JP2009278512A - Moving image decoding apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a moving image decoding apparatus preventing a collapsed image from being displayed, even when image size information is not appropriately reported, under the situation where encoded data of moving images in different image sizes may be received in the middle of moving images. <P>SOLUTION: The moving image decoding apparatus includes an image size determination section 9 for discriminating whether or not the number of decoded macro blocks at a point of time, when one frame of a video stream is decoded by a variable length decoder 1, corresponds to an image size indicated by image size information, and for storing an alternative image in a display memory 8 in place of a decoded image generated by an adder 5 if they do not correspond to each other. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  In the present invention, when decoding encoded data of a moving image that has been digitally compressed and encoded to generate a decoded image, a moving image that can cope with the intermittent switching of the image size of the moving image The present invention relates to a decoding device.

Currently, in broadcasting / communication systems that distribute moving images using digital broadcasting (satellite, terrestrial, cable) or the Internet, international standard video codes such as “MPEG” and “ITU-T H.26x” are used. Is used.
In such a broadcasting / communication system, on the transmission side, the image size may be changed in the middle of a moving image due to switching of programs having different image sizes, even within one channel.
Also, on the receiving side, switching between a plurality of different channels may result in receiving encoded data of moving images having different image sizes in the middle of the moving image.

When the image size of a moving image is changed, image size information (image size change information) is notified from the transmitting side encoding device to the receiving side decoding device, and the receiving side decoding device receives the image size information. By receiving, decoding processing can be performed with an appropriate image size.
However, when a situation occurs in which the image size information is not properly notified to the decoding device on the receiving side, the decoding device does not appropriately change the image size, and performs the decoding process with the image size before the change.
In this case, the decoded image by the decoding device is a corrupted image because the screen position is shifted.

The following Patent Document 1 discloses a device for appropriately processing an image size change on the encoding device side.
As a result, even when the image size is changed in the middle of a moving image due to switching of programs with different image sizes, it is possible to prevent the display of a corrupted decoded image, but the channel can be switched on the receiving side. The situation cannot be dealt with.

JP-A-11-96412 (paragraph number [0007], FIG. 1)

  Since the conventional moving image decoding apparatus is configured as described above, encoded data of moving images having different image sizes in the middle of moving images by switching between different channels on the receiving side. When the image size information is not properly notified to the receiving side in a situation where the received image is received, the screen position is shifted, and thus there is a problem that the decoded image may be corrupted.

  The present invention has been made to solve the above-described problems, and appropriately notifies image size information in a situation where encoded data of a moving image having a different image size is received in the middle of the moving image. An object of the present invention is to obtain a moving picture decoding apparatus capable of preventing display of a corrupted image even when it is not performed.

  The moving picture decoding apparatus according to the present invention determines whether or not the correspondence relationship between the number of decoded macroblocks at the time when one frame of encoded data is decoded by the decoding processing means and the image size indicated by the image size information is the same. If the correspondence does not match, a correspondence determining means for storing the substitute image in the display memory is provided instead of the decoded image generated by the decoded image generating means.

  According to the present invention, it is determined whether or not the correspondence between the number of decoded macroblocks at the time when one frame of the encoded data is decoded by the decoding processing means and the image size indicated by the image size information match, If there is no match, the correspondence relationship determining means for storing the substitute image in the display memory is provided instead of the decoded image generated by the decoded image generating means. Even in a situation where encoded data of different moving images is received, even when image size information is not properly notified, there is an effect that display of a corrupted image can be prevented.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a moving picture decoding apparatus according to Embodiment 1 of the present invention. In the figure, a variable-length decoding unit 1 uses MPEG (satellite, terrestrial wave, cable), the Internet, etc. When a video stream (moving image encoded data) compressed by a video encoding method such as the above is received, a decoding process is performed on a frame of the video stream in units of macroblocks, thereby predicting residual information (for example, MPEG4). Then, a process of extracting information such as a quantized DCT (Discrete Cosine Transformation) coefficient and a motion vector is performed.
Further, the variable length decoding unit 1 performs a process of outputting the number of decoded macroblocks when the decoding of one frame of the video stream is completed.
The variable length decoding unit 1 constitutes decoding processing means.

The macroblock position specifying unit 2 receives, for example, image size information indicating the image size of the moving image from the encoding device on the transmission side, and outputs macroblock position information indicating the position of the macroblock to be processed according to the image size information. Processing to output to the buffer 6, the frame memory 7 and the display memory 8 is performed.
In the example of FIG. 1, it is assumed that the image size information is transmitted through a different route from the video stream using, for example, RTSP (Realtime Streaming Protocol), but the image size information is superimposed on the video stream. If so, the variable length decoding unit 1 decodes the image size information superimposed on the video stream, and outputs the image size information to the macroblock position designation unit 2 and the image size determination unit 9. Also good.

The prediction residual calculation unit 3 uses the prediction residual information output from the variable length decoding unit 1 and the information of the past prediction residual macroblock in the processing frame stored in the line buffer 6, so that the currently processed macro is processed. A process of calculating a prediction residual macroblock of the block is performed.
For example, in MPEG-4, as information of past prediction residual macroblocks, DC (Direct Current) of DCT coefficients of macroblocks positioned on the upper side and the left side of the currently processed macroblock in the raster scan order. The component is stored in the line buffer 6, and the prediction residual calculation unit 3 refers to the information of the past prediction residual macroblock, and among the DCT coefficients of the currently processed macroblock, the DC component and AC A part of the (Alternating Current) coefficient is decoded by ADPCM (Adaptive Differential Pulse Code Modulation), and the decoding result is output to the adder 5 as a prediction residual macroblock, and the decoding result is also stored in the past prediction residual macro. Line buff as block information And stores it in the 6.

The motion compensated prediction unit 4 outputs a region indicated by the motion vector output from the variable length decoding unit 1 in the decoded image of the previous frame stored in the frame memory 7 to the adder 5 as a predicted image macroblock. To implement.
The adder 5 adds the prediction residual macroblock output from the prediction residual calculation unit 3 and the prediction image macroblock output from the motion compensated prediction unit 4 to generate a decoded image of the macroblock. A process of writing the decoded image in the display memory 8 and the frame memory 7 according to the macroblock position information output from the position specifying unit 2 is performed.

The line buffer 6 is a memory for storing information of past prediction residual macroblocks in the frame being processed.
The frame memory 7 is a memory for storing the decoded image of the previous frame.
The display memory 8 is a memory for storing a decoded image displayed by a display device (not shown).
The macroblock position designation unit 2, the prediction residual calculation unit 3, the motion compensation prediction unit 4, the adder 5, the line buffer 6, and the frame memory 7 constitute decoded image generation means.

  The image size determination unit 9 determines whether or not the correspondence between the number of decoded macroblocks at the time when one frame of the video stream is decoded by the variable length decoding unit 1 and the image size indicated by the image size information match. If the relationship does not match, a process of storing the substitute image in the display memory 8 instead of the decoded image output from the adder 5 is performed. The image size determining unit 9 constitutes a correspondence determining unit.

Next, the operation will be described.
When the moving picture decoding apparatus employs an image encoding method such as MPEG-4, for example, the video frame of the video stream is divided into rectangular areas called macroblocks and processed in raster scan order in units of macroblocks. .
For example, in the macro block of “Simple Profile” in MPEG-4, the luminance component is composed of 16 × 16 pixels, and the two color difference components are each composed of 8 × 8 pixels.

  When the variable length decoding unit 1 receives a video stream compressed by a video encoding method such as MPEG using digital broadcasting (satellite, terrestrial wave, cable), the Internet, or the like, the frame of the video stream is macroblocked. By dividing each block and performing decoding processing in units of macroblocks, information such as prediction residual information and motion vectors of each macroblock is extracted.

In addition, the variable length decoding unit 1 counts the number of decoded macroblocks when decoding processing is performed in units of macroblocks, and determines the number of decoded macroblocks at the time when decoding of one frame of the video stream is completed. To the unit 9.
For example, if the image size of the moving image is QCIF size, the number of decoded macroblocks for one frame is 99. If the image size of the moving image is CIF size, the number of decoded macroblocks for one frame is 396. It is a piece.

The macroblock position specifying unit 2 receives, for example, image size information indicating the image size of the moving image from the encoding device on the transmission side, and macroblock position information indicating the position of the macroblock to be processed according to the image size information Are output to the line buffer 6, the frame memory 7 and the display memory 8.
Here, FIG. 2 is an explanatory diagram showing the frame memory 7 and the display memory 8 and the line buffer 6 when the video encoding method is MPEG-4.
As apparent from FIG. 2, when the image sizes of the moving images are different, the positions of the macro blocks written in the frame memory 7, the display memory 8 and the line buffer 6 are changed.

Therefore, the macroblock position specifying unit 2 is transmitted from the encoding device on the transmission side in order to prevent the display of a corrupted image even if a video stream of a moving image having a different image size is received in the middle of the moving image. In accordance with the image size information, the position of the macro block written in the frame memory 7, the display memory 8 and the line buffer 6 is optimized.
However, when the image size information is not properly notified from the encoding device on the transmission side, it is impossible to optimize the positions of the macroblocks written in the frame memory 7, the display memory 8, and the line buffer 6.

  In the example of FIG. 1, it is assumed that the image size information is transmitted via a different path from the video stream using, for example, RTSP, but the image size information is superimposed on the video stream. Alternatively, the variable length decoding unit 1 may decode the image size information superimposed on the video stream and output the image size information to the macroblock position designation unit 2 and the image size determination unit 9.

When receiving the number of decoded macroblocks at the time when one frame of the video stream is decoded from the variable length decoding unit 1, the image size determining unit 9 receives the number of decoded macroblocks and the image size transmitted from the encoding device on the transmission side. It is determined whether or not the correspondence relationship with the image size of the moving image indicated by the information matches.
For example, when the image size of the moving image indicated by the image size information is the QCIF size, the number of decoded macroblocks for one frame is 99, or when the image size of the moving image is the CIF size, it corresponds to one frame. If the number of decoded macroblocks is 396, it is determined that the correspondences match.
On the other hand, when the image size of the moving image is QCIF size, when the number of decoded macroblocks for one frame is not 99, or when the image size of the moving image is CIF size, the number of decoded macroblocks for one frame Is not 396, it is determined that the correspondence does not match.

When the correspondences match, the image size determination unit 9 is appropriately notified of the image size information from the encoding device on the transmission side, and the prediction residual information previously extracted by the variable length decoding unit 1 and Even if a decoded image is generated from the motion vector and the decoded image is displayed, a corrupted image is not displayed. Therefore, a control signal instructing output of the decoding result is output to the variable length decoding unit 1.
When the variable length decoding unit 1 receives the control signal instructing output of the decoding result from the image size determination unit 9, the variable length decoding unit 1 outputs the prediction residual information of the macroblock extracted previously to the prediction residual calculation unit 3, and the macro The motion vector of the block is output to the motion compensation prediction unit 4.

When the prediction residual calculation unit 3 receives the prediction residual information from the variable length decoding unit 1, the prediction residual information and information of the past prediction residual macroblock in the frame being processed stored in the line buffer 6. Then, the prediction residual macroblock of the macroblock currently being processed is calculated.
For example, in MPEG-4, as the information of past prediction residual macroblocks, the DC components of the DCT coefficients of the macroblocks located on the upper side and the left side of the currently processed macroblock are displayed in the line buffer as raster scan order. 6, the prediction residual calculation unit 3 refers to the information of the past prediction residual macroblock and refers to one of the DC component and the AC coefficient among the DCT coefficients of the currently processed macroblock. The result is decoded by ADPCM, the decoding result is output to the adder 5 as a prediction residual macroblock, and the decoding result is stored in the line buffer 6 as past prediction residual macroblock information.

When the motion compensated prediction unit 4 receives the motion vector from the variable length decoding unit 1, the motion compensation prediction unit 4 adds, as a predicted image macroblock, a region indicated by the motion vector in the decoded image of the previous frame stored in the frame memory 7. 5 is output.
When the adder 5 receives the prediction residual macroblock from the prediction residual calculation unit 3 and receives the prediction image macroblock from the motion compensation prediction unit 4, the adder 5 adds the prediction residual macroblock and the prediction image macroblock, A decoded image of the macroblock is generated.
When the adder 5 generates a decoded image of the macroblock, the adder 5 stores the decoded image at a position on the display memory 8 and the frame memory 7 indicated by the macroblock position information output from the macroblock position specifying unit 2.

When the correspondence between the number of decoded macroblocks and the image size does not match, the image size determination unit 9 has not been notified of the image size information properly from the encoding device on the transmission side, and the variable length decoding unit 1 first. When the decoded image is generated from the prediction residual information and the motion vector extracted by the above and the decoded image is displayed, the corrupted image is displayed. Therefore, the control signal for stopping the output of the decoding result is variable. Output to the long decoding unit 1.
Thereby, the prediction residual information and the motion vector are not output from the image size determination unit 9, and the decoded image of the macroblock is not output from the adder 5.

When the image size determination unit 9 outputs a control signal for stopping output of the decoding result to the variable length decoding unit 1, the image size determination unit 9 performs a process of writing a substitute image prepared in advance in the display memory 8.
As an alternative image, a test pattern image prepared in advance, a decoded image of the previous frame, or the like is used.
As the test pattern image, for example, a blue or black image can be considered. An image subjected to a predetermined processing such as a fade-out process can also be considered.

  As is clear from the above, according to the first embodiment, the correspondence between the number of decoded macroblocks at the time when one frame of the video stream is decoded by the variable length decoding unit 1 and the image size indicated by the image size information is one. If the correspondence does not match, an image size determination unit 9 for storing the substitute image in the display memory 8 is provided instead of the decoded image generated by the adder 5. Therefore, in the situation where a video stream of a moving image having a different image size is received in the middle of the moving image, even when the image size information is not properly notified, the display of the corrupted image can be prevented. Play.

Embodiment 2. FIG.
3 is a block diagram showing a moving picture decoding apparatus according to Embodiment 2 of the present invention. In the figure, the same reference numerals as those in FIG.
The image size determination unit 10 determines whether or not the correspondence between the number of decoded macroblocks at the time when one frame of the video stream is decoded by the variable length decoding unit 1 and the image size indicated by the image size information match. If the relationship does not match, the variable-length decoding unit 1 is instructed to re-execute the decoding process from the beginning of the video stream, and the current image size is estimated from the number of decoded macroblocks. Is notified to the macroblock position designation unit 2. Note that the image size determination unit 10 constitutes a correspondence determination unit.

Next, the operation will be described.
Compared to the first embodiment, the operation is the same except that an image size determination unit 10 is provided instead of the image size determination unit 9, and thus the operation of the image size determination unit 10 will be mainly described.

When receiving the number of decoded macroblocks at the time when one frame of the video stream is decoded from the variable length decoding unit 1, the image size determining unit 10 determines the number of decoded macroblocks as in the image size determining unit 9 of FIG. It is determined whether or not the correspondence relationship with the image size of the moving image indicated by the image size information transmitted from the transmission-side encoding device matches.
For example, when the image size of the moving image indicated by the image size information is the QCIF size, the number of decoded macroblocks for one frame is 99, or when the image size of the moving image is the CIF size, it corresponds to one frame. If the number of decoded macroblocks is 396, it is determined that the correspondences match.
On the other hand, when the image size of the moving image is QCIF size, when the number of decoded macroblocks for one frame is not 99, or when the image size of the moving image is CIF size, the number of decoded macroblocks for one frame Is not 396, it is determined that the correspondence does not match.

  When the correspondences match, the image size determination unit 10 is appropriately notified of the image size information from the encoding device on the transmission side, and the prediction residual information previously extracted by the variable length decoding unit 1 and Even if a decoded image is generated from a motion vector and the decoded image is displayed, a corrupted image is not displayed. Therefore, as with the image size determination unit 9 in FIG. The signal is output to the variable length decoding unit 1.

  When the correspondence between the number of decoded macroblocks and the image size does not match, the image size determination unit 10 has not been notified of the image size information from the encoding device on the transmission side, and the variable length decoding unit 1 first. When the decoded image is generated from the prediction residual information and the motion vector extracted by the above and the decoded image is displayed, the corrupted image is displayed. Therefore, the decoding process from the beginning of the video stream is re-executed. A control signal instructing execution is output to the variable length decoding unit 1. At this time, if the prediction residual information and the motion vector extracted previously are recorded for one frame, the variable length decoding unit 1 can reuse the numerical value without performing the decoding process.

In addition, the image size determination unit 10 estimates the current image size from the number of decoded macroblocks at the time when one frame of the video stream is decoded, and notifies the macroblock position designation unit 2 of the current image size.
For example, if the number of decoded macroblocks is 99, the current image size is estimated to be QCIF size, and if the number of decoded macroblocks is 396, the current image size is estimated to be CIF size.

  As a result, a decoded image is generated from the prediction residual information and the motion vector newly extracted by the variable length decoding unit 1, and the position of the macroblock corresponding to the estimated current image size (line buffer 6, frame memory 7). Since the decoded image is written at an appropriate position on the display memory 8, an image without collapse is displayed.

  As is apparent from the above, according to the second embodiment, the correspondence between the number of decoded macroblocks at the time when one frame of the video stream is decoded by the variable length decoding unit 1 and the image size indicated by the image size information is one. If the correspondence does not match, the variable length decoding unit 1 is instructed to re-execute the decoding process from the beginning of the video stream, and the current image is determined from the number of decoded macroblocks. Since the image size determination unit 10 that estimates the size and notifies the macroblock position specifying unit 2 of the current image size is provided, a video stream of a moving image having a different image size is received in the middle of the moving image. Even if the image size information is not properly notified under such circumstances, the display of a corrupted image can be prevented.

It is a block diagram which shows the moving image decoding apparatus by Embodiment 1 of this invention. (A) is explanatory drawing which shows the frame memory 7 and the display memory 8 in case a video coding system is MPEG-4. (B) is an explanatory view showing the line buffer 6 when the video encoding method is MPEG-4. It is a block diagram which shows the moving image decoding apparatus by Embodiment 2 of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Variable length decoding part (decoding processing means), 2 Macroblock position designation part (decoded image generation means), 3 Prediction residual calculation part (decoded image generation means), 4 Motion compensation prediction part (decoded image generation means), 5 Adder (decoded image generating means), 6 line buffer (decoded image generating means), 7 frame memory (decoded image generating means), 8 display memory, 9, 10 image size determining unit (corresponding relationship determining means).

Claims (3)

  A decoding processing unit that inputs encoded data of a moving image and performs a decoding process of the encoded data, and a decoded image is generated from the decoding processing result of the decoding processing unit, and an image size indicating an image size of the moving image Correspondence between decoded image generation means for storing the decoded image in a display memory according to information, the number of decoded macroblocks when one frame of encoded data is decoded by the decoding processing means, and the image size indicated by the image size information It is determined whether or not the relationship is matched, and if the correspondence is not matched, a corresponding relationship determining unit that stores a substitute image in the display memory instead of the decoded image generated by the decoded image generating unit; A video decoding device comprising:   2. The moving picture decoding apparatus according to claim 1, wherein the correspondence determination unit stores a test pattern image prepared in advance or a decoded image of the previous frame as a substitute image in a display memory.   A decoding processing unit that inputs encoded data of a moving image and performs a decoding process of the encoded data, and a decoded image is generated from the decoding processing result of the decoding processing unit, and an image size indicating an image size of the moving image Correspondence between decoded image generation means for storing the decoded image in a display memory according to information, the number of decoded macroblocks when one frame of encoded data is decoded by the decoding processing means, and the image size indicated by the image size information It is determined whether or not the relationship is the same, and if the correspondence is not the same, the decoding processing means is instructed to perform the decoding process from the beginning of the encoded data, and the number of decoded macroblocks is determined. A moving picture decoding apparatus comprising: a correspondence determination unit that estimates a current image size and notifies the decoded image generation unit of the current image size.

Images