JP4228537B2 - Image signal decoding / encoding device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention superimposes each encoding parameter corresponding to each macroblock at the time of encoding the encoded image signal on the decoded image signal obtained by decoding the encoded image signal that has been subjected to high-efficiency predictive encoding.The present invention relates to an image signal decoding / encoding device that re-encodes the decoded image signal.
[0002]
  And the present invention is particularlyImage signal decoding / coding that enables good re-encoding when re-encoding a decoded image signal edited by switching a plurality of decoded image signals on which the respective encoding parameters before decoding are superimposed An object is to provide a device.
[0003]
[Prior art]
In recent years, there has been a service for compressing digitized video and audio signals by high-efficiency predictive coding and using the compressed information to distribute information over transmission lines such as satellite waves, terrestrial waves, and telephone lines. It has been put into practical use. In such a service, MPEG2, which is an international standard, is used as a high-efficiency predictive coding method for moving image signals and audio signals. MPEG2 is an encoding method that compresses the amount of information of an image signal using correlation between adjacent pixels of the image signal (correlation in the spatial direction) and correlation between frames or fields (correlation in the time direction).
[0004]
Image encoding in the MPEG2 standard is processed by the following algorithm. First, temporally continuous image frames are divided into a reference frame and a prediction frame. The reference frame can be restored using only the encoded data of the frame by using only the spatial correlation. By using both the temporal direction correlation and the spatial direction correlation from the reference frame, the prediction frame can further increase the coding efficiency with respect to the reference frame. The predicted frame is restored from the decoded reference frame and encoded data.
[0005]
A specific encoding system used in MPEG2 image encoding will be described with reference to FIG. The I frame, which is a reference frame, periodically exists and serves as a reference for decoding processing. Further, the prediction frame includes a P frame that is encoded only by prediction from a temporally previous reference frame and a B frame that is predictively encoded from two temporally preceding and following reference frames. The P frame itself is a prediction frame and also serves as a reference frame for the following B frame and P frame.
[0006]
The image signal of the I frame is divided into processing units called macroblocks of horizontal 16 pixels × vertical 16 pixels as luminance signals. The divided macroblock data is further divided into two-dimensional blocks of 8 × 8 pixel units, and a DCT (discrete cosine transform) process, which is a kind of orthogonal transform, is performed. Since the signal after DCT conversion shows a value according to the frequency component of the two-dimensional block, the component is concentrated in a low frequency in a general image. In addition, information degradation of high frequency components has a property that is visually less noticeable than information degradation of low frequency components. Therefore, the amount of information is compressed by finely quantizing the low-frequency components finely and roughly quantizing the high-frequency components, and variable-length coding the continuous length of the quantized coefficient component and the coefficient component without coefficient (coefficient 0). ing.
[0007]
The image signal of the P frame is divided into 16 × 16 pixel macroblocks as in the case of the I frame. In the P frame, a motion vector between the reference frame and each macroblock is calculated. Motion vector detection is generally obtained by block matching. The absolute sum of absolute differences (or the sum of squared differences) of each pixel of the macroblock and the reference frame where the horizontal / vertical position where the macroblock exists is moved by 16 x 16 pixels. And the value of the motion vector taking the minimum value is output as the detected motion vector.
[0008]
Each pixel of the macroblock is compared with each pixel of the two-dimensional block cut out by the motion vector. When an accurate motion vector is detected, the information amount of the difference block is significantly smaller than the information amount of the original macroblock, so that a rougher quantization process than that of the I frame is possible. Actually, select whether to encode the difference block or non-difference block (Intra block) (prediction mode determination), and DCT / variable length encoding process similar to I frame for the selected block The amount of information is compressed.
[0009]
The B frame is processed in the same manner as the P frame. However, since there are reference frames before and after the frame, a motion vector is detected between each frame. There are three types of prediction options for the B frame: prediction from the previous reference frame (Forward prediction), prediction from the subsequent reference frame (Backward prediction), and the average value per pixel of the two prediction blocks (Average prediction). Predictive mode determination is performed from four types of Intra blocks. Since the B frame can be predicted from temporally preceding and following reference frames, the prediction efficiency is further improved than the P frame. Therefore, the quantization is generally rougher than the P frame. The selected block is encoded in the same manner as the I and P frames.
[0010]
Since the B frame is decoded, a prediction process from a temporally subsequent reference frame is performed, so that the reference frame is encoded prior to the B frame. Therefore, the input image signals are rearranged in the order shown in FIG. 4 and encoded. In the decoding process, reverse sorting is performed and output, whereby decoded images are reproduced in the order of input image signals.
[0011]
A conventional general encoding apparatus and decoding apparatus for realizing MPEG2 image encoding will be described with reference to FIGS. 6 and 7, respectively.
[0012]
In the encoding device shown in FIG. 6, the input digital image signal is recorded in the frame memory (input image memory) 1, and a delay is performed for rearrangement in the order of encoding according to the encoding syntax. . From the digital signal output from the input image memory 1, a macro block is cut out by the two-dimensional block conversion circuit 2.
[0013]
In the reference frame, the macroblock data is subjected to DCT transform in units of 8 × 8 pixels by the orthogonal transform circuit 4 and sent to the quantization circuit 4. The DCT coefficient quantized in the quantization circuit 4 is encoded in a variable length or a fixed length by referring to an address corresponding to the coefficient in the encoding table 16 in the encoding circuit 5. The encoded data and the additional information (encoding parameter) indicating the location of the macro block in the screen are multiplexed and output as a bit stream.
[0014]
On the other hand, the DCT coefficients quantized by the quantization circuit 4 are subjected to inverse quantization and inverse DCT processing in the inverse quantization circuit 7 and the inverse orthogonal transform circuit 8, and the encoded bit stream is decoded to generate the deblock circuit 9 and It is stored in the reference image memory 11 via the adder circuit 10.
[0015]
Subsequently, in the prediction frame, a motion vector between images is obtained by the motion vector detection circuit 12 between the macroblock data cut out from the input image memory 1 and the image stored in the reference image memory 11. Based on the output motion vector, the prediction block is cut out from the reference image memory 11 by the motion compensation prediction circuit 13, and the motion compensation prediction circuit 13 selects a prediction mode and subtracts a difference signal from the input image block to be encoded. 14 and sent to the orthogonal transformation circuit 3. The difference signal is processed in the same manner as each block of the frame, and the DCT coefficient is quantized and output as a bit stream together with a motion vector and a prediction mode (encoding parameter).
[0016]
Regarding the control of the code amount, the code amount of the output stream is compared with the target code amount in the code amount control circuit 15, and the quantization level of the quantization circuit 4 in order to approach the target code amount. (Quantization scale) is controlled. For the three types of frame types having different information amounts, the target code amount for each frame is calculated using the nature and appearance frequency of each frame type for the set encoding rate. Further, the target code amount is limited so that the overflow / underflow of the buffer does not occur by virtually simulating the stream buffer of the decoding device. The quantization scale uses the fact that the scale and the output code amount are generally inversely proportional, and calculates the quantization scale value for the target code amount for each frame type and performs the quantization process. By varying the quantization scale in a direction approaching the target code amount for each block, the encoded bit stream is suppressed within the target code amount.
[0017]
In the decoding device shown in FIG. 7, the input bit stream is first stored in the stream buffer 31. A virtually simulated buffer value is written in the encoded bitstream, and when the decoding process is performed after the bitstream for that buffer value has been stored, the buffer breaks down and the decoding process stops. It is preventing. The variable length decoding circuit 32 separates the encoding parameters such as the quantization scale, the prediction mode, and the motion vector, and decodes the quantized DCT coefficient. The decoded DCT coefficients are subjected to inverse quantization and inverse DCT processing in the same inverse quantization circuit 33 and inverse orthogonal transform circuit 34 as in the encoding circuit, and the Intra block or the difference block is decoded. To the adder 36.
[0018]
In the case of a prediction block, the adder 36 adds the prediction block extracted from the reference image memory 38 based on the prediction mode decoded by the motion compensation prediction circuit 37 and the motion vector value, and the image signal of the macroblock is obtained. Is restored. In the case of an I or P frame, the restored macroblock data is written into the reference image memory 38. In the case of the B frame, it is stored in the output frame memory 39 and output as an image signal. Image data of I and P frames written in the reference image memory 38 is added to the output frame memory 39 at the output timing in accordance with the timing of FIG.
[0019]
In such an information distribution system using image encoding, there is a need to decode a once encoded stream and perform encoding again. One is a case where information is transmitted from a place where information is collected and recorded to a place where information is distributed. The transmission path may be a wired / wireless communication line or a recording medium. However, if the bandwidth owned by the transmission path and the transmission path of the system to be distributed are different, it is necessary to change the bit rate of the bit stream. There are other cases where the recorded image signal is edited and processed. When connecting a plurality of image signals or inserting a character telop or the like on the image signal, it is difficult to realize without decoding the stream.
[0020]
In such an apparatus (transcoder) that performs re-encoding processing, in order to reduce encoding degradation during re-encoding, on an image signal (decoded baseband signal) transmitted by the transcoder, A scheme is considered in which frame coding information such as a frame type at the time of coding and macroblock coding parameters are superimposed and transmitted.
[0021]
In MPEG2 encoding, the quality of the image signal generally deteriorates in the order of I frame, P frame, and B frame. Since the I frame is a reference frame, it is quantized more finely than other frames and is not affected by reference frame deterioration because there is no reference from other images. As it becomes P frame and B frame, it is quantized roughly and easily affected by reference frame deterioration. When a frame that was a B frame at the time of re-encoding is encoded as an I-frame, an image with a large deterioration is used as a reference frame, so the predicted frame is affected by the deterioration and Degradation of image quality increases. By referring to the encoding parameters and matching the frame type at the time of re-encoding, the deterioration factor is reduced. Also, by using the macroblock coding parameters in the coding parameters, the amount of processing for motion vector detection can be reduced, and the amount of information the macroblock has from the quantization scale and the number of coding bits. And good code amount control can be expected during bit rate conversion.
[0022]
The encoding parameter for re-encoding needs to be transmitted from the decoding processing means to the re-encoding processing means in association with the video signal temporally and spatially.
[0023]
As a transmission method for realizing this, there is SMPTE 319M which defines a standard for superimposing MPEG-2 video encoding parameters on a standard TV quality 4: 2: 2 digital component video signal for transmission. In addition, the SMPTE 351M proposal is currently under deliberation to define a standard for superimposing MPEG-2 video encoding parameters on a 4: 2: 2 digital component video signal for transmission.
[0024]
A specific method for superimposing encoding parameters will be described using SMPTE 319M as an example. In the baseband signal, a luminance signal (Y) and two color difference signals (Cb, Cr) are transmitted at a sampling ratio of 4: 2: 2, and each pixel value is composed of 10 bits (1024 levels). In MPEG2 image coding, the amount of information of a decoded image is composed of 8 bits (256 levels), so the lower 2 bits of the decoded image signal become a transmission unnecessary area. Coding parameters are superimposed on this area. In the example of SMPTE 319M Table 3, the least significant bits of Cb and Cr in the transmission unnecessary area are used as the overlapping area.
[0025]
A coding parameter standardized by SMPTE 319M (see SMPTE 319M Figure 1) will be described as an example of a coding parameter to be specifically superimposed. For each 16 × 16 pixel area constituting each macroblock, a macroblock coding parameter that is coding information for each macroblock and a part of frame coding information (coding information for each image) Write picture coding parameters. (The macroblock encoding parameter and the picture encoding parameter are collectively referred to as an encoding parameter here.)
When the LSB of Cb and Cr are combined, there are 256 bits, and 256 bits indicate the information of SMPTE 319M Figure 1. As macroblock coding parameters, additional information such as motion vector values, prediction modes, quantization scales, 8 × 8 block patterns (coded # block # pattern) having effective DCT coefficients, and additional information and DCT coefficients The number of bits required for encoding is recorded as information.
[0026]
Also, in the picture # element # information area, frame coding information as shown in SMPTE 319M Table 4 is divided for each macroblock (this divided is a picture coding parameter) and is written. This area has 32 bits for each macroblock, and by collecting picture # element # information of all macroblocks over the entire frame, frame encoded information copied a plurality of times can be restored.
[0027]
An example of such a transcoder that transmits coding parameters from the decoding device to the coding device is shown in FIG. In this example, a synthesizing device is inserted between the decoding device and the encoding device, and an image edited with respect to the decoded baseband signal is re-encoded by the encoding device. The configuration diagrams of the decoding device and the encoding device shown in FIG. 8 are shown and described in FIGS. 9 and 10, respectively. 6 and 7 are denoted by the same reference numerals, and the description thereof is omitted.
[0028]
The variable length decoding circuit 32 of the decoding apparatus shown in FIG. 9 calculates the frame encoding information, the macroblock encoding parameter, and the number of encoding bits of the macroblock at the time of decoding, and inputs them to the encoding information generation circuit 41. . In the encoding information generation circuit 41, the picture encoding parameter and the macroblock encoding parameter, which are encoding parameters, are formatted into a bit pattern as shown in SMPTE 319M Figure 1 and stored in the encoding information memory 42.
[0029]
In the encoded information superimposing circuit 43, the image information stored in the output frame buffer 39 and the encoding parameter stored in the encoded information memory 42 are input and encoded into the LSB bits of the color difference signal of the image information. Superimpose parameters. The encoding information memory 42 is provided for a plurality of frames so as to correspond to the output rearrangement of I, P, and B frames, and the encoding information superimposing circuit 43 changes the order of encoding parameters taken out according to the frame type. .
[0030]
In the encoding apparatus shown in FIG. 10, the encoding information separation circuit 21 extracts the encoding parameter from the image signal of the input image memory. The extracted encoding parameters are stored in the encoding information memory 22. The encoding syntax control circuit 24 detects the frame type from the extracted encoding parameters, and performs control to rearrange the input images in the encoding order. When the encoding parameter of the frame cannot be extracted, a non-detection signal is sent from the encoding information separation circuit 21, and the encoding syntax is configured in the encoding apparatus as in the conventional normal encoding process. Then, the encoding process is performed (the normal encoding process similar to the process for the image signal on which the encoding parameter is not superimposed is performed). In the macroblock information generation circuit 23, the encoding parameters stored in the encoding information memory 22 are read in a different order according to the encoding syntax, and the encoding parameters for each macroblock are extracted.
[0031]
The extracted encoding parameters are transmitted to the motion compensation prediction circuit 13 and the code amount control circuit 15. The motion compensated prediction circuit 13 cuts out a prediction block from the reference image memory 11 using a motion vector and a prediction mode existing in the encoding parameter, generates a difference signal from the input image block to be encoded, and sends it to the orthogonal transformation circuit 3. Send it out. When the encoding parameter in the frame or the encoding parameter for each macroblock cannot be extracted, a non-detection signal is sent from the encoding information separation circuit 21 to the motion compensation prediction circuit 13, and the conventional encoding is performed. Similar to the processing, motion vector detection and prediction mode selection processing are performed.
[0032]
In the code amount control circuit 15, the quantization scale and the number of bits required for encoding for each macroblock are input as encoding parameters, and the set code amount and the comparison processing at the encoding bit rate that is the current control target are compared. As a result, an appropriate quantization scale is determined and an encoding process is performed.
[0033]
[Problems to be solved by the invention]
When an encoded image signal is once decoded and edited, baseband decoded signals output from a plurality of decoding apparatuses may be switched and edited. As a transcoding device that performs this editing process, for example, a device in which an image switch is inserted between a plurality of decoding devices and one coding device as shown in FIG. 8 can be considered. The decoded baseband signal (image signal) is subjected to switching processing by an image switch according to the switching timing, and the baseband signal subjected to the switching processing is input to the encoding device and re-encoded. In this case, there is a problem that it is impossible to recognize that the continuity of the encoded information in units of frames at the switching timing is cut off, and erroneous encoded information is used during re-encoding. Also, when the switching timing occurs within a frame, the macroblock encoding parameter becomes an output of a different decoding apparatus, that is, an encoding parameter of a different stream, depending on the portion in one screen (in one image unit), and re-encoding If it is used incorrectly at the time of conversion, there will be a contradiction in macroblock encoding parameters such as motion vectors and prediction modes, resulting in a problem of image quality degradation. Further, even when it can be recognized that the encoding parameter is destroyed, re-encoding is performed free of charge, and a good transcoding function cannot be realized.
[0034]
  The present invention provides a plurality of image signal decodingProcessing partWhen re-encoding the decoded image signal edited by switching the decoded image signal on which the pre-decoding encoding parameters output from are superimposed, it is possible to perform good re-encoding.PaintingAn object of the present invention is to provide an image signal decoding / encoding device.
[0035]
[Means for Solving the Problems]
  In order to solve the above problems, the present invention provides the following apparatus.
(1)  A plurality of decoding parameters obtained by decoding a coded image signal that has been subjected to high-efficiency predictive coding, and each coding parameter corresponding to each macroblock at the time of coding the coded image signal is superimposed and output. Each of the coding parameters includes a macroblock coding parameter that is coding information for each macroblock and a picture coding parameter that is a part of the coding information for each image. A plurality of image signal decoding processing units to include,
  An image signal switching unit that switches and outputs the decoded image signals respectively supplied from the plurality of image signal decoding processing units;
  Image signal decoding comprising: an image signal encoding processing unit that performs high-efficiency predictive encoding on the decoded image signal output from the image signal switching unit based on the respective encoding parameters separated from the decoded image signal -In the encoding device,
  In each image signal decoding processing unit,
  Adding an identification signal of the image signal decoding processing unit to each encoding parameter, and providing superimposing means for superimposing the encoding parameter with the identification signal added on the decoded image signal;
  In the image signal encoding processing unit,
  Image signal decoding obtained by decoding the decoded image signal from the respective encoding parameters separated from the decoded image signal corresponding to one image unitProcessing partDetecting means for detecting the identification signal of
  For each identification signal that is the same as the detected identification signal, was the encoding information obtained in units of one image obtained by combining the picture encoding parameters extracted from the respective encoding parameters in which the identification signal was detected? When the encoded information for each image is obtained, it is determined whether or not all the detected identification signals are the same, and when the detected identification signals are not the same, It is determined whether there is one type of identification signal detected from each encoding parameter from which encoding information of one image unit is obtained, and based on the determination result, the acquired one image unit First determination means for determining whether the encoded information is valid or invalid at the time of encoding;
  Second determining means for determining whether the macroblock coding parameter in each separated coding parameter is valid or invalid at the time of coding based on the detected identification signal;
  The image signal encoding processing unit uses the macroblock encoding parameter and the picture encoding parameter for encoding for the macroblock encoding parameter and the picture encoding parameter determined to be valid, and the macroblock determined to be invalid As for the encoding parameter and the picture encoding parameter, information corresponding to the macroblock encoding parameter and the picture encoding parameter is generated and used for encoding.
An image signal decoding / encoding device characterized by that.
[0036]
DETAILED DESCRIPTION OF THE INVENTION
An example of a transcoding apparatus (schematic configuration is the same as in FIG. 8) that enables switching editing using an embodiment of the present invention will be described. FIG. 1 shows a block diagram of an image signal decoding apparatus in one embodiment. The same parts as those shown in FIG. 9 which is the conventional example are denoted by the same reference numerals, and detailed description thereof will be omitted.
[0037]
The variable length decoding circuit 32 calculates the picture coding parameter, the macroblock coding parameter, and the number of coded bits of the macroblock when decoding in the same manner as in the conventional example, and inputs it to the coding information generation circuit 52. In the encoded information generation circuit 52, in addition to the picture encoding parameter and the macroblock encoding parameter, information of the decoding apparatus identification signal memory 51 in which a value (Decoder ID) indicating the identification of the decoding apparatus itself is recorded is input. Then, it is formatted into a bit pattern and stored in the encoded information memory 42.
[0038]
Coding parameters (macroblock coding parameters that are coding information in units of macroblocks) and picture coding parameters that are part of frame coding information (coding information in units of images) are encoded here. As an example of a specific format (referred to as a parameter), the one shown in SMPTE 319M Figure 1 described above is used, but about 4 bits in the reserved area in the macroblock coding parameter area of that format. (For example, the 0th bit in the vertical direction and the 6th to 9th bits in the horizontal direction) are used as an area for adding a Decoder ID.
[0039]
Decoder ID can be set from the outside. By setting different IDs among a plurality of decoding devices at the time of transcoding, it can be made an identifiable signal.
[0040]
In the encoded information superimposing circuit 43, the image information stored in the output frame memory 39 and the encoding parameter which is the encoded information stored in the encoded information memory 42 are input as in the conventional case, and the image information The encoding parameter is superimposed on the LSB bit of the color difference signal.
[0041]
Next, FIG. 2 shows a block diagram of an image signal encoding apparatus according to an embodiment, which will be described. In addition, the same code | symbol is attached | subjected to the part similar to what is shown in FIG. 10 which is a prior art example, and the specific description of the part is abbreviate | omitted.
[0042]
The macroblock information generation circuit 70 extracts the Decoder ID in the encoding parameter together with the function shown in the conventional example, and stores it in the Decoder ID memory 71.
[0043]
When the macro block information for one frame (one image unit) is stored, it can be recognized whether or not the frame is constituted by an output from a single decoding device. In the frame information determination circuit 73, a Decoder ID detected from each macroblock for one frame and frame encoding information (ie, a predetermined number of picture encoding parameters) that is extracted frame-by-frame encoding information are combined into one frame. From the encoded information), the validity of the frame encoded information is determined and the frame information is acquired. If encoding parameters can be extracted from all macroblocks and the decoder ID values of the macroblocks in the decoder ID memory 71 are all the same, it is determined that the output is from a single decoding device. Also, when compared with the Decoder ID of the previous frame, if the Decoder ID is different, it is determined as the switching point of the image signal in units of frames.
[0044]
A specific frame information acquisition algorithm will be described with reference to FIG. First, for each identical Decoder ID detected from each macroblock for one frame, frame coding information obtained by combining picture coding parameters extracted from each coding parameter in which the Decoder ID is detected is stored. (Step S1).
[0045]
When frame coding information is not obtained for any Decoder ID (step S2), the frame type managed in advance in the coding apparatus is used (step S3). If there is a Decoder ID from which frame coding information can be extracted (step S2), the coding parameters can be extracted from all the macroblocks, and the Decoder ID for each macroblock is the same (this) Is set as condition 1) (step S4). If this condition 1 is satisfied, the extracted frame coding information is selected as a use candidate (step S5).
[0046]
When the condition 1 is not satisfied (step S4), it is checked whether or not there are a plurality of decoder IDs from which the frame encoded information can be extracted (step S6), and in the case of one, the extracted frame encoded information is used. A candidate is selected (step S5). In the case of a plurality of frames, the frame coding information corresponding to the same ID as the Decoder ID detected one frame before is preferentially selected as a use candidate (Step S7, Step S5).
[0047]
When the frame encoding information selected as the use candidate corresponds to the same Decoder ID as the Decoder ID corresponding to the frame encoding information selected one frame before (step S8), the selected frame The frame type in the encoded information and other additional information in the frame encoded information are used for re-encoding as they are (step S9).
[0048]
If they are not the same ID (step S8), or if there is no frame encoding information corresponding to the same ID as the Decoder ID detected one frame before in step S7, the frame type of the selected frame encoding information is checked. (Step S11). In the case of No in step S7, the selection from the plurality of frame encoding information is simply performed by detecting the frame encoding information corresponding to the first acquired decoder ID (ie, the same decoder ID as the first acquired decoder ID). Frame encoding information obtained from each encoding parameter) (step S10).
[0049]
When the frame type is I frame in step S11, since the encoding is completed within the frame, additional information is useful even for a frame having an independent Decoder ID. Therefore, the frame type in the frame coding information selected in step S10 and other additional information in the frame coding information are used as they are for re-coding.
[0050]
If the frame type is P frame (step S12), it is determined that the input has been switched (since it is determined that the Decercer ID is different from the previous frame in step S7), and the frame type is changed to I frame and selected. The additional information in the frame encoded information is discarded as being unusable (step S13).
[0051]
In the case of a B frame, in order to limit the number of consecutive B frames, only when the previous frame is an I frame or a P frame (step S14), the frame type is set as the B frame as it is (step S15). In other cases, the frame type is set to I frame, and in any case, the selected additional information in the frame encoded information is discarded (step S13).
[0052]
Next, the encoding syntax control circuit 24 performs control for rearranging the input images in the encoding order based on the selected frame type, and for the frame encoding information determined to be valid by the frame information determination circuit 73. Then, the frame unit is controlled using the value. Similarly to the reference image memory 11, the Decoder ID memory 71 stores and controls the Decoder ID of the image serving as the base frame.
[0053]
The significant / ineffective detection circuit 72 is supplied with a decoder ID from the decoder ID memory 71 and a macroblock coding parameter from the coding information memory 22 and uses the macroblock coding parameter separated from the input image signal. It is determined whether or not it is possible. The significant / ineffective detection circuit 72 reads the Decoder ID detected from the encoding target macroblock and, if the prediction mode in the macroblock encoding parameter in the encoding target macroblock is not an Intra block, the macroblock code The motion vector value is read from the quantization parameter, and the decoder ID address value of the macroblock in which the predicted reference image data exists is calculated.
[0054]
When the motion vector value is horizontal X pixel, vertical Y pixel, and the macroblock head position point is horizontal H and vertical V, there are four types of address values where the required Decoder ID exists. the position is,
(1) Horizontal (H + X) / 16, Vertical (V + Y) / 16
(2) Horizontal (H + X + 15) / 16, Vertical (V + Y) / 16
(3) Horizontal (H + X) / 16, Vertical (V + Y + 15) / 16
(4) Horizontal (H + X + 15) / 16, Vertical (V + Y + 15) / 16
It becomes. When the Decoder ID read from these address values and the Decoder ID of the own macroblock match, it is assumed that the macroblock coding parameter is significant. However, when the prediction mode of the macroblock coding parameter is forward prediction, only the Decoder ID of the forward reference image is referenced, and in the backward prediction, only the Decoder ID of the backward reference image is referenced. Just judge.
[0055]
Further, the decoder ID recognized for each frame is input from the encoding syntax control circuit 24 to the significant / ineffective detection circuit 72 together with the frame type. If the input Decoder ID is different from the Decoder ID of its own macroblock, the macroblock coding parameter is assumed to be useless.
[0056]
By referring to the output signal of the significant / ineffective detection circuit 72, the frame information acquired by the frame information determination circuit 73, and the non-detection signal of the encoded information for each macroblock by the encoded information separation circuit 21, the code for each macroblock is referred to. Process. In other words, if the significant / ineffective detection circuit 72 determines that it is significant, the encoding is performed based on the separated macroblock encoding parameter corresponding to the encoding target macroblock and the frame information acquired by the frame information determination circuit 73. Encode the target macroblock. When the significant / ineffective detection circuit 72 determines that it is ineffective and when the non-detection signal is output from the encoded information separation circuit 21, information corresponding to the macroblock encoding parameter is newly generated, and the information Based on the frame information acquired by the frame information determination circuit 73, the encoding target macroblock is encoded.
[0057]
As described above, according to the apparatus of the present embodiment, once a plurality of encoded image signals are decoded, editing is performed by switching the decoded plurality of image signals, and then the encoding process is performed again. When performing, the encoding parameter at the time of encoding before decoding is held on the image signal output by the decoding device, and an identification signal for identifying each decoding device is added to the encoding parameter for each macroblock and written. I made it. Therefore, by detecting the identification signal at the time of re-encoding, it is possible to determine the switching of the decoded image signal in units of frames (decoding device switching in units of frames) and also recognize the switching of the decoding devices in the frame. Can do.
[0058]
In the state where the identification signal is continuous, the frame information (information obtained from the frame encoded information separated and extracted from the decoded image signal) remains completely maintained, so that re-encoding processing is performed using the frame information. Since it can be performed, the position of the I frame of the stream that has been previously encoded can be aligned with the position of the I frame at the time of re-encoding, and image degradation at the time of re-encoding can be reduced. Furthermore, by controlling the re-encoding process based on the detected combination of identification signals, it is possible to prevent a change that causes a malfunction of the encoding syntax or disturbance of the code amount control at the switching point.
[0059]
In addition, by using the detected identification signal, it is managed whether the macroblock coding parameter of the macroblock to be coded is consistent with the control at the time of re-encoding, and the prediction of the reference image when performing the prediction process It can be easily recognized whether or not there is an inconsistent area in the block, and re-encoding processing can be prevented from being performed using an incorrect encoding parameter. Thereby, it becomes possible to appropriately use coding parameters effective for re-encoding, and a good transcoding function can be realized.
[0060]
【The invention's effect】
  As described above, the present inventionAccording to the image signal decoding / encoding device,Good re-encoding when re-encoding a decoded image signal that has been edited by switching a plurality of decoded image signals (decoded image signals on which encoding parameters before decoding are superimposed)It can be performed.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of an image signal decoding apparatus according to an embodiment.
FIG. 2 is a configuration diagram of an image signal encoding device according to an embodiment.
FIG. 3 is a diagram illustrating an encoding system used in MPEG2 image encoding.
[Fig. 4] Fig. 4 is a diagram illustrating the encoding order rearrangement timing at the time of MPEG2 encoding.
FIG. 5 is a diagram illustrating a relationship between a stream arrival order during decoding and an output order of decoded images.
FIG. 6 is a configuration diagram of an MPEG2 image encoding device.
FIG. 7 is a configuration diagram of an MPEG2 image decoding device.
FIG. 8 is a configuration diagram of a transcoding device that enables switching editing.
FIG. 9 is a configuration diagram of a conventional MPEG2 image decoding apparatus having a transcoding function.
FIG. 10 is a configuration diagram of a conventional MPEG2 image encoding device having a transcoding function.
FIG. 11 is a diagram illustrating a frame information acquisition algorithm according to an embodiment.
[Explanation of symbols]
21 Encoded information separation circuit
22 Encoded information memory
24 Coding syntax control circuit
51 Decoding device identification signal memory
52 Encoded information generation circuit
70 Macroblock information generation circuit
71 Decoder ID memory
72 Majority detection circuit
73 Frame information judgment circuit

Claims (1)

A plurality of decoding parameters obtained by decoding a coded image signal that has been subjected to high-efficiency predictive coding, and each coding parameter corresponding to each macroblock at the time of coding the coded image signal is superimposed and output. Each of the coding parameters includes a macroblock coding parameter that is coding information for each macroblock and a picture coding parameter that is a part of the coding information for each image. A plurality of image signal decoding processing units to include,
An image signal switching unit that switches and outputs the decoded image signals respectively supplied from the plurality of image signal decoding processing units;
Image signal decoding comprising: an image signal encoding processing unit that performs high-efficiency predictive encoding on the decoded image signal output from the image signal switching unit based on the respective encoding parameters separated from the decoded image signal -In the encoding device,
In each image signal decoding processing unit,
Adding an identification signal of the image signal decoding processing unit to each encoding parameter, and providing superimposing means for superimposing the encoding parameter with the identification signal added on the decoded image signal;
In the image signal encoding processing unit,
Detecting means for detecting an identification signal of an image signal decoding processing unit that has performed decoding into the decoded image signal from within each of the encoding parameters separated from the decoded image signal corresponding to one image unit;
For each identification signal that is the same as the detected identification signal, was the encoding information obtained in units of one image obtained by combining the picture encoding parameters extracted from the respective encoding parameters in which the identification signal was detected? When the encoded information for each image is obtained, it is determined whether or not all the detected identification signals are the same, and when the detected identification signals are not the same, It is determined whether there is one type of identification signal detected from each encoding parameter from which encoding information of one image unit is obtained, and based on the determination result, the acquired one image unit First determination means for determining whether the encoded information is valid or invalid at the time of encoding;
Second determining means for determining whether the macroblock coding parameter in each separated coding parameter is valid or invalid at the time of coding based on the detected identification signal;
The image signal encoding processing unit uses the macroblock encoding parameter and the picture encoding parameter for encoding for the macroblock encoding parameter and the picture encoding parameter determined to be valid, and the macroblock determined to be invalid For the encoding parameter and picture encoding parameter, information corresponding to the macroblock encoding parameter and picture encoding parameter is generated and used for encoding.
An image signal decoding / encoding device characterized by that.

Images