CN101998117A - Video transcoding method and device - Google Patents

Abstract

The embodiment of the present invention discloses a video transcoding method and device. The video transcoding method includes: decoding the input code stream, extracting the macroblock information of the decoded input code stream, and determining the pixel according to the input code stream Transcoding rate; when the determined pixel transcoding rate is less than a preset first threshold value, encode the decoded input code stream according to the macroblock information; when the pixel transcoding rate is greater than When the first threshold is set, a macroblock coding mode is re-determined according to the macroblock information, and the decoded input code stream is coded according to the re-determined macroblock coding mode. The embodiments of the present invention increase the transcoding speed on the premise of ensuring the restored video quality, and better meet the requirements of real-time video transcoding.

Description

Video transcoding method and device

technical field

Embodiments of the present invention relate to the field of communication technologies, and in particular, to a video transcoding method and device.

Background technique

With the continuous development of mobile communication technology, the rapid increase of network access speed, the increasingly intelligent mobile terminal, and the increasingly optimized digital compression technology, the mobile terminal has developed from a simple communication and contact tool to a multimedia intelligent platform. . At the same time, traditional value-added services such as MMS and picture and ring download can no longer meet the needs of users. The emergence of mobile streaming media will change this situation. Mobile streaming media can provide users with services including video on demand, mobile video chat, and mobile video surveillance. Mobile streaming media is a combination of mobile communication and streaming media transmission, using streaming media technology on mobile terminals for data transmission through mobile networks. However, the diversity of mobile terminals, the volatility of wireless networks, and the processing capabilities of mobile terminals limit the development of mobile streaming media.

The video transcoding of code rate conversion can provide the most suitable video stream according to the processing capability of the mobile terminal and the conditions of the network bandwidth, so it is a key technology to promote the development of mobile streaming media. The most direct method of code rate conversion is to use cascaded full-decoding and full-coding transcoders. The full-decoding and full-coding transcoder is divided into two parts: first, the decoder completely decodes the input video stream, and then an encoder encodes and compresses the decoded video according to the target bit rate. By re-encoding the decoded video stream according to the target bit rate, the best image quality can be obtained.

In the process of realizing the present invention, the inventor found that the prior art has at least the following problems: the video transcoding scheme provided by the prior art needs to fully decode the input code stream, and then re-decode the decoded input code stream. Coding operations such as motion estimation and coding mode judgment have high computational complexity, which makes the transcoding speed low and cannot better meet the requirements of real-time video transcoding.

Contents of the invention

Embodiments of the present invention provide a video transcoding method and device to realize real-time selection of a macroblock encoding mode according to the pixel transcoding rate of the input code stream and output code stream, and increase the transcoding speed under the premise of ensuring the restoration of video quality .

An embodiment of the present invention provides a video transcoding method, including:

Decoding the input code stream, extracting the macroblock information of the decoded input code stream, and determining the pixel transcoding rate according to the input code stream;

When the determined pixel transcoding rate is less than a preset first threshold value, encode the decoded input code stream according to the macroblock information;

When the pixel transcoding rate is greater than the first threshold value, re-determine the macro-block coding mode according to the macro-block information, and code the decoded input code stream according to the re-determined macro-block coding mode .

The embodiment of the present invention also provides a video transcoding device, including:

Decoding module, used to decode the input code stream;

An extraction module, configured to extract the macroblock information of the input stream decoded by the decoding module;

A pixel transcoding rate determining module, configured to determine a pixel transcoding rate according to the input code stream;

An encoding module, configured to decode the input of the decoding module according to the macroblock information extracted by the extracting module when the pixel transcoding rate determined by the pixel transcoding rate determination module is less than a preset first threshold value Encode the code stream;

An encoding mode determination module, configured to re-determine the macroblock encoding mode according to the macroblock information extracted by the extraction module when the pixel transcoding rate determined by the pixel transcoding rate determination module is greater than the first threshold value; The encoding module is further configured to encode the input code stream decoded by the decoding module according to the macroblock encoding mode re-determined by the encoding mode determining module.

In the embodiment of the present invention, the macroblock encoding mode is determined in real time according to the pixel transcoding rate and the macroblock information of the input code stream, and then encoding is performed according to the determined macroblock encoding mode, thereby improving the conversion rate while ensuring the restored video quality. The coding speed better meets the requirements of real-time video transcoding.

Description of drawings

In order to more clearly illustrate the present invention or the technical solutions in the prior art, the accompanying drawings that need to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings in the following description are only the present invention. For some embodiments of the invention, those skilled in the art can also obtain other drawings based on these drawings without creative effort.

FIG. 1 is a flowchart of an embodiment of the video transcoding method of the present invention;

FIG. 2 is a flowchart of another embodiment of the video transcoding method of the present invention;

FIG. 3 is a schematic structural diagram of an embodiment of a video transcoding device according to the present invention;

FIG. 4 is a schematic structural diagram of another embodiment of the video transcoding device of the present invention;

Fig. 5 is the comparative schematic diagram of PSNR performance of Carphone.qcif sequence of the present invention;

Fig. 6 is a comparative schematic diagram of the transcoding speed of the Carphone.qcif sequence of the present invention;

Fig. 7 is a schematic diagram of PSNR performance comparison of the Foreman.cif sequence of the present invention;

Fig. 8 is a schematic diagram of the comparison of the transcoding speed of the Foreman.cif sequence of the present invention.

Detailed ways

The technical solutions in the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

An embodiment of the present invention provides a video transcoding method. By multiplexing information such as image information, motion vectors, macroblock coding types, and prediction residuals in the input code stream, operations such as coding mode judgment and motion estimation are simplified to reduce the cost of transcoding. Computational complexity, improve transcoding speed.

Fig. 1 is a flowchart of an embodiment of the video transcoding method of the present invention, as shown in Fig. 1, including:

Step 101, decode the input code stream, extract the macroblock information of the decoded input code stream, and determine the pixel transcoding rate according to the input code stream.

Step 102, when the determined pixel transcoding rate is less than a preset first threshold value, encode the decoded input code stream according to the macroblock information.

Among them, the pixel conversion rate is the ratio of the average number of bits per pixel in the input code stream to the average number of bits per pixel in the preset output code stream, where the average number of bits per pixel in the preset output code stream The number of bits is a value representing the desired output, which is related to relevant indicators in the transcoding process and relevant information in the input code stream (such as code rate, etc.). The pixel transcoding rate indicates the degree of similarity between the input code stream and the expected output code stream. The smaller the pixel transcoding rate, the higher the similarity between the input code stream and the expected output code stream.

When the pixel conversion rate is less than the preset first threshold value, there is a strong similarity between the decoded input code stream and the expected output code stream, so there is no need to make major adjustments to the input code stream, The decoded input code stream may be encoded according to the macroblock information of the decoded input code stream.

Step 103, when the pixel transcoding rate is greater than the first threshold value, re-determine the macro-block coding mode according to the macro-block information, and code the decoded input stream according to the newly-determined macro-block coding mode.

Specifically, when the pixel transcoding rate is greater than the first threshold value, the difference between the decoded input code stream and the expected output code stream is relatively large, so according to the macroblock information of the decoded input code stream The decoded input stream is encoded, and the video quality of the transcoded output stream cannot be guaranteed. In this case, it is necessary to re-determine the macroblock coding mode according to the macroblock information in the decoded input code stream, and then encode the decoded input code stream according to the newly determined macroblock coding mode.

In this embodiment, the pixel conversion rate can be determined according to the average number of bits occupied per pixel in the input code stream and the preset average number of bits occupied per pixel in the output code stream.

Specifically, the average number of bits per pixel in the code stream is bpp, and bpp can be calculated by formula (1),

bpp＝R/(F×W×H)       (1)

In formula (1), R identifies the code rate of the code stream, F represents the frame rate of the code stream, W and H represent the width and height of the image, respectively.

After decoding the input code stream, the code rate and frame rate of the input code stream can be obtained, as well as the width and height of the image in the decoded input code stream, and the average The number of bits occupied by a pixel; the above is only a way to obtain the code rate and frame rate of the input code stream, and the width and height of the image in the input code stream, and then obtain the average number of bits occupied by each pixel in the input code stream. The present invention The embodiment is not limited thereto, the code rate and frame rate of the input code stream, and the width and height of the image in the input code stream may not be obtained after decoding, and the above parameters can also be obtained in other ways, for example: pre-input The above parameters; the embodiments of the present invention do not limit the way to obtain the code rate and frame rate of the input code stream, and the width and height of the image in the input code stream.

In addition, the code rate and frame rate of the output code stream can be preset according to the transcoding requirements, as well as the width and height of the image in the output code stream. According to these parameters, the average bit occupied by each pixel in the preset output code stream can be calculated. number.

Among them, the macroblock information includes: block type information, block mode information, motion vector information, block coding type (Coded Block Pattern; hereinafter referred to as: CBP) information, quantization parameters, and macroblock residual information required for coding; The block coding mode includes: block type information, block mode information and motion vector information.

The above-mentioned steps 102 and 103 are executed in no particular order, and may be executed in parallel.

In this embodiment, the macroblock encoding mode is determined in real time according to the pixel transcoding rate and the macroblock information of the input code stream, and then encoding is performed according to the determined macroblock encoding mode, thereby improving the transcoding while ensuring the restored video quality. The speed better meets the requirements of real-time video transcoding.

Fig. 2 is a flowchart of another embodiment of the video transcoding method of the present invention, as shown in Fig. 2, including:

Step 201, decode the input code stream, and extract the macroblock information of the decoded input code stream.

Step 202: Determine the pixel transcoding rate according to the average number of bits per pixel in the input code stream and the preset average number of bits per pixel in the output code stream.

Specifically, the average number of bits occupied by each pixel in the code stream is bpp, and bpp can be calculated by formula (1).

After decoding the input code stream, the code rate and frame rate of the input code stream can be obtained, as well as the width and height of the image in the decoded input code stream, and the average The number of bits occupied by a pixel; the above is only a way to obtain the code rate and frame rate of the input code stream, and the width and height of the image in the input code stream, and then obtain the average number of bits occupied by each pixel in the input code stream. The present invention The embodiment is not limited thereto, the code rate and frame rate of the input code stream, and the width and height of the image in the input code stream may not be obtained after decoding, and the above parameters can also be obtained in other ways, for example: pre-input The above parameters; the embodiments of the present invention do not limit the way to obtain the code rate and frame rate of the input code stream, and the width and height of the image in the input code stream.

In addition, the code rate and frame rate of the output code stream can be preset according to the transcoding requirements, as well as the width and height of the image in the output code stream. According to these parameters, the average bit occupied by each pixel in the preset output code stream can be calculated. number.

The ratio of the average number of bits per pixel in the input code stream to the preset average number of bits per pixel in the output code stream is the pixel conversion rate. Use ω to represent the pixel transcoding rate, bpp _i to represent the average number of bits per pixel in the input code stream, and bpp _o to represent the average number of bits per pixel in the preset output code stream, then,

ω＝bpp _i /bpp _o (2)

The pixel transcoding rate indicates the degree of similarity between the input code stream and the expected output code stream. The smaller the pixel transcoding rate, the higher the similarity between the input code stream and the expected output code stream.

Step 203, judging whether the pixel transcoding rate is greater than a preset first threshold. If the pixel transcoding rate is less than the first threshold value, perform steps 204 to 206; if the pixel transcoding rate is greater than the first threshold value, perform step 207.

In this embodiment, the first threshold value α is preset, and the value of α here is an empirical value, and α can take multiple values within a certain range, and the optimal number can be obtained after multiple simulations and debugging, such as α can be 0.5, or 0.45, etc. Other threshold values in this paper can also be obtained by similar methods. After setting the first threshold value α, it is judged whether the pixel transcoding rate ω is greater than α. If ω>α, it means that the difference between the decoded input code stream and the expected output code stream is large, so the encoding of the decoded input code stream according to the macroblock information of the decoded input code stream cannot guarantee The video quality of the transcoded output stream. In this case, it is necessary to re-determine the macroblock coding mode according to the macroblock information in the decoded input code stream, and then encode the decoded input code stream according to the newly determined macroblock coding mode;

If ω<α, it means that there is a strong similarity between the decoded input code stream and the expected output code stream, so there is no need to make major adjustments to the input code stream. The macroblock information encodes the decoded input stream.

Step 204: After encoding each frame of the decoded input code stream according to the macroblock information of the decoded input code stream, calculate the first distortion degree of the coded output frame, and the corresponding input frame distortion degree of the output frame second degree of distortion, and calculate the ratio of the first degree of distortion to the second degree of distortion.

The input code stream is composed of images frame by frame. Each frame image includes multiple macroblocks. The header bits in the macroblocks represent the bits occupied by information such as macroblock coding mode and motion vector. The length of the header bits is the same as The quantization parameter (Quantization Parameter; hereinafter referred to as: QP) is basically irrelevant; while the texture bits in the macroblock refer to the bits occupied by the macroblock residual, and the length of the texture bits is closely related to the QP.

When ω<α, the decoded input code stream can be encoded directly according to the macroblock information of the decoded input code stream. During encoding, the motion information of the input macroblock is fully multiplexed, so the number of header bits of the output macroblock is very close to the number of header bits of the corresponding input macroblock. At this time, the main difference between the coded output code stream and the input code stream lies in the number of texture bits. Since code rate transcoding is usually code rate reduction transcoding, that is, the QP used by the encoded output macroblock is generally not less than the QP of the input macroblock, that is, the number of texture bits of the output macroblock will be less than the texture bits of the input macroblock number. When the QP of the output macroblock is slightly larger than the QP of the input macroblock, the number of texture bits of the output macroblock is slightly smaller than the number of texture bits of the input macroblock, and the code rate of the output code stream is not much different from the code rate of the input code stream , so the quality of the output stream can be guaranteed. When the QP of the output macroblock is larger than the QP of the input macroblock, the number of texture bits of the output macroblock is much smaller than the number of texture bits of the input macroblock, and the code rate of the output code stream is relatively different from that of the input code stream. If it is too large, the quality of the output code stream will be greatly reduced.

In order to ensure the video quality of the coded output code stream, in this embodiment, after each frame of the decoded input code stream is encoded according to the macroblock information of the decoded input code stream, the value of the coded output frame is calculated. A ratio of the first degree of distortion to the second degree of distortion of the input frame corresponding to the output frame. Specifically, it can be:

First, calculate the first ratio of the number of texture bits of each macroblock in the encoded output frame to the total number of bits of the macroblock, and calculate the average value of the first ratios of all macroblocks in the encoded output frame, the first The mean value of a ratio is the first degree of distortion; specifically, the first ratio is represented by R(M _o ),

$\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; o</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; Text</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; o</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; Total</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; o</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

In formula (3), Tex(M _o ) is the number of texture bits of the macroblock in the output frame, and Total(M _o ) is the total number of bits of the macroblock. After the first ratio of each macroblock in the output frame is calculated, the average value of the first ratios of all macroblocks in the output frame is calculated, and the average value of the first ratio is the first degree of distortion R _o .

Then, calculate the second ratio of the number of texture bits of each macroblock in the input frame corresponding to the above output frame to the total number of bits of the macroblock, and calculate the average value of the second ratios of all macroblocks in the input frame, the first The average of the two ratios is the second degree of distortion; specifically, the second ratio is represented by R(M _i ),

$\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; Text</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; Total</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

In formula (4), Tex(M _i ) is the number of texture bits of the macroblock in the input frame, and Total(M _i ) is the total number of bits of the macroblock. After calculating the second ratio of each macroblock in the input frame, calculate the average value of the second ratios of all macroblocks in the input frame, and the average value of the second ratio is the second distortion degree R _i .

Finally, the ratio R _o /R _i of the first distortion degree to the second distortion degree is calculated.

Step 205, judging whether the ratio of the first distortion level to the second distortion level is smaller than a preset second threshold value. If the degree of distortion is smaller than the preset second threshold value, perform step 207; if the degree of distortion is greater than the second preset threshold value, perform step 206.

In this embodiment, the second threshold value β is set in advance, where the value of β is an empirical value, and β can take multiple values within a certain range, and the optimal number can be obtained after multiple simulations and debugging, such as β Can take 0.55 or 0.65 and so on. After setting the second threshold value β, it is judged whether the ratio R _o /R _i of the first distortion degree to the second distortion degree is smaller than β. If R _o /R _i <β, it can be determined that the video quality of the output code stream has dropped too much, and the subsequent code streams to be encoded in the decoded input code stream cannot be further processed according to the macroblock information of the decoded input code stream To encode, it is necessary to re-determine the macroblock coding mode according to the macroblock information of the decoded input code stream, and then encode the subsequent code stream to be coded according to the re-determined macroblock coding mode until the decoded input code The coding of the stream ends; if R _o /R _i > β, it can be determined that the video quality of the output code stream meets the predetermined transcoding requirements, and the decoded input code stream can be continued according to the macroblock information of the decoded input code stream After encoding, after each frame is encoded, R _o /R _i is calculated, and the relationship between R _o /R _i and β is judged. As long as R _o /R _i > β, you can continue to use the decoded input The macroblock information of the code stream encodes the decoded input code stream, and once R _o /R _i <β, it is necessary to re-determine the macroblock coding mode according to the macroblock information of the decoded input code stream, and then according to the re-determined The macroblock encoding mode encodes subsequent code streams to be encoded in the decoded input code stream until the encoding of the decoded input code stream ends.

Step 206, encode the decoded input code stream according to the macroblock information of the decoded input code stream.

Preferably, when the decoded input code stream is encoded according to the macroblock information of the decoded input code stream, CBP information in the macroblock information can also be extracted, and the main function of the CBP information is to indicate the macroblock residual ( Luminance and chrominance) after discrete cosine transform (Discrete Cosine Transform; hereinafter referred to as: DCT) and quantized part is zero. Since the motion information of the input code stream is completely multiplexed when encoding according to the macroblock information of the decoded input code stream, it can be approximately considered that the residual of the encoded output macroblock is similar to the residual distribution of the input macroblock At this time, it can be considered that the part of the output macroblock that is zero is still zero after encoding and transcoding, so the macroblock whose CBP information is indicated as zero in the frame to be encoded can be directly entropy encoded, so that the CBP information indication can be omitted Time for performing DCT, quantization, inverse quantization, and inverse discrete cosine transform (Inverse Discrete Cosine Transform; hereinafter referred to as: IDCT) for a macroblock that is zero. Using this method can increase the transcoding speed, but at the same time it will also reduce the peak signal-to-noise ratio (Peak Signal to Noise Ratio; hereinafter referred to as: PSNR) value of the encoded output stream.

Step 207, re-determine the macroblock coding mode according to the macroblock information of the decoded input code stream, and code the subsequent code stream to be coded according to the re-determined macroblock coding mode.

When ω>α or R _o /R _i <β, it can be determined that the video quality of the output code stream has degraded too much, and the decoded input code stream cannot be encoded according to the macroblock information of the decoded input code stream. However, it is necessary to re-determine the macroblock coding mode according to the macroblock information of the decoded input code stream, and then encode the subsequent code streams in the decoded input code stream according to the re-determined macro block coding mode until the decoding After the encoding of the input code stream ends.

For example, when ω>α or R _o /R _i <β, the block type of the macroblock to be encoded can be determined according to the macroblock information of the decoded input code stream, and then different macroblock coding can be selected according to different block types mode, etc. The macroblock information here includes: block type information, block mode information, motion vector information, CBP information, quantization parameters, and macroblock residuals and other information required for encoding; the macroblock coding mode includes: Block mode information and motion vector information.

The above-mentioned steps 201 to 207 are executed for each frame in the decoded input code stream until the encoding of the decoded input code stream ends.

In this embodiment, after the pixel transcoding rate is determined, before judging whether the pixel transcoding rate is greater than the preset first threshold value, that is, between step 202 and step 203, it may also be based on the decoded input The macroblock information in the code stream determines whether the macroblock to be encoded is a Skip block; if the macroblock to be encoded is a Skip block, the Skip block can be skipped and the next macroblock of the Skip block can be directly encoded , so that the motion estimation and the subsequent encoding process can be omitted, and the transcoding speed can be greatly improved.

In the H.264 predicted frame, the rate-distortion optimization method is used to select the coding mode of the macroblock, that is, the prediction error and the information coding rate are combined to select the coding mode:

J _mode (MB _{x, y} , I _{x, y} ) = D _REC (MB _{x, y} , I _{x, y} ) + λR _REC (B _{x, y} , I _{x, y} ) (5)

In formula (5), I _{x, y} represent the optional prediction modes of MB _{x, y} , including intra-frame coding mode and inter-frame coding mode, and the value of Lagrangian parameter λ is λ=2 ^QP/6-2 (QP is quantization parameter), R _REC and D _REC represent the number of coded bits in the corresponding prediction mode and the distortion of the decoded restored image caused by quantization, respectively. For the mth macroblock M(F, m) in the frame F, the optimal rate-distortion cost obtained after mode decision and motion estimation is denoted as RD(M(F, m)), then for the Skip mode, since Almost no bits are occupied in entropy coding, so its rate-distortion cost function is:

RD _Skip (M(F, m)) = D _REC (M(F, m), Skip) (6)

Note that the current frame on the decoding side is F _DC , the reference frame corresponding to the current frame on the decoding side is F _DR , the current frame on the encoding side is F _EC , and the reference frame corresponding to the current frame on the encoding side is F _ER . If formula (7) holds true, then the macroblock to be coded can be coded as a Skip block, where the value of ρ is an empirical value, and ρ can take multiple values within a certain range, and the best value can be obtained after multiple simulations and debugging. Quantity, such as ρ can take 0.55 or 0.5 and so on.

${\mathrm{<span\; class="notranslate">\; RD</span>}}_{\mathrm{<span\; class="notranslate">\; Skip</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; EC</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; RD</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; ER</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>\frac{\mathrm{<span\; class="notranslate">\; RD</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; DC</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; RD</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; DR</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

The method is used to predict the encoding mode of the macroblock of the input code stream. If the macroblock to be encoded is a Skip block, motion estimation and subsequent encoding process will not be performed on the macroblock to be encoded, thereby increasing the transcoding speed.

In the embodiment of the present invention, the macroblock encoding mode is determined in real time according to the pixel transcoding rate and the macroblock information of the input code stream, and then encoding is performed according to the determined macroblock encoding mode, thereby improving the conversion rate while ensuring the restored video quality. The code speed better meets the requirements of real-time video transcoding. At the same time, the embodiment of the present invention also predicts the encoding mode of the macroblock of the input code stream. If the macroblock to be encoded is a Skip block, the awaiting Coded macroblocks are used for motion estimation and subsequent encoding process, thus further improving the transcoding speed.

Those of ordinary skill in the art can understand that all or part of the steps for realizing the above-mentioned method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the It includes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

FIG. 3 is a schematic structural diagram of an embodiment of a video transcoding device of the present invention. The video transcoding device of this embodiment may be a communication device, or be located in a communication device, and the communication device may be a gateway, a media server, or a digital TV. As shown in FIG. 3 , the video transcoding device may include: a decoding module 31 , an extraction module 32 , a pixel transcoding rate determination module 33 , an encoding module 34 and an encoding mode determination module 35 .

Wherein, the decoding module 31 is used to decode the input stream;

The extracting module 32 can extract the macroblock information of the input stream decoded by the decoding module 31;

The pixel transcoding rate determination module 33 can determine the pixel transcoding rate according to the input code stream;

The encoding module 34 can encode the input code stream decoded by the decoding module 31 according to the macroblock information extracted by the extracting module 32 when the pixel transcoding rate determined by the pixel transcoding rate determination module 33 is less than the preset first threshold value ; When the pixel conversion rate is less than the preset first threshold value, there is a strong similarity between the decoded input code stream and the expected output code stream, so there is no need to make major adjustments to the input code stream , the encoding module 34 can encode the decoded input stream according to the macroblock information of the decoded input stream;

The encoding mode determining module 35 may re-determine the macroblock encoding mode according to the macroblock information extracted by the extracting module 32 when the pixel transcoding rate determined by the pixel transcoding rate determining module 33 is greater than the first threshold value; at this time, the encoding module 34 The input code stream decoded by the decoding module 31 can also be encoded according to the macroblock encoding mode re-determined by the encoding mode determination module 35; specifically, when the pixel transcoding rate is greater than the first threshold value, the decoded input code stream The difference between the stream and the expected output stream is large, so encoding the decoded input stream based on the macroblock information of the decoded input stream cannot guarantee the video quality of the transcoded output stream. In this case, the coding mode determination module 35 needs to re-determine the macroblock coding mode according to the macroblock information in the input code stream, and the coding module 34 encodes the decoded input code stream according to the newly determined macroblock coding mode.

The macroblock information in this embodiment includes: information required for encoding such as block type information, block mode information, motion vector information, CBP information, quantization parameters, and macroblock residuals; the macroblock coding mode includes: block type information, Block mode information and motion vector information.

In this embodiment, when the pixel transcoding rate determined by the pixel transcoding rate determining module 33 is lower than the preset first threshold value, the encoding module 34 can decode the macroblock information obtained by the decoding module 31 according to the macroblock information extracted by the extracting module 32. The input code stream is encoded; when the pixel transcoding rate determined by the pixel transcoding rate determination module 33 is greater than the preset first threshold value, the encoding mode determination module 35 determines the input code stream according to the pixel transcoding rate and the macro of the decoded input code stream The block information determines the macroblock encoding mode in real time, and then the encoding module 34 performs encoding according to the macroblock encoding mode determined by the encoding mode determination module 35, thereby improving the transcoding speed and better satisfying Requirements for real-time video transcoding.

Fig. 4 is a schematic structural diagram of another embodiment of the video transcoding device of the present invention. The video transcoding device of this embodiment may be a communication device, or be located in a communication device, and the communication device may be a gateway, a media server or a digital TV, etc. . As shown in FIG. 4 , the video transcoding device includes: a decoding module 41 , an extraction module 42 , a pixel transcoding rate determination module 43 , an encoding module 44 , an encoding mode determination module 45 and a Skip block determination module 46 .

Wherein, the decoding module 41 is used to decode the input stream;

The extraction module 42 can extract the macroblock information of the input stream decoded by the decoding module 41;

The pixel transcoding rate determining module 43 can determine the pixel transcoding rate according to the input code stream; the pixel transcoding rate determining module 43 can specifically determine the pixel transcoding rate according to the average number of bits per pixel in the input code stream and the preset average number of bits per pixel in the output code stream. The number of bits occupied determines the pixel transcoding rate; specifically, the average number of bits occupied per pixel in the code stream is bpp, and bpp can be calculated by formula (1). After the decoding module 41 decodes the input code stream, the input code stream can be obtained. The code rate and frame rate of the code stream, as well as the width and height of the image in the decoded input code stream, according to these parameters, the average number of bits per pixel in the decoded input code stream can be calculated; the above is just to obtain the input code The code rate and frame rate of the stream, as well as the width and height of the image in the input code stream, and then obtain the average number of bits per pixel in the input code stream. Embodiments of the present invention are not limited thereto. The input code stream The bit rate and frame rate, as well as the width and height of the image in the input bit stream are not necessarily obtained after decoding, and the above parameters can also be obtained in other ways, for example: input the above parameters in advance. The embodiment of the present invention does not limit the way of obtaining the code rate and frame rate of the input code stream, and the width and height of the image in the input code stream.

In addition, the code rate and frame rate of the output code stream can be preset according to the transcoding requirements, as well as the width and height of the image in the output code stream. According to these parameters, the average bit occupied by each pixel in the preset output code stream can be calculated. number.

The encoding module 44 can encode the input code stream decoded by the decoding module 41 according to the macroblock information extracted by the extracting module 42 when the pixel transcoding rate determined by the pixel transcoding rate determination module 43 is less than the preset first threshold value ; When the pixel conversion rate is less than the preset first threshold value, there is a strong similarity between the decoded input code stream and the expected output code stream, so there is no need to make major adjustments to the input code stream , the encoding module 44 can encode the decoded input stream according to the macroblock information of the decoded input stream;

The encoding mode determining module 45 may re-determine the macroblock encoding mode according to the macroblock information extracted by the extracting module 42 when the pixel transcoding rate determined by the pixel transcoding rate determining module 43 is greater than the first threshold value; at this time, the encoding module 44 The input code stream decoded by the decoding module 41 can also be encoded according to the macroblock encoding mode re-determined by the encoding mode determination module 45; specifically, when the pixel transcoding rate is greater than the first threshold value, the decoded input code stream The difference between the stream and the expected output stream is large, so encoding the decoded input stream based on the macroblock information of the decoded input stream cannot guarantee the video quality of the transcoded output stream. In this case, the coding mode determining module 45 needs to re-determine the macroblock coding mode according to the macroblock information in the input code stream, and the coding module 44 encodes the decoded input code stream according to the newly determined macroblock coding mode.

Wherein, the encoding module 44 may include a distortion calculation submodule 441 , a first encoding submodule 442 and a second encoding submodule 443 .

The distortion calculation sub-module 441 is used to calculate the first distortion of the encoded output frame after encoding each frame of the input code stream decoded by the decoding module 41 according to the macroblock information extracted by the extraction module 42, and the The second distortion degree of the input frame corresponding to the output frame, and calculate the ratio of the first distortion degree to the second distortion degree; specifically, the distortion degree calculation sub-module 441 can calculate the texture bit of each macroblock in the encoded output frame Number and the first ratio of the total number of bits of the macroblock, and calculate the mean value of the first ratio of all macroblocks in the output frame after encoding, the mean value of the first ratio is the first degree of distortion; degree of distortion calculation submodule 441 may also calculate the second ratio of the number of texture bits of each macroblock in the input frame corresponding to the output frame to the total number of bits of the macroblock, and calculate the mean value of the second ratio of all macroblocks in the input frame, the second ratio The mean of is the second degree of distortion.

The first encoding submodule 442 can encode the input code stream decoded by the decoding module 41 according to the macroblock information extracted by the extracting module 42 when the ratio calculated by the distortion calculation submodule 441 is greater than the preset second threshold value; The coding mode determination module 45 is also used to re-determine the macroblock coding mode according to the macroblock information extracted by the extraction module 42 when the ratio calculated by the distortion degree calculation submodule 441 is smaller than the above-mentioned second threshold value; at this time, the second coding submodule 443 may encode subsequent to-be-coded code streams in the decoded input code stream according to the macroblock coding mode re-determined by the coding mode determination module 45 until the encoding of the decoded input code stream ends.

Specifically, when the ratio calculated by the distortion degree calculation sub-module 441 is less than the preset second threshold value, it can be determined that the video quality of the output code stream has degraded too much, and it is impossible to continue decoding according to the macroblock information extracted by the extraction module 42. The final input code stream is encoded, and the encoding mode determination module 45 is required to redetermine the macroblock encoding mode according to the macroblock information extracted by the extracting module 42, and then the second encoding submodule 443 performs decoding on the decoded macroblock encoding mode according to the newly determined macroblock encoding mode In the input code stream, the subsequent code stream to be encoded is encoded until the encoding of the decoded input code stream is completed; when the ratio calculated by the distortion degree calculation submodule 441 is greater than the preset second threshold value, the output can be determined The video quality of the code stream meets the predetermined transcoding requirements, so the first encoding submodule 442 can continue to encode the input code stream decoded by the decoding module 41 according to the macroblock information extracted by the extraction module 42, and then encode each frame Afterwards, the degree of distortion calculation submodule 441 will calculate the ratio of the first degree of distortion to the second degree of distortion, as long as the ratio calculated by the degree of distortion calculation submodule 441 is greater than the preset second threshold value, the first encoding submodule 442 can continue to encode the decoded input code stream according to the macroblock information extracted by the extraction module 42, and once the ratio calculated by the distortion degree calculation sub-module 441 is less than the preset second threshold value, the encoding mode determination module is required 45 to redetermine the macroblock encoding mode according to the macroblock information extracted by the extracting module 42, and then the second encoding sub-module 443 performs subsequent coding in the decoded input code stream according to the macroblock encoding mode re-determined by the encoding mode determination module 45 The code stream is encoded until the encoding of the decoded input code stream ends.

In this embodiment, the extracting module 42 can extract block coding type CBP information in the macroblock information; the first coding submodule 442 is specifically configured to directly perform entropy coding on the macroblock whose CBP information indicates zero.

In this embodiment, after the pixel transcoding rate determining module 43 determines the pixel transcoding rate, the Skip block determining module 46 may first determine whether the macroblock to be encoded is a Skip block according to the macroblock information extracted by the extracting module 42; , the encoding module 44 can no longer perform motion estimation and subsequent encoding process on the macroblock to be encoded when the Skip block determination module 46 determines that the macroblock to be encoded is a Skip block, and directly the next macroblock of the Skip block Encoding, thus greatly improving the transcoding speed.

Specifically, the Skip block determination module 46 determines that the macroblock to be encoded is a Skip block step comprising:

When the mth macroblock in frame F is M(F, m), the optimal rate-distortion cost obtained after mode decision and motion estimation is RD(M(F, m)), and the current frame on the decoding side is F _DC , the reference frame corresponding to the current frame on the decoding side is F _DR , the current frame on the encoding side is F _EC , the reference frame corresponding to the current frame on the encoding side is F _ER , the RD _Skip (M(F _EC , m)) of the macroblock to be encoded satisfy

${\mathrm{RD}}_{\mathrm{Skip}}\left(m(f_{\mathrm{EC}},m)\right)<\mathrm{\&rho;}\&Center\; Dot;\mathrm{RD}\left(m(f_{\mathrm{ER}},m)\right)\&CenterDot;\frac{\mathrm{RD}\left(m(f_{\mathrm{DC}},m)\right)}{\mathrm{RD}\left(m(f_{\mathrm{DR}},m)\right)}$ hour,

The Skip block determination module 46 determines that the macroblock to be encoded is a Skip block; wherein, p is a positive number greater than 0, and RD _Skip (M(F _EC , m)) is the optimal rate-distortion cost of the macroblock to be encoded,

RD _Skip (M(F _EC ,m))=D _REC (M(F _EC ,m),Skip).

It should be noted that the above-mentioned modules can be implemented by a processor or other hardware, for example, by a central processing unit (Central Processing Unit; hereinafter referred to as: CPU) of a personal computer (Personal Computer; hereinafter referred to as: PC), or a digital Signal processor (Digital Signal Processor; hereinafter referred to as: DSP), or Field Programmable Gate Array (Field Programmable Gate Array; hereinafter referred to as: FPGA), or a dedicated transcoding processor or other hardware implementation.

In the embodiment of the present invention, the macroblock encoding mode is determined in real time according to the pixel transcoding rate and the macroblock information of the input code stream, and then encoding is performed according to the determined macroblock encoding mode, thereby improving the conversion rate while ensuring the restored video quality. The coding speed better meets the requirements of real-time video transcoding. At the same time, the embodiment of the present invention also predicts the encoding mode of the macroblock of the input code stream. If the macroblock to be encoded is a Skip block, the motion estimation and subsequent encoding process of the macroblock to be encoded will no longer be performed, thereby further Improved transcoding speed.

Embodiments of the present invention provide a video transcoding method and device, and the transcoding speed can be greatly improved through the method and device provided by the embodiments of the present invention. The embodiment of the present invention uses two standard test sequences of quarter common intermediate resolution image format (Quarter Common Intermediate Format; hereinafter referred to as: QCIF) and common intermediate resolution image format (Common Intermediate Format; hereinafter referred to as: CIF): "Carphone.qcif" and "Foreman.cif" each take 300 frames, respectively use the cascaded transcoder with full search algorithm, full solution and full encoding, and the video transcoding device proposed by the embodiment of the present invention to perform code rate transcoding.

For the QCIF format sequence in the experiment, first generate a test code stream with a code rate of 768kbps and a frame rate of 25 frames per second, and then generate an initial target code rate through the cascaded transcoder and the video transcoding device of the embodiment of the present invention respectively It is 64kbps, and each increment increases by 64kbps until the output code stream of 768kbps. The frame rate of the output code stream is 25 frames per second by default. Record the transcoding results of the corresponding cascaded transcoder and the video transcoding device of the embodiment of the present invention . For the CIF format sequence, first generate a test code stream with a code rate of 2024kbps and a frame rate of 25 frames per second, and then generate the initial target code rate through the cascade transcoder and the video transcoding device of the embodiment of the present invention respectively 64kbps, increasing by 64kbps each time until the output code stream of 1984kbps, recording the transcoding results of the corresponding cascaded transcoders and the video transcoding device of the embodiment of the present invention. In the experiment, α = 5, β = 0.55, and ρ = 0.55.

Fig. 5 is the comparative schematic diagram of the PSNR performance of the Carphone.qcif sequence of the present invention; Fig. 6 is the comparative schematic diagram of the transcoding speed of the Carphone.qcif sequence of the present invention; Fig. 7 is the comparative schematic diagram of the PSNR performance of the Foreman.cif sequence of the present invention; Fig. 8 is this Schematic diagram of the transcoding speed comparison of the invented Foreman.cif sequence. In FIGS. 5 to 8 , -◆- represents the transcoding result obtained by the cascaded transcoder of full-decoding and full-coding of the full search algorithm, and -■- represents the transcoding result of the video transcoding device according to the embodiment of the present invention.

Figures 5 to 8 respectively show the performance comparison results of the "Carphone.qcif" and "Foreman.cif" sequences being transcoded by the video transcoding device and the cascaded transcoder according to the embodiment of the present invention. It can be seen from Fig. 5 and Fig. 7 that, compared with the PSNR performance of the cascaded transcoder, the PSNR performance of the video transcoding device according to the embodiment of the present invention is about 1dB when the target bit rate is low, and the PSNR difference is about 1dB when the target bit rate is high. When the PSNR difference is within 0.5dB; it can be seen from Figure 6 and Figure 8 that compared with the cascaded transcoder, the video transcoding device of the embodiment of the present invention can increase the transcoding speed by about 45%, up to 60 %. Therefore, the video transcoding method and device proposed in the embodiment of the present invention is a better video bit rate transcoding solution, which can be used in actual video communication systems after comprehensively considering the restoration of video quality after transcoding and the speed of transcoding operations.

Those skilled in the art can understand that the drawing is only a schematic diagram of a preferred embodiment, and the modules or processes in the drawing are not necessarily necessary for implementing the present invention.

Those skilled in the art can understand that the modules in the device in the embodiment can be distributed in the device in the embodiment according to the description in the embodiment, or can be located in one or more devices different from the embodiment according to corresponding changes. The modules in the above embodiments can be combined into one module, and can also be further split into multiple sub-modules.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that: it still Modifications or equivalent replacements can be made to the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims (15)

1. a video transcoding method is characterized in that, comprising:

Input code flow is decoded, extract the macro block information of decoded input code flow, determine pixel transcoding rate according to described input code flow;

When the described pixel transcoding rate of determining during, described decoded input code flow is encoded according to described macro block information less than preset first threshold;

When described pixel transcoding rate during, redefine macro-block coding pattern according to described macro block information, and described decoded input code flow is encoded according to the macro-block coding pattern that redefines greater than described first threshold value.

2. method according to claim 1 is characterized in that, describedly determines that according to described input code flow pixel transcoding rate comprises:

According to the bit number that average every pixel in average every pixel in the described input code flow bit number that takies and the output code flow of presetting takies, determine pixel transcoding rate.

3. method according to claim 1 is characterized in that, described described decoded input code flow the coding according to described macro block information comprises:

After according to described macro block information each frame of described decoded input code flow being encoded, first distortion factor of the output frame behind the calculation code, and second distortion factor of the incoming frame of described output frame correspondence, and calculate the ratio of described first distortion factor and described second distortion factor;

When described ratio during, described decoded input code flow is encoded according to described macro block information greater than the preset second threshold value;

When described ratio during less than described second threshold value, redefine macro-block coding pattern according to described macro block information, and according to the macro-block coding pattern that redefines follow-up code stream to be encoded is encoded, until end-of-encode to described decoded input code flow.

4. method according to claim 3 is characterized in that, first distortion factor of the output frame behind the described calculation code comprises:

Calculate first ratio of total bit number of the texture bit number of each macro block in the output frame behind the described coding and described macro block, and calculate the average of first ratio of all macro blocks in the output frame behind the described coding, the average of described first ratio is first distortion factor;

Second distortion factor of the incoming frame of the described output frame correspondence of described calculating comprises:

Calculate second ratio of total bit number of the texture bit number of each macro block in the incoming frame of described output frame correspondence and described macro block, and calculate the average of second ratio of all macro blocks in the described incoming frame, the average of described second ratio is described second distortion factor.

5. method according to claim 3 is characterized in that, described described decoded input code flow the coding according to described macro block information comprises:

Extract the block encoding Type C BP information in the described macro block information;

Described CBP information is designated as zero macro block directly carries out entropy coding.

6. method according to claim 1 is characterized in that, also comprises:

According to described macro block information, determine whether macro block to be encoded is the Skip piece;

When described macro block to be encoded is the Skip piece, directly next macro block of described Skip piece is encoded.

7. method according to claim 6 is characterized in that, the step that described definite macro block to be encoded is the Skip piece comprises:

M macro block in frame F is that (F m), is that (M (F, m)), decoding side present frame is F to RD through the optimum rate distortion costs that obtains after mode adjudging and the estimation to M _DC, decoding side present frame corresponding reference frame is F _DR, coding side present frame is F _EC, coding side present frame corresponding reference frame is F _ER,

The RD of described macro block to be encoded _Skip(M (F _EC, m)) satisfy

${\mathrm{RD}}_{\mathrm{Skip}}\left(M(F_{\mathrm{EC}},m)\right)<\mathrm{\&rho;}\&CenterDot;\mathrm{RD}\left(M(F_{\mathrm{ER}},m)\right)\&CenterDot;\frac{\mathrm{RD}\left(M(F_{\mathrm{DC}},m)\right)}{\mathrm{RD}\left(M(F_{\mathrm{DR}},m)\right)}$ The time,

Described macro block to be encoded is the Skip piece; Wherein, ρ is the positive number greater than 0, RDS _Kip(M (F _EC, m)) and be the optimum rate distortion costs of described macro block to be encoded,

RD _Skip(M(F _EC，m))＝D _REC(M(F _EC，m)，Skip)。

8. according to claim 1,3,5 or 6 described methods, it is characterized in that described macro block information comprises: block type information, piecemeal pattern information, motion vector information, CBP information, quantization parameter and macro block residual error;

Described macro-block coding pattern comprises: block type information, piecemeal pattern information and motion vector information.

9. a video code conversion device is characterized in that, comprising:

Decoder module is used for input code flow is decoded;

Extraction module is used to extract the macro block information of the decoded input code flow of described decoder module;

Pixel transcoding rate determination module is used for determining pixel transcoding rate according to described input code flow;

Coding module when being used for the pixel transcoding rate determined when described pixel transcoding rate determination module less than preset first threshold, is encoded to the decoded input code flow of described decoder module according to the macro block information that described extraction module extracts;

The coding mode determination module, when being used for the pixel transcoding rate determined when described pixel transcoding rate determination module greater than described first threshold value, the macro block information that extracts according to described extraction module redefines macro-block coding pattern; The macro-block coding pattern that described coding module also is used for redefining according to described coding mode determination module is encoded to the decoded input code flow of described decoder module.

10. video code conversion device according to claim 9, it is characterized in that, described pixel transcoding rate determination module specifically is used for the bit number that average every pixel takies in bit number that takies according to the average every pixel of described input code flow and the output code flow of presetting, and determines pixel transcoding rate.

11. video code conversion device according to claim 9 is characterized in that, described coding module comprises:

Distortion factor calculating sub module, be used for after the macro block information that extracts according to described extraction module is encoded to each frame of the decoded input code flow of described decoder module, first distortion factor of the output frame behind the calculation code, and second distortion factor of the incoming frame of described output frame correspondence, and calculate the ratio of described first distortion factor and described second distortion factor;

The first coding submodule when being used for the ratio that calculates when described distortion factor calculating sub module greater than the preset second threshold value, is encoded to the decoded input code flow of described decoder module according to the macro block information of described extraction module extraction.

12. video code conversion device according to claim 11, it is characterized in that, when described coding mode determination module also was used for the ratio that calculates when described distortion factor calculating sub module less than described second threshold value, the macro block information that extracts according to described extraction module redefined macro-block coding pattern;

Described coding module also comprises: the second coding submodule, the macro-block coding pattern that is used for redefining according to described coding mode determination module is encoded to follow-up code stream to be encoded, until the end-of-encode to described decoded input code flow.

13. according to the described video code conversion device of claim 11, it is characterized in that, described distortion factor calculating sub module is used for calculating first ratio of total bit number of the texture bit number of each macro block of output frame behind the described coding and described macro block, and calculate the average of first ratio of all macro blocks in the output frame behind the described coding, the average of described first ratio is first distortion factor;

Described distortion factor calculating sub module also is used for calculating second ratio of total bit number of the texture bit number of each macro block of incoming frame of described output frame correspondence and described macro block, and calculate the average of second ratio of all macro blocks in the described incoming frame, the average of described second ratio is described second distortion factor.

14. video code conversion device according to claim 11 is characterized in that,

Described extraction module also is used for extracting the block encoding Type C BP information of described macro block information;

The described first coding submodule is used for that described CBP information is designated as zero macro block and directly carries out entropy coding.

15. according to the described video code conversion device of claim 9, it is characterized in that, also comprise:

Skip piece determination module is used for the macro block information according to described extraction module extraction, determines whether macro block to be encoded is the Skip piece;

Described coding module also is used for directly next macro block of described Skip piece being encoded when described Skip piece determination module determines that described macro block to be encoded is the Skip piece;

Described Skip piece determination module determines that macro block to be encoded is that the step of Skip piece comprises:

M macro block in frame F is that (F m), is that (M (F, m)), decoding side present frame is F to RD through the optimum rate distortion costs that obtains after mode adjudging and the estimation to M _DC, decoding side present frame corresponding reference frame is F _DR, coding side present frame is F _EC, coding side present frame corresponding reference frame is F _ER,

The RD of described macro block to be encoded _Skip(M (F _EC, m)) satisfy

${\mathrm{RD}}_{\mathrm{Skip}}\left(M(F_{\mathrm{EC}},m)\right)<\mathrm{\&rho;}\&CenterDot;\mathrm{RD}\left(M(F_{\mathrm{ER}},m)\right)\&CenterDot;\frac{\mathrm{RD}\left(M(F_{\mathrm{DC}},m)\right)}{\mathrm{RD}\left(M(F_{\mathrm{DR}},m)\right)}$ The time,

Described Skip piece determination module determines that described macro block to be encoded is the Skip piece; Wherein, ρ is the positive number greater than 0, RD _Skip(M (F _EC, m)) and be the optimum rate distortion costs of described macro block to be encoded,

RD _Skip(M(F _EC，m))＝D _REC(M(F _EC，m)，Skip)。

Images