KR100312419B1 - Error propagation prevention method due to transmission error in video coder - Google Patents

Abstract

1. TECHNICAL FIELD OF THE INVENTION

The present invention relates to a method for preventing error propagation due to transmission errors in a video encoding apparatus and a computer-readable recording medium having recorded thereon a program for realizing the method.

2. The problem to be solved by the invention

In the block-based video encoding method using motion estimation and compensation such as MPEG-2, the present invention maintains a trace of motion in a block unit for a certain frame period in order to prevent propagation of a damaged image due to a transmission error. In addition, the receiving end notifies the transmitting end of the error block to the transmitting end, and the transmitting end records the error propagation prevention method due to transmission error and a program for realizing the method in the video encoding apparatus capable of preventing the propagation of the error. To provide a computer-readable recording medium.

3. Summary of Solution to Invention

The present invention provides a method for preventing error propagation due to transmission error in a video encoding apparatus, the first encoding of which the encoder encodes through block motion estimation and prediction compensation between frames in order to prevent propagation of a damaged image due to transmission error. step; A second step of recording, by the encoder, for each block of the current frame whether it has been used for encoding specific blocks of the next frame; And a third step of encoding, in an intra mode, the blocks in the current image affected by the error-producing blocks according to whether the location of the error-prone image and the block has been transmitted from the decoder.

4. Important uses of the invention

The present invention is used in a video encoding apparatus.

Description

Error propagation prevention method due to transmission error in video coder}

The present invention relates to a method for preventing error propagation due to a transmission error in a video encoding apparatus and a computer-readable recording medium recording a program for realizing the method. In particular, in a channel where transmission errors of a transmission channel are frequent, such as a wireless channel, The present invention relates to a method for preventing error propagation by transmission error for application to real time video communication.

In general, a video encoding apparatus determines an appropriate search region in a previous image to estimate a motion vector of a 16x16 block, and finds a 16x16 block to be encoded of the current image within the range most similarly within the search region, and finds a relative displacement. Transmit as a motion vector.

However, even though the image is damaged due to transmission error in the previous image, the error continues from the previous image to the current image and the next image due to the continuous playback of the current image using a part of the damaged region by the moving compensation method. By propagation, there has been a problem that an image damaged by a transmission error continues to propagate until an encoded image in the next frame is obtained.

The present invention has been proposed to solve the above problems. In the block-based video encoding method using motion estimation and compensation such as MPEG-2, a block unit motion is prevented to prevent propagation of a damaged image due to a transmission error. The transmitter and the receiver store the trajectory of the signal for a predetermined frame period, and then the receiver receives the error block to the transmitter, and the transmitter encodes it in the intra mode to prevent the propagation of the error. It is an object of the present invention to provide a method of preventing and a computer-readable recording medium having recorded thereon a program for realizing the method.

1 is an exemplary configuration diagram of a video encoding apparatus using motion compensation prediction and transform encoding to which the present invention is applied.

* Explanation of symbols for the main parts of the drawings

110: input processing unit 120: motion estimation and compensation unit

130: encoding mode determiner 140: two-dimensional discrete cosine transform unit

150: quantization unit 160: inverse quantization unit

170: 2D inverse discrete cosine transform unit 180: previous image playback unit

190: Previous image memory 200: Scanner

210: variable length encoder

According to an aspect of the present invention, in order to prevent propagation of an image damaged by a transmission error, an encoder estimates and predicts a block-by-block motion between frames in a video encoding apparatus. A first step of encoding through compensation; A second step of recording, by the encoder, for each block of the current frame whether it has been used for encoding specific blocks of the next frame; And a third step of encoding, in an intra mode, blocks in the current image affected by the error-producing blocks according to whether an error-prone image and a block position are transmitted from a decoder. It is characterized by.

In addition, the present invention provides a video encoding apparatus having a processor to prevent propagation of an image damaged by a transmission error, comprising: a first function of encoding by a encoder through motion estimation and prediction compensation in units of blocks between frames; A second function for recording, for each block of the current frame, whether the encoder has been used to encode specific blocks of the next frame; And a third function for encoding, in an intra mode, the blocks in the current image affected by the error-producing blocks according to whether an error-prone image and a block position are transmitted from a decoder. A computer readable recording medium having recorded a program is provided.

If a transmission error occurs, this causes an error in the reproduced video, which continues to propagate until the next intra-frame coded (intra-encoded) video comes out.

Accordingly, in the present invention, in the block-based video encoding method using motion estimation prediction such as CCITT H.261, H.263 or ISO / IEC 13818-2 MPEG-2, in order to prevent propagation of a damaged image due to transmission error For example, the trajectory of block-based movement is stored in the transceiver for a predetermined frame period, and then the receiver receives the error block to the transmitter, and the transmitter encodes it in an intra mode to prevent the propagation of an error.

The above objects, features and advantages will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is an exemplary configuration diagram of a video encoding apparatus using motion compensation prediction and transform encoding according to the present invention.

As shown in FIG. 1, the video encoding apparatus to which the present invention is applied includes an input processor 110 that receives and transmits image-related data from the outside, and an input processor by using a previous image stored in the previous image memory 190. Receiving image-related data from the (110) receives the motion estimation and compensation unit 120 for motion estimation and compensation, and whether to encode the image data transferred from the input processing unit 110 in units of macro blocks (MB: Macro Block) The encoding mode determiner 130 determines whether to encode the image data transmitted from the motion estimation and compensator 120 in macroblock units, and the image data transferred from the encoding mode determiner 130 is two-dimensional. 2D discrete cosine transform unit 140 for discrete cosine transform (DCT) and discrete cosine transformed image data by discrete cosine transform unit 140 to quantize ( Two-dimensional image data output from the quantization unit 150 for quantization, the inverse quantization unit 160 for inverse quantization of the quantized image data through the quantization unit 150, and the inverse quantization unit 160 The 2D inverse discrete cosine transform unit 170 and the 2D inverse discrete cosine transform unit 170 for transforming the inverse discrete cosine transformed by the inverse discrete cosine transform unit 170 and the motion estimation and compensation unit 120 The previous image reproducing unit 180 for receiving the image data and reproducing the previous image, the previous image memory 190 for storing the previous image reproduced through the previous image reproducing unit 180, and the quantization unit 150 The scanner 200 for scanning the quantized image data in the form of a zig-zag, and the variable length encoder 210 for receiving the image data scanned through the scanner 200 and encoding and outputting the variable length. ).

Such a compression encoding method using the video encoding apparatus is a method commonly used in the IOU-T H.261, H.263 video telephone standards, or TV or HDTV video compression coding standards such as OSO / DEC PEG-2, Applications include two-way video communications such as video telephony, conferencing, and distance education, and one-way or distributed video transmission services such as digital TV and digital video storage media.

The operation of the video encoding apparatus to which the present invention to which the present invention having the above structure is applied is applied will be described in detail as follows.

In order to estimate the motion vector of the 16x16 block, the motion estimation and compensator 120 determines an appropriate search region in the previous image and finds the most similar place in the search region for the 16x16 block to be encoded of the current image within the range. Send the displacement as a motion vector.

At this time, a problem in the motion estimation and compensation process is that, even though the transmission error occurred in the previous image and the image was damaged, the error was transferred due to the playback of the current image by continuously using a part of the damaged region by the moving compensation method. It is the fact that it continues to propagate from the video to the current video and the next video.

This propagation is continued until an intra-coded picture that is encoded by only the discrete cosine transform through the two-dimensional discrete cosine transform 140 without moving compensation in the previous image.

In addition, most transmission errors can be detected using an error detection technique. However, since real-time video communication cannot normally require retransmission, decoding is performed as it is, or an error concealment technique is used. Therefore, error occurrence and propagation are inevitable.

Therefore, in the present invention, in order to prevent the propagation of such an error, the encoder records the trajectory of the movement of each processing Macro Block (MB) and transmits the location information of the MB or MB in which the error occurs at the receiving end. The present invention proposes a method for preventing the propagation of an error by tracking the positions of MBs in which an error is propagated and encoding the MBs in an intra mode.

Hereinafter, an error propagation prevention process due to a transmission error in the video encoding apparatus of the present invention as described above will be described in detail.

First, on the encoder side, a memory for recording k, k + 1, K + 2, ... which MB in the current video is used to decode which MB of the next video is used.

Now, assuming that the kth image is encoded (where the first image is usually encoded in the intra mode), the i + th vertically and the jth MB horizontally during the encoding process of the k + 1th image. It can be known which MB of the k-th image is encoded using the MB.

Then, i and j values, which are their indexes, are recorded in the trajectory memories of the MBs used for encoding MB (i, j) (k + 1). If the above process is performed for all MBs, the index memory Tmem (i, j) (k) will record the indexes of all the MBs in the k + 1th image encoded using itself.

This will similarly fill the trajectory memory of the k + 1th image during the encoding process of the K + 2th image. In addition, in the case of a certain MB, there may not be any MB of the next image encoded using the self.

Now, a method of using the trajectory memory will be described while describing an encoding and decoding process.

The data encoded by the encoder is usually output through a memory buffer at the output end.

The role of the memory buffer is to convert a variable data rate into an equality rate.

At this time, the size of the buffer is usually about 1/20 to 1/10 of the transfer rate.

For example, if the transmission rate is 10 Mbit / s, the buffer size is set around 0.5 to 1 Mbits.

Also, if the decoder also puts the same amount of buffer at the input and transfers the bit string input at the constant rate to the variable length decoder at the variable rate necessary for decoding, the transmission delay is caused by this buffer. Assuming that the buffer size is set to 1/15 of the transmission rate, the encoding and decoding delay of two frames occurs due to the buffer.

Therefore, when the encoder starts encoding the k + 2th image, the decoder starts decoding the kth image.

Next, assuming that a transmission error is detected in a part of the image when decoding the k-th image in the decoder.

First, data containing the corresponding error are discarded at the decoder, and data is usually obtained at the same position of the previous image in the error concealment process. Of course, this is not a complete concealment, so the image is already damaged.

At this time, the decoder immediately transmits the corresponding position in the damaged image to the appropriate channel on the encoder side.

Subsequently, the encoder encoding the k + 2th image finds the positions of the decoded k + 1th image blocks from the position of the error blocks of the kth image transferred from the decoder. The blocks are propagated blocks because they are decoded using the previously failed blocks at the decoder.

Then, the location of the blocks in the k + 2th image is determined by referring to the trajectory memory of the k + 1th image.

If the position has already been encoded, only the error block is recorded. If the block has not been encoded, the block is encoded in the intra mode.

The error block that has already been encoded is encoded in the intra mode or the motion compensation is performed at another position so as not to be decoded using the corresponding blocks when processing the k + 3th image.

Therefore, from the perspective of the decoder, some of the images damaged by the transmission error of the k-th image are propagated up to the k + 1 th image, but partially restored from the k + 2 th image, and completely restored from the k + 3 th image.

In summary, the present invention is intended to minimize the propagation of an error when a transmission error occurs when a decoder decodes an image transmitted by an encoder by encoding a block in real time video transmission.

Real-time video transmission means that the receiver directly sees the video transmitted by encoding such as a video call, a video conference, a broadcast, etc. within a very short time excluding the transmission delay. In contrast, non-real-time transmission usually checks the transmitted image after the transmitted image is stored in the receiver or in the storage device of the receiver, and is much longer than the data transmission time, such as Internet broadcasting or video mail. I can do it.

In the interframe video encoding method (ITU-T H.261 and H.263, ISO / IEC MPEG-1, MPEG-2, etc.), the problem of error propagation uses the past image to encode the current image, thus decoding the past image. The error is propagated to the decoding of the current image.

When transmitting real-time video, ARQ (Automatic Repeat reQuest) method is difficult to use basically. This is because the position of the video encoded by the encoder and the position of the video decoded by the decoder are largely different in the video encoding scheme. The reason for the difference is that the encoder stores the data encoded in the encoder in the buffer, and the data stored in the buffer is transmitted to the decoder in the form of FIFO (First In First Out). Large enough to hold Chapter 3). Therefore, when the decoder decodes the k-th video, the encoder is encoding a video that is k + 1 or k + 2 or higher. Therefore, even if the decoder requests an error when the k-th image is decoded and needs to be resent, it is practically difficult for the encoder to satisfy the request. This is because past image data has already disappeared in the encoder. Therefore, while trying to minimize the error by using an error correction code or the like, if a transmission error occurs, the error of the decoded video is inevitable. Here, it is considered that the error propagates to the next image due to the characteristics of the interframe encoding, which becomes an unobtrusive part until the next intra-mode image. Therefore, encoders and decoders need to prevent such error propagation as much as possible.

An existing method used to prevent error propagation is to periodically insert an image encoded in an intra mode or to perform intra mode encoding on a block basis after a predetermined time for each block to be encoded. If the period is shortened, the propagation of the error is shortened but the compression efficiency of the intra mode is lowered. Therefore, the overall compression efficiency is lowered. If the longer the period, the compression efficiency is relatively increased, but the propagation time is longer when an error occurs.

The present invention intends to more actively prevent the propagation of such an error, and when an error occurs during decoding at the decoder, the error location is transmitted to the encoder, and the encoder preferentially intra-spaces the area around the location. By encoding in the mode, the main means is to prevent error propagation in a short time. For example, suppose that video telephony transmits about 10 images per second and prevents the propagation of errors by sequentially encoding every block in every 99 images in intra mode. That is, about 5 seconds of error propagation. However, if the difference between the video encoding and the decoding time of the encoder and decoder corresponds to 2, that is, 0.2 seconds, the decoder informs the encoder of the position when an error occurs, and the encoder has the data. By encoding in the intra mode based on the position prior to encoding, the error propagation time can be limited to about 3 or more, that is, about 0.3 seconds. This method is the same as ARQ in that it indicates the location of an error, but ARQ is different from this method because it is based on the premise of retransmission. Since the encoder is encoded in the intra mode regardless of the position, there is a big difference in the effect as in the above example.

Therefore, although the present scheme uses the concept of the existing method of preventing error propagation by encoding a part of the ARQ scheme and the intra mode, the specific methods and effects may be different. In addition, the present method is based on a detailed understanding of the interframe video encoding scheme, the effects of transmission errors, and the structure of the encoder and decoder.

The method of the present invention as described above may be implemented as a program and stored in a computer-readable recording medium (CD-ROM, RAM, ROM, floppy disk, hard disk, magneto-optical disk, etc.).

Although the technical idea of the present invention has been described in detail according to the above preferred embodiment, it should be noted that the above-described embodiment is for the purpose of description and not of limitation. In addition, those skilled in the art will understand that various embodiments are possible within the scope of the technical idea of the present invention.

As described above, the present invention maintains the trajectory of block-by-block movement at the transmitting and receiving end for a predetermined frame period, and then, at the receiving end, informs the transmitting end of an error block, and at the transmitting end, encodes it in an intra mode to prevent the propagation of an error. As a result, propagation of a damaged image due to a transmission error can be prevented.

Claims (4)

In the video signal transmission device error transmission prevention method due to transmission error, In order to prevent propagation of an image damaged due to a transmission error, a first step of encoding by a encoder through block estimation of motion estimation and prediction compensation between frames; A second step of recording, by the encoder, for each block of the current frame whether it has been used for encoding specific blocks of the next frame; And A third step of encoding, in an intra mode, blocks in the current image affected by the error-producing blocks according to whether an error-prone image and a block position are transmitted from a decoder. Error propagation prevention method due to transmission error in the video encoding apparatus comprising a. The method of claim 1, In the third step, The encoder, A method for preventing error propagation due to transmission error in a video encoding apparatus, characterized by encoding blocks in a current image affected by error generating blocks using blocks of a previous frame in which no error has occurred, or encoding in an intra mode. The method according to claim 1 or 2, The third step, A fourth step of determining whether a location of an error image and a block are transmitted from the decoder; As a result of the determination in the fourth step, if the location of the error image and the block is transmitted from the decoder, the encoder selects the blocks of the previous frame that are not affected by the error occurrence blocks. A fifth step of encoding using the image or encoding in the intra mode; And A sixth step of skipping to the second step if the location of the image and the block in which the error occurs is not transmitted from the decoder as a result of the determination in the fourth step Error propagation prevention method due to transmission error in the video encoding apparatus comprising a. In order to prevent propagation of a damaged image due to a transmission error, in a video encoding apparatus having a processor, A first function of encoding, by the encoder, by frame estimation through motion estimation and prediction compensation between frames; A second function for recording, for each block of the current frame, whether the encoder has been used to encode specific blocks of the next frame; And A third function of encoding, in an intra mode, blocks in the current image affected by the error-producing blocks according to whether an error-prone image and a block position are transmitted from a decoder. A computer-readable recording medium having recorded thereon a program for realizing this.