodiment, description thereof is omitted.

  FIG. 12A is a diagram showing a slice group set by the redundant slice setting unit 4 provided in another embodiment of the moving image encoding apparatus, and FIG. 12B is another example of the slice group. FIG.

  The slice group is a unit for dividing into redundant slices, and is composed of a plurality of macro blocks. 12A and 12B are diagrams showing examples of slice groups. Referring to FIG. 12A, the moving image 12b is divided into left and right regions by a bent line that extends vertically downward from the upper side of the moving image 12b, bends horizontally to the left at a substantially central portion, and further bends vertically downward to reach the lower side. The left area is a slice group 24a, and the right area is a slice group 24b. FIG. 12B shows a method in which two rectangular slice groups of a slice group 25a and a slice group 25b are arranged in the moving image 12b, and the rest is a background slice group 25c.

  The redundant slice setting unit 4 first sets a slice group based on the character multiple region information 3 output from the character multiple region detection unit 2. Next, based on the set slice group, a redundant slice is set in the character multiplex area by the method described in Embodiment 1, and moving image data is output to the conversion unit 7 in units of macroblocks.

  FIG. 13A is a diagram showing a moving image 12b input to the redundant slice setting unit 4, and FIG. 13B is a diagram for explaining slice groups 24a and 24b set by the redundant slice setting unit 4. FIG. 13C is a diagram for explaining the redundant slice 26 set based on the slice group 24a.

  On the left side of the moving image 12b input to the redundant slice setting unit 4, as shown in FIG. 13 (a), two lines of character strings “Aiue Kakikakeko” are arranged in the vertical direction. As shown in FIG. 13B, the redundant slice setting unit 4 sets a slice group 24a, which is an area obtained by dividing the moving image 12b so as to include a character multiplexed area of the character string “Aiue Kakikukeko”.

  Next, based on the information indicating which macroblock the slice group 24a is composed of, the macroblock numbers 0, 1, 2, 11, 12, 13, 22, 23, 24, and 33 included in the slice group 24a are displayed. For the macroblocks of 34, 35, 44, 45, 55, 56, 66, 67, 77, 78, 88, and 89, the character multiplexing region is obtained by the method described with reference to FIG. A redundant slice 26 is set (numbers in FIG. 9C indicate macroblock numbers).

  Here, the macroblocks with the macro block numbers 12, 13, 22, 23, 24, 33, 34, 35, 44, 45, 55, and 56 that contain the character string “Aiueo Kakikukiko” have one redundant slice 26. The redundant slice setting unit 4 outputs the redundant slice position information 5 to the variable length encoding unit 6 so as to be configured. As a result, the redundant slice setting unit 4 determines the slice group type (information indicating which macroblock the slice group 24a is composed of), the start macroblock number of the redundant slice 26, and the end of the redundant slice 26. The variable length encoding unit 6 is notified of the macro block number = 56.

  When the slice group is not used as in the first embodiment, the numbers 12, 13, 22, 23, 24, 33, 34, 35, 44, 45, and 55 included in the redundant slice 26 in FIG.・ Since 56 macroblocks cannot form one redundant slice, all of the macroblocks of numbers 12, 13, 14,. Redundant slice consisting of macroblocks numbered 12 and 13; redundant slice B consisting of macroblocks numbered 22, 23 and 24; and redundancy consisting of macroblocks numbered 33, 34 and 35 Slice C, redundant slice D composed of macroblocks numbered 44 and 45, redundant slice E composed of macroblocks numbered 55 and 56, It is necessary to divide the redundant slice 26 into a plurality of redundant slice.

  On the other hand, in the second embodiment, one redundant slice 26 composed of macroblocks of numbers 12, 13, 22, 23, 24, 33, 34, 35, 44, 45, 55, and 56 is used by the slice group 24a. Therefore, the amount of codes can be reduced as compared with the configuration in which all macroblocks from No. 12 to No. 56 are made one redundant slice, and the redundant slices A, B, C, D, and E can be reduced to a plurality of redundant slices. The amount of codes can also be reduced compared with the configuration in which the redundant slice 26 is divided. For this reason, even if the character multiplex area has a complicated shape as shown in FIG. 13A, encoding can be performed with a small number of redundant slices to be set, and as a result, the code amount of the redundant slice is suppressed. Can improve the overall image quality.

  FIG. 14 is a flowchart showing the operation of the video encoding apparatus using slice groups. In the flowchart of FIG. 2, the same components as those described above are denoted by the same reference numerals, and detailed description thereof is omitted. First, the character multiplex area detector 2 detects the character multiplex area of the moving image 12b shown in FIG. 13A (step S21). Then, the slice group 24a shown in FIG. 13B is set by the redundant slice setting unit 4 (step S181). Next, the redundant slice 26 shown in FIG. 13C is set based on the slice group 24a (step S182).

  Thereafter, the variable length encoding unit 6 performs variable length encoding on the moving image data in units of normal slices (step S23). Then, it is determined whether or not it is the frame end (step S26). If it is not the frame end (NO in step S26), the process returns to step S23. If it is the frame end (YES in step S26), the variable length coding unit 6 performs variable length coding on the moving image data belonging to the redundant slice 26 (step S183). Then, the moving image data variable-length encoded in normal slice units and the variable-length encoded moving image data belonging to the redundant slice 26 are transmitted to the decoding device 17 (step S184). Next, it is determined whether or not it is a sequence end (step S28). If it is determined that it is not the sequence end (NO in step S28), the process returns to step S21. If it is determined that the sequence is ended, the entire process is terminated.

  In addition, although the example in which the redundant slice setting unit sets the slice group is shown, the slice group may be configured to be selected from a plurality of predetermined slice group groups.

  In the first embodiment and the second embodiment described above, the example in which the moving image data in the character multiplex area is redundantly encoded by the redundant slice has been described, but the present invention is not limited to this. For example, the moving image data in the character multiplex area may be redundantly encoded by the object encoding method.

  FIG. 15 is a diagram for explaining the configuration of still another embodiment of the moving picture coding apparatus. The video data belonging to the character multiplex area may be redundantly encoded in units of video objects defined in MPEG-4. In MPEG-4, a plurality of video objects (VO) can be stored in one video sequence (video screen flow). In the example shown in FIG. 15, the video object 28 that covers the entire frames 27a, 27b, 27c, 27d, and 27e, and the character multiplexing areas 29a, 29b, 29c, 29d, and 29e included in each frame are targeted. The video object 30 to be stored is stored in the video sequence. The object encoding unit provided in the moving image encoding apparatus encodes moving image data belonging to the video object 28 for the entire frame and moving image data belonging to the video object 30 for the character multiplexed region. Can be encoded redundantly for moving image data belonging to a character multiplex region. In addition, the shape of the character multiplex area | region used as the object of a video object may be a rectangular shape, and arbitrary shapes may be sufficient as it.

  FIG. 16 is a block diagram illustrating a configuration of a moving image encoding apparatus 1a that encodes a video object. The same components as those described above are denoted by the same reference numerals, and detailed description thereof will be omitted. The moving image encoding apparatus 1a encodes the character multiplexed region detecting unit 2 that detects a character multiplexed region from a moving image and the video object 28 (frames 27a, 27b, 27c, 27d, and 27e) shown in FIG. 15 is encoded based on the character multiplex area detected by the VO encoding unit 161 of 1 and the character multiplex area detection unit 2 (character multiplex areas 29a, 29b, 29c, 29d, and 29e) shown in FIG. The second VO encoding unit 162 that performs the processing, the encoded data encoded by the first VO encoding unit 161 and the encoded data encoded by the second VO encoding unit 162 are transmitted to the decoding device And a transmission unit 163.

  In the first embodiment and the second embodiment described above, an example in which a redundant slice is set in the character multiplex area has been shown. However, the present invention is not limited to the character multiplex area, and the same processing is performed for an important area in the image. It is also applicable when For example, the present invention can be applied to a portion that is more important than the background region, such as a human face portion.

  Even if an error is mixed in the encoded data of the person's face part, as in the case where an error is mixed in the character multiplex area, if the macroblock at the same position in the previous frame is not the corresponding person's face part, it will be displayed correctly In addition, since the human face portion has a problem that it is difficult to substitute the peripheral pixels of the current frame, if the present invention is applied to the human face portion, the effect of applying the present invention to the character multiplex region Similar effects can be obtained.

  The present invention is not limited to the above-described embodiments, and various modifications can be made within the scope of the claims, and the technical means disclosed in different embodiments can be appropriately combined. Such embodiments are also included in the technical scope of the present invention.

  INDUSTRIAL APPLICABILITY The present invention can be applied to a moving image encoding apparatus that encodes moving image data, and in particular, a moving image that converts and quantizes moving image data representing a moving image having a character multiplex region and performs variable length encoding. It can be applied to an image encoding device. Specifically, it can be used for a videophone system and a video distribution system (streaming), and can also be used for a receiver that receives a television broadcast, encodes the video, and wirelessly transmits it to a portable display unit. Can be used.

It is a block diagram which shows the structure of one Embodiment of the moving image encoder in this invention. It is a flowchart which shows operation | movement of the said moving image encoder. (A) is a figure which shows the moving image input into the character multiplex area | region detection part of the said moving image encoder, (b) is a figure which shows the character multiplex area | region information which the said character multiplex area detection part produces | generates. . (A) is a figure which shows the moving image input into the redundant slice setting part of the said moving image encoder, (b) is a figure for demonstrating the redundant slice set by the said redundant slice setting part. . It is a flowchart which shows operation | movement of the said redundant slice setting part. It is a figure which shows the normal slice and redundant slice which become a unit of the encoding by the variable length encoding part of the said moving image encoder. It is a flowchart which shows the modification of operation | movement of the said moving image encoder. (A) is a figure which shows the other example of the moving image input into the redundant slice setting part of the said moving image encoder, (b) is another example of the redundant slice set by the said redundant slice setting part. It is a figure for demonstrating. It is a figure for demonstrating the said normal slice encoded by arbitrary orders. It is a flowchart which shows the other modification of operation | movement of the said moving image encoder. It is a figure which shows the transmission / reception timing between the said moving image encoder and decoding apparatus. (A) is a figure which shows the slice group set by the redundant slice setting part provided in other embodiment of the moving image encoder of this invention, (b) is another example of the said slice group. FIG. (A) is a figure which shows the moving image input into the said redundant slice setting part, (b) is a figure for demonstrating the slice group set by the said redundant slice setting part, (c) is said figure It is a figure for demonstrating the redundant slice set based on a slice group. It is a flowchart which shows operation | movement of the moving image encoder of said other embodiment. It is a figure for demonstrating the structure of further another embodiment of the moving image encoder of this invention. It is a block diagram which shows the structure of the said moving image encoder. It is a block diagram which shows the structure of the conventional moving image encoder.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Moving image encoding apparatus 2 Character multiple area detection part (character multiple area detection means)
3 Character multiple area information 4 Redundant slice setting part (Redundant slice setting means)
5 Redundant slice position information 6 Variable length coding unit (variable length coding means)
10 Transmission section (Transmission means)
11 Receiver (Receiving means)
12, 12a, 12b Video 15, 15a, 15b Redundant slices 16a, 16b, 16c, 16d, 16e, 16f, 16g Normal slice 17 Decoding device 24a, 24b, 25a, 25b, 25c Slice group 26 Redundant slice 28 Video object 30 Video object (character multi-region video object)

Claims (7)

  A moving image encoding device that converts and quantizes moving image data representing a moving image having a character multiplex region, and performs variable length encoding, wherein the redundant region overlaps the character multiplex region along an outer edge of the character multiplex region A moving picture encoding apparatus characterized in that the moving picture data belonging to is encoded in a variable length more redundantly than moving picture data belonging to a region other than the redundant region. Variable length encoding means for encoding moving image data belonging to each normal slice in units of normal slices obtained by dividing the moving image into strips;
Redundant slice setting means for setting a redundant slice that overlaps the character multiplex region along the outer edge of the character multiplex region;
2. The moving picture coding apparatus according to claim 1, wherein the variable length coding means further codes moving picture data belonging to the redundant slice. A character multiple region detection means for detecting the character multiple region from the moving image and generating character multiple region information;
The moving picture encoding apparatus according to claim 2, wherein the redundant slice setting unit sets the redundant slice based on the character multiple region information generated by the character multiple region detection unit. Transmission means for transmitting moving image data belonging to the normal slice encoded by the variable-length encoding means to a decoding device;
Receiving means for receiving a decoding error notification of the moving image data belonging to the character multiplex region of the moving image data belonging to the normal slice;
The variable length encoding means encodes moving image data belonging to the redundant slice when the decoding error notification is received by the receiving means,
The moving picture encoding apparatus according to claim 2, wherein the transmission means transmits the moving picture data belonging to the redundant slice encoded by the variable length encoding means to the decoding apparatus. The variable length encoding means encodes moving image data belonging to a normal slice having the redundant slice before moving image data belonging to a normal slice not having the redundant slice,
The transmission means decodes the moving picture data belonging to the normal slice having the redundant slice encoded by the variable length coding means before the moving picture data belonging to the normal slice having no redundant slice. The moving picture coding apparatus according to claim 2, which is transmitted to the apparatus.   The moving picture code according to claim 2, wherein the redundant slice setting means sets a slice group that is an area obtained by dividing the moving picture so as to include the character multiplex area, and sets the redundant slice based on the slice group. Device. An object encoding means for encoding moving image data belonging to a video object for the entire moving image;
2. The moving image encoding apparatus according to claim 1, wherein the object encoding means further encodes moving image data belonging to a character multi-region video object for the character multi-region.

Images