CN102595142A - Method and system for power grid holographic digital map video encoding - Google Patents

Abstract

The invention relates to a method and a system for power grid holographic digital map video encoding. The method comprises the steps as follows: a laying control module divides a power grid holographic digital map video into a base layer and a time domain enhancement layer image sequence which are respectively transmitted to the base layer and a time domain enhancement layer encoding module for decoding; a video decoding quality evaluation module evaluates the quality of the base layer and a time domain enhancement layer encoding reconstructed image, and transmits an evaluation index to a video decoding control module; a code rate real time statistics module counts the code rate of the current base layer and the time domain enhancement layer encoding code rate, as well as a total decoding code rate, and transmits that to the video decoding control module; the video decoding control module controls the laying control module, the base layer and a video coder in the time domain enhancement layer encoding module according to the time domain enhancement layer encoding code rate and the total decoding rate; according to the code rate, a video package transmission control module confirms the inserted time domain enhancement layer encoding code flow rate when transmitting base layer decoding code flow. The invention is widely suitable for transmitting high-quality video images under the condition of a certain bandwidth.

Description

Holographic numerical map method for video coding of a kind of electrical network and system

Technical field

The present invention relates to technical field of video coding, particularly about a kind of holographic numerical map method for video coding of electrical network and system that the holographic numerical map video of the electrical network that contains multiple thematic data and project data is encoded.

Background technology

Video coding and decoding technology is most important with transmission for video storage.H.261 the development of video encoding standard has been passed through, MPEG-1, MPEG-2, H.263, MPEG-4, H.264 wait several stages, compression efficiency constantly promotes.H.264 as video encoding standard of new generation, with H.263, MPEG-4 compares, video compression efficiency has promoted about one times.Existing video coding technique all adopts lossy compression to obtain higher compression ratios; Main process comprises prediction, conversion, quantification, rate-distortion optimization control and entropy coding; Its basic ideas are only the information that changes to be encoded, and abandon the unessential information of part.Wherein, rate-distortion optimization control has directly determined limiting under the bandwidth condition quality of coded image restoration and reconstruction.The holographic numerical map of electrical network comprises abundant literal, indicates line and texture information; If in the holographic numerical map process of remote transmission electrical network, directly adopt existing video coding technique; The loss in detail that will inevitably cause above-mentioned information; Particularly under the lower bandwidth condition, literal with indicate line and can become very fuzzy, sometimes even be difficult to offer a clear explanation.Therefore; Must be to the characteristics of the holographic numerical map of electrical network; Design is based on the method for video coding and the system of new rate-distortion optimization control strategy; Guarantee to keep image detail information well, make literal on the numerical map, indicate line etc. and after coding transmission, can clearly recover out limiting under the bandwidth condition.

Summary of the invention

To the problems referred to above, the purpose of this invention is to provide a kind of holographic numerical map method for video coding of electrical network and the system that can under the qualification bandwidth, transmit the holographic numerical map video of electrical network high-qualityly.

For realizing above-mentioned purpose; The present invention takes following technical scheme: the holographic numerical map method for video coding of a kind of electrical network; It is characterized in that; It may further comprise the steps: 1) system is set, and this system comprises hierarchical control module, one basic layer of video packets transmission control module that coding module, a time domain enhancement layer coding module, a video encoding quality evaluation module, a video coding control module, a code check real-time statistics module and an output are connected with network communicating system that an input is connected with video acquisition system; 2) the hierarchical control parameter that transmits according to the video coding control module; The hierarchical control module will be divided into the image sequence to be encoded of abandoning coded video sequences and passing to the video encoding quality evaluation module from the holographic numerical map video of the electrical network that video acquisition system receives earlier, again image sequence to be encoded will be divided into the basic tomographic image sequence of passing to basic layer coding module and the time domain enhancement layer image sequence of passing to time domain enhancement layer coding module; 3) the basic layer quality of coded picture grade adjustment parameter that transmits according to the video coding control module; Basic layer coding module is through built-in video encoder; Basic tomographic image sequence is encoded; The basic layer of output encoding code stream sends control module for code check real-time statistics module and video packets, and the basic layer of output coding and rebuilding image sequence is given video encoding quality evaluation module and time domain enhancement layer coding module; 4) the time domain enhancement layer coding image quality level adjustment parameter that transmits according to the video coding control module; Time domain enhancement layer coding module is based on basic layer coding and rebuilding image sequence; Through built-in video encoder; Time domain enhancement layer image sequence is encoded, and output time domain enhancement layer coding code stream sends control module for code check real-time statistics module and video packets, and output time domain enhancement layer coding reconstructed image sequence is given the video encoding quality evaluation module; 5) the video encoding quality evaluation module compares basic layer coding and rebuilding image sequence and time domain enhancement layer coding reconstructed image sequence respectively with image sequence to be encoded, obtains basic layer of quality of coded picture evaluation index d _{B_ssim}With time domain enhancement layer coding image quality evaluation index d _{Ts_ssim}, pass to the video coding control module; 6) code check real-time statistics module is carried out real-time statistics to the code check of basic layer encoding code stream and time domain enhancement layer coding code stream respectively, with the basic layer encoder bit rate r that obtains _bWith time domain enhancement layer coding code check r _Ts, and the code check r that always encodes _e, pass to video coding control module and video packets and send control module; 7) according to basic layer quality of coded picture evaluation index d _{B_ssim}, time domain enhancement layer coding image quality evaluation index d _{Ts_ssim}, basic layer encoder bit rate r _b, time domain enhancement layer coding code check r _TsWith total coding code check r _eThe video coding control module is calculated the hierarchical control parameter of passing to the hierarchical control module; Pass to the basic layer quality of coded picture grade adjustment parameter of basic layer coding module, pass to the time domain enhancement layer coding image quality level adjustment parameter of time domain enhancement layer coding module; 8) video packets is sent control module sending basic layer encoding code stream to network communicating system when; According to basic layer encoder bit rate, the time domain enhancement layer coding code check and the code check of always encoding, confirm the time domain enhancement layer coding code stream flow that in basic layer encoding code stream, inserts.

Above-mentioned steps 6) in, code check r always encodes _eBe basic layer encoder bit rate r _bWith time domain enhancement layer coding code check r _TsSum.

Above-mentioned steps 5) in, the quality of coded picture evaluation index of basic layer and time domain enhancement layer all adopts the SSIM index.

Above-mentioned steps 7) in, the hierarchical control parameter is basic frame rate f _b, time domain enhancement layer frame per second f _TsWith frame-skipping frame per second f _Skip, basic layer quality of coded picture grade adjustment parameter is basic layer coded frame level quantization parameter q _b, time domain enhancement layer coding image quality level adjustment parameter is time domain enhancement layer coding frame level quantization parameter q _TsEvery coding finishes a two field picture, and these parameter values are all upgraded through following algorithm iteration by the video coding control module:

1) judges r _c＞=r _eWhether set up:

If r _c＞=r _e, get into step 2);

If r _c＜r _e, get into step 9);

2) judge f _Ts=f _In-f _bWhether set up:

If f _Ts=f _In-f _b, get into step 3);

If f _Ts≠ f _In-f _b, get into step 5);

3) f _Ts, f _bAnd f _SkipValue remain unchanged, and judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, get into step 4);

4) judge d _{Ts_ssim}Whether＜=0.8 set up:

If d _{Ts_ssim}＜=0.8, promote time domain enhancement layer coding picture quality, q _Ts=Max (q _b+ 2, q _Ts-1), q _bValue remain unchanged, get into step 12);

If d _{Ts_ssim}＞0.8, q _bAnd q _TsValue remain unchanged, get into step 12);

5) judge f _TsWhether=0 set up:

If f _Ts=0, get into step 6);

If f _Ts≠ 0, get into step 7);

6) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, q _bValue remain unchanged q _Ts=Min (51, q _b+ 2),

Calculate basic frame rate f _bMedian f ' _b,

And pass through

Confirm f _Ts,

Pass through f again _b=Min (5, f ' _b) confirm f _bValue, and pass through f _Skip=f _In-(f _b+ f _Ts) confirm f _Skip, get into step 12);

7) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, get into step 8);

8) judge d _{Ts_ssim}Whether＜=0.8 set up:

If d _{Ts_ssim}＜=0.8, promote time domain enhancement layer coding picture quality, q _Ts=Max (q _b+ 2, q _Ts-1),

f _Ts, f _b, f _SkipAnd q _bValue remain unchanged, get into step 12);

If d _{Ts_ssim}＞0.8, f _b, q _bAnd q _TsValue remain unchanged, and

$f_{\mathrm{ts}}=\mathrm{Min}(f_{\mathrm{in}}-f_b,\frac{(r_c-r_e)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}}),$

f _skip＝f _in-(f _b+f _ts)，

Get into step 12);

9) judge r _b＜=r _cWhether set up:

If r _b＜=r _c, get into step 10);

If r _b＞r _c, get into step 11);

10) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, f _bAnd q _TsValue remain unchanged q _b=Max (1, q _b-1),

$f_{\mathrm{ts}}=\frac{(r_c-r_b)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}},$

f _skip＝f _in-(f _b+f _ts)，

Get into step 12);

If d _{B_ssim}＞0.85, reduce basic layer encoder bit rate, q _b=Min (51, q _b+ 2), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

11) reduce basic layer encoder bit rate, q _b=Min (51, q _b+ 2), and judge f _TsWhether=0 set up:

If f _Ts=0 sets up f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If f _Ts＞0, stop the time domain enhancement layer coding, f _Ts=0, f _Skip=f _In-f _b, f _bAnd q _TsValue remain unchanged, get into step 12);

12) finish.

Above-mentioned frame-skipping frame per second f _SkipInitial value f _Skip0Be set at 0, basic frame rate f _bInitial value f _B0With time domain enhancement layer frame per second f _TsInitial value f _Ts0Then confirm by following formula respectively:

$f_{b0}=\mathrm{Min}(5,\frac{7864.32*r_c}{W*H})$

$f_{\mathrm{ts}0}=\mathrm{Min}((f_{\mathrm{in}}-f_{b0}),\mathrm{Max}(0,\frac{7864.32*r_c-f_{b0}*W*H}{\mathrm{\α}*W*H}))$

In the following formula, current available network bandwidth r _cUnit be kilobits/second (kbps), W presentation code picture traverse, H presentation code picture altitude, f _InThe holographic numerical map video acquisition of expression electrical network input frame rate, α representes the encoder bit rate proportioning parameter of the basic relatively layer of time domain enhancement layer, default value is made as 0.6.

Above-mentioned steps 8) in, video packets is sent the number N of the time domain enhancement layer coding image that control module confirms through following formula to insert between every adjacent two basic layers of coded image _Ts:

$N_{\mathrm{ts}}=\mathrm{Max}(0,\frac{(r_c-r_b)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}*f_b})$

In the following formula, r _cBe current available network bandwidth, r _bBe basic layer encoder bit rate, r _TsBe time domain enhancement layer coding code check, f _bBe basic frame rate, f _TsBe time domain enhancement layer frame per second.

A kind of holographic numerical map video coding system of electrical network of realizing said method is characterized in that: it comprises hierarchical control module, one basic layer of video packets transmission control module that coding module, a time domain enhancement layer coding module, a video encoding quality evaluation module, a video coding control module, a code check real-time statistics module and an output are connected with network communicating system that an input is connected with video acquisition system; First output of hierarchical control module connects basic layer coding module, and second output connects time domain enhancement layer coding module, and the 3rd output connects the video encoding quality evaluation module; First output of basic layer coding module connects video encoding quality evaluation module and time domain enhancement layer coding module, and second output connects code check real-time statistics module and video packets is sent control module; First output of time domain enhancement layer coding module connects the video encoding quality evaluation module, and second output connects code check real-time statistics module and video packets is sent control module; The output of video encoding quality evaluation module connects the video coding control module; First output of video coding control module connects basic layer coding module, and second output connects time domain enhancement layer coding module, and the 3rd output connects the hierarchical control module; The output of code check real-time statistics module connects the video coding control module and video packets is sent control module.

Video encoder in above-mentioned basic layer coding module and the time domain enhancement layer coding module adopts H.264HighProfile video encoder.

The present invention is owing to take above technical scheme; It has the following advantages: 1, the present invention is owing to be divided into basic tomographic image sequence that is loaded with key message and the time domain enhancement layer image sequence that keeps the picture continuity with the holographic numerical map video of electrical network to be encoded; Realize optimization coding through the video coding control module to basic tomographic image sequence and time domain enhancement layer image sequence; Thereby guarantee video through after the coding transmission, can revert to the high-quality video image that comprises clear literal, indicates detailed information such as line and texture at the remote decoder end.2, the present invention is owing to adopt the code check of basic layer of code check real-time statistics module real-time statistics and time domain enhancement layer coding code stream; And corresponding total coding code check; And in view of the above cataloged procedure is controlled; Therefore have good network Bandwidth adaptation property, make encoder bit rate can adapt to the fluctuation of the network bandwidth well.3, the present invention removes code check real-time statistics module is set; Also be provided with the video encoding quality evaluation module; Obtain the quality of coded picture evaluation index of basic tomographic image coding module and time domain enhancement layer image coding module, thereby can make video in the balance of getting aspect image quality and the continuity through video coding control module adjustment respective coding code check and frame per second; When making full use of current available bandwidth, obtain high-quality as far as possible image.The present invention can be widely used in and limit high-quality ground transmitting video image under the bandwidth.

Description of drawings

Fig. 1 is that system of the present invention forms sketch map

Embodiment

Below in conjunction with accompanying drawing and embodiment the present invention is carried out detailed description.

As shown in Figure 1; Coding method of the present invention realizes through a system, and this system comprises that a hierarchical control module 1, one basic layer of coding module 2, a time domain enhancement layer coding module 3, a video encoding quality evaluation module 4, a video coding control module 5, a code check real-time statistics module 6 and a video packets send control module 7.

The hierarchical control parameter that hierarchical control module 1 of the present invention transmits according to video coding control module 5; Earlier the holographic numerical map video of the electrical network that receives is divided into and abandons coded video sequences and image sequence to be encoded, again image sequence to be encoded is divided into basic tomographic image sequence and time domain enhancement layer image sequence.Hierarchical control module 1 is passed to video encoding quality evaluation module 4 with image sequence to be encoded, and basic tomographic image sequence is passed to basic layer coding module 2, and time domain enhancement layer image sequence is passed to time domain enhancement layer coding module 3.Wherein:

Abandon coded video sequences, be meant under the extremely low situation of available transmission bandwidth,, cause the serious deterioration of decoded image quality, thereby before coding, initiatively abandon the image encoded sequence for avoiding causing packet loss because of the coding bit rate output surpasses available bandwidth; Basic tomographic image sequence can independently be carried out encoding and decoding, but frame per second is not high, can be used for guaranteeing under the lower bandwidth situation, obtains to guarantee effective transmission of key message than images with high image quality; Time domain enhancement layer image sequence need be carried out encoding and decoding based on basic tomographic image sequence, when more than needed bandwidth, can be used for improving video frame rate, makes picture more smooth.

As shown in Figure 1, hierarchical control module 1 has two inputs and three outputs: input 11 is connected (not shown) with video acquisition system, receives the holographic numerical map video of electrical network; Input 12 is connected with the output of video coding control module 5, receives the hierarchical control parameter; Output 13 is connected with the input of basic layer coding module 2, transmits basic tomographic image sequence; Output 14 is connected with the input of time domain enhancement layer coding module 3, transmission time domain enhancement layer image sequence; Output 15 is connected with the input of video encoding quality evaluation module 4, transmits image sequence to be encoded.

Wherein, hierarchical control parameter is the basic frame rate f that characterizes basic tomographic image sequence, time domain enhancement layer image sequence and abandon coded video sequences proportion in video _b, time domain enhancement layer frame per second f _TsWith frame-skipping frame per second f _SkipWhen starting coding, frame-skipping frame per second f _SkipInitial value f _Skip0Be set at 0, basic frame rate f _bInitial value f _B0With time domain enhancement layer frame per second f _TsInitial value f _Ts0Then confirm by following formula respectively:

$f_{b0}=\mathrm{Min}(5,\frac{7864.32*r_c}{W*H})$

$f_{\mathrm{ts}0}=\mathrm{Min}((f_{\mathrm{in}}-f_{b0}),\mathrm{Max}(0,\frac{7864.32*r_c-f_{b0}*W*H}{\mathrm{\α}*W*H}))$

In the following formula, r _cRepresent current available network bandwidth, unit is kilobits/second (kbps), W presentation code picture traverse, H presentation code picture altitude, f _InThe holographic numerical map video acquisition of expression electrical network input frame rate, α representes the encoder bit rate proportioning parameter of the basic relatively layer of time domain enhancement layer, default value is made as 0.6.First two field picture is basic tomographic image, and the time domain enhancement layer image is evenly distributed between the basic tomographic image, and the time domain enhancement layer image number between the two whenever adjacent basic tomographic images is identical.

Basic layer coding module 2 of the present invention is through built-in video encoder; Basic layer quality of coded picture grade according to video coding control module 5 transmits is adjusted parameter; Basic tomographic image sequence to receiving is encoded; The basic layer of output encoding code stream sends control module for code check real-time statistics module and video packets, and the basic layer of output coding and rebuilding image sequence is given video encoding quality evaluation module and time domain enhancement layer coding module.

As shown in Figure 1, basic layer coding module 2 has two inputs and two outputs: input 21 is connected with the output 13 of hierarchical control module 1, receives basic tomographic image sequence; Input 22 is connected with the output of video coding control module 5, receives basic layer quality of coded picture grade adjustment parameter; Output 23 is connected with the input of time domain enhancement layer coding module 3 and video coding evaluation module 4, the basic layer of transmission coding and rebuilding image sequence; Output 24 is connected with the input that code check real-time statistics module 6 and video packets are sent control module 7, the basic layer of transmission encoding code stream.

Wherein, basic layer coding module 2 built-in video encoders can adopt any video encoder, and current default employing is High Profile video encoder H.264.Basic layer quality of coded picture grade adjustment parameter is a basic frame level quantization parameter, and frame type comprises I frame and P frame.Basic layer coded frame level quantization parameter q _bInitial value can be set at 20, but be not limited thereto.In cataloged procedure; Video coding control module 5 can constantly be adjusted basic frame level quantization parameter according to the quality of coded picture evaluation index; Decoding and rebuilding quality to guarantee the holographic numerical map video of electrical network is not less than pre-determined grade, guarantees that literal, sign line and the texture information on the picture is clear.

The basic layer coding and rebuilding image sequence that time domain enhancement layer coding module 3 of the present invention transmits based on basic layer coding module 2; Time domain enhancement layer coding image quality level according to video coding control module 5 transmits is adjusted parameter; Through built-in video encoder the time domain enhancement layer image sequence that receives is encoded; Output time domain enhancement layer coding code stream sends control module 7 for code check real-time statistics module 6 and video packets, and output time domain enhancement layer coding reconstructed image sequence is given video encoding quality evaluation module 4.

As shown in Figure 1, time domain enhancement layer coding module 3 has three inputs and two outputs: input 31 is connected with the output 14 of hierarchical control module 1, receives time domain enhancement layer image sequence; Input 32 is connected with the output 23 of basic layer coding, receives basic layer coding and rebuilding image sequence; Input 33 is connected with the output of video coding control module 5, receives time domain enhancement layer coding image quality level adjustment parameter; Output 34 is connected with the input of video encoding quality evaluation module 4, transmission time domain enhancement layer coding reconstructed image sequence; Output 35 is connected with the input that code check real-time statistics module 6 and video packets are sent control module 7, transmission time domain enhancement layer coding code stream.

Wherein, time domain enhancement layer coding module 3 built-in video encoders can adopt any video encoder, and current default employing is High Profile video encoder H.264.Time domain enhancement layer coding image quality level adjustment parameter is a time domain enhancement layer frame level quantization parameter, and frame type comprises P frame and B frame.Time domain enhancement layer coding frame level quantization parameter q _TsInitial value can be set at 22.In cataloged procedure; Video coding control module 5 can constantly be adjusted time domain enhancement layer frame level quantization parameter according to the quality of coded picture evaluation index; Guarantee to have the better image quality based on the time domain enhancement-layer video coded video sequences of basic tomographic image sequence; When more than needed bandwidth, time domain enhancement-layer video coded video sequences can be used for improving video frame rate, makes picture more smooth.

Video encoding quality evaluation module 4 of the present invention is according to image sequence to be encoded; Respectively the picture quality of basic layer coding and rebuilding image sequence and time domain enhancement layer coding reconstructed image sequence is estimated, and the basic layer quality of coded picture evaluation index d that will obtain _{B_ssim}With time domain enhancement layer coding image quality evaluation index d _{Ts_ssim}Pass to the video coding control module.

As shown in Figure 1, video encoding quality evaluation module 4 has three inputs and an output: input 41 is connected with the output 15 of hierarchical control module 1, receives image sequence to be encoded; Input 42 is connected with the output 23 of basic layer coding module 2, receives basic layer coding and rebuilding image sequence; Input 43 is connected with the output 34 of time domain enhancement layer coding module 3, receives time domain enhancement layer coding reconstructed image sequence; Output 44 is connected with the input of video coding control module 5, the basic layer of transmission quality of coded picture evaluation index d _{B_ssim}With time domain enhancement layer coding image quality evaluation index d _{Ts_ssim}

Wherein, the quality of coded picture evaluation index of basic layer and time domain enhancement layer all can adopt the SSIM index, but also can adopt the class SSIM index that derives on this basis, like MS-SSIM, 3SSIM, stSSIM etc.Propose the Image quality assessment:From error visibility to structural similarity (image quality measure: visual to structural similarity) that the SSIM index is delivered on IEEE Trans.Image Processing (IEEE image processing magazine) in 2004 by Z.Wang etc. from mistake.

Code check real-time statistics module 6 of the present invention is carried out real-time statistics to the code check of basic layer encoding code stream and the code check of time domain enhancement layer coding code stream respectively, obtains current basic layer encoder bit rate r _bWith time domain enhancement layer coding code check r _Ts, and the code check r that always encodes _e, and pass to video coding control module 5 and send control module 7 with video packets.Wherein, the code check r that always encodes _eBe basic layer encoder bit rate r _bWith time domain enhancement layer coding code check r _TsSum.

As shown in Figure 1, code check real-time statistics module 6 has two inputs and an output: input 61 is connected with the output 24 of basic layer coding module 2, receives basic layer encoding code stream; Input 62 is connected with the output 35 of time domain enhancement layer coding module 3, receives time domain enhancement layer coding code stream; Output 63 is connected with the input that video coding control module 5 and video packets are sent control module 7, the basic layer of transmission encoder bit rate r _b, time domain enhancement layer coding code check r _TsWith total coding code check r _e

Video coding control module 5 of the present invention is according to the quality of coded picture evaluation index that receives; And basic layer encoder bit rate, the time domain enhancement layer coding code check and the code check of always encoding; 1. calculate the hierarchical control parameter, pass to hierarchical control module 1 and carry out the classification of successive image; 2. calculate basic layer quality of coded picture grade adjustment parameter; Pass to the interior Rate Control module of video encoder of basic layer coding module 2; Calculate time domain enhancement layer coding image quality level adjustment parameter; Pass to the interior Rate Control module of video encoder of time domain enhancement layer coding module 3,, improve quality of coded picture to make full use of current available bandwidth.

As shown in Figure 1, video coding control module 5 has two inputs and three outputs: input 51 is connected with the output 44 of video encoding quality evaluation module 4, received code image quality evaluation index; Input 52 is connected with the output 63 of code check real-time statistics module 6, receives basic layer encoder bit rate, the time domain enhancement layer coding code check and the code check of always encoding; Output 53 is connected with the input 22 of basic layer coding module 2, the basic layer of transmission quality of coded picture grade adjustment parameter; Output 54 is connected with the input 33 of time domain enhancement layer coding module 3, transmission time domain enhancement layer coding image quality level adjustment parameter; Output 55 is connected transmitting layered Control Parameter with the input 12 of hierarchical control module 1.

Wherein, behind the intact two field picture of every coding, video coding control module 5 is according to current available network bandwidth r _c, upgrade basic frame rate f as the hierarchical control parameter through following algorithm iteration _b, time domain enhancement layer frame per second f _TsWith frame-skipping frame per second f _Skip, as basic layer of coded frame level quantization parameter q of the quality of coded picture grade adjustment parameter of basic layer and time domain enhancement layer _bWith time domain enhancement layer coding frame level quantization parameter q _Ts:

1) judges r _c＞=r _eWhether set up:

If r _c＞=r _e, get into step 2);

If r _c＜r _e, get into step 9);

2) judge f _Ts=f _In-f _bWhether set up:

If f _Ts=f _In-f _b, get into step 3);

If f _Ts≠ f _In-f _b, get into step 5);

3) f _Ts, f _bAnd f _SkipValue remain unchanged, and judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, get into step 4);

4) judge d _{Ts_ssim}Whether＜=0.8 set up:

If d _{Ts_ssim}＜=0.8, promote time domain enhancement layer coding picture quality, q _Ts=Max (q _b+ 2, q _Ts-1), q _bValue remain unchanged, get into step 12);

If d _{Ts_ssim}＞0.8, q _bAnd q _TsValue remain unchanged, get into step 12);

5) judge f _TsWhether=0 set up:

If f _Ts=0, get into step 6);

If f _Ts≠ 0, get into step 7);

6) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, q _bValue remain unchanged q _Ts=Min (51, q _b+ 2),

Calculate basic frame rate f _bMedian f ' _b, And pass through

Confirm f _Ts, pass through f again _b=Min (5, f ' _b) confirm f _bValue, and pass through f _Skip=f _In-(f _b+ f _Ts) confirm f _Skip, get into step 12);

7) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, get into step 8);

8) judge d _{Ts_ssim}Whether＜=0.8 set up:

If d _{Ts_ssim}＜=0.8, promote time domain enhancement layer coding picture quality, q _Ts=Max (q _b+ 2, q _Ts-1), f _Ts, f _b, f _SkipAnd q _bValue remain unchanged, get into step 12);

If d _{Ts_ssim}＞0.8, f _b, q _bAnd q _TsValue remain unchanged, and

$f_{\mathrm{ts}}=\mathrm{Min}(f_{\mathrm{in}}-f_b,\frac{(r_c-r_e)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}}),$

f _skip＝f _in-(f _b+f _ts)，

Get into step 12);

9) judge r _b＜=r _cWhether set up:

If r _b＜=r _c, get into step 10);

If r _b＞r _c, get into step 11);

10) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, f _bAnd q _TsValue remain unchanged q _b=Max (1, q _b-1)

$f_{\mathrm{ts}}=\frac{(r_c-r_b)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}},$

f _skip＝f _in-(f _b+f _ts)，

Get into step 12);

If d _{B_ssim}＞0.85, reduce basic layer encoder bit rate, q _b=Min (51, q _b+ 2), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

11) reduce basic layer encoder bit rate, q _b=Min (51, q _b+ 2), and judge f _TsWhether=0 set up:

If f _Ts=0 sets up f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If f _Ts＞0, stop the time domain enhancement layer coding, f _Ts=0, f _Skip=f _In-f _b, f _bAnd q _TsValue remain unchanged, get into step 12);

12) finish.

Video packets of the present invention is sent control module sending basic layer encoding code stream to network communicating system when; According to basic layer encoder bit rate, the time domain enhancement layer coding code check and the code check of always encoding; Confirm the time domain enhancement layer coding code stream flow that in basic layer encoding code stream, inserts; To guarantee that under the prerequisite that does not break through the target bandwidth video coding code stream of conveying can revert to the image of better smoothness in decoding end.

As shown in Figure 1, video packets is sent control module 7 has three inputs and an output: input 71 is connected with the output 24 of basic layer coding module 2, receives basic layer encoding code stream; The output 35 of input 72 time domain enhancement layer coding modules 3 connects, and receives time domain enhancement layer coding code stream; Input 73 is connected with the output 63 of code check real-time statistics module 6, receives basic layer encoder bit rate, the time domain enhancement layer coding code check and the code check of always encoding; Output 74 is connected (not shown) with network communicating system, the output video encoding code stream.

Wherein, video packets is sent the number N of the time domain enhancement layer coding image that control module 7 confirms according to following formula to insert between every adjacent two basic layers of coded image _Ts:

$N_{\mathrm{ts}}=\mathrm{Max}(0,\frac{(r_c-r_b)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}*f_b})$

Above-mentioned each embodiment only is used to explain the present invention, and wherein the structure of each parts, connected mode etc. all can change to some extent, and every equivalents of on the basis of technical scheme of the present invention, carrying out and improvement all should not got rid of outside protection scope of the present invention.

Claims (10)

1. the holographic numerical map method for video coding of an electrical network is characterized in that it may further comprise the steps:

1) system is set, this system comprises hierarchical control module, one basic layer of video packets transmission control module that coding module, a time domain enhancement layer coding module, a video encoding quality evaluation module, a video coding control module, a code check real-time statistics module and an output are connected with network communicating system that an input is connected with video acquisition system;

2) the hierarchical control parameter that transmits based on the video coding control module; The hierarchical control module will be divided into the image sequence to be encoded of abandoning coded video sequences and passing to the video encoding quality evaluation module from the holographic numerical map video of the electrical network that video acquisition system receives earlier, image sequence to be encoded will be divided into the basic tomographic image sequence of passing to basic layer coding module and the time domain enhancement layer image sequence of passing to time domain enhancement layer coding module again;

3) the basic layer quality of coded picture grade adjustment parameter that transmits based on the video coding control module; Basic layer coding module is through built-in video encoder; Basic tomographic image sequence is encoded; The basic layer of output encoding code stream sends control module for code check real-time statistics module and video packets, and the basic layer of output coding and rebuilding image sequence is given video encoding quality evaluation module and time domain enhancement layer coding module;

4) the time domain enhancement layer coding image quality level adjustment parameter that transmits according to the video coding control module; Time domain enhancement layer coding module is based on basic layer coding and rebuilding image sequence; Through built-in video encoder; Time domain enhancement layer image sequence is encoded, and output time domain enhancement layer coding code stream sends control module for code check real-time statistics module and video packets, and output time domain enhancement layer coding reconstructed image sequence is given the video encoding quality evaluation module;

5) the video encoding quality evaluation module compares basic layer coding and rebuilding image sequence and time domain enhancement layer coding reconstructed image sequence respectively with image sequence to be encoded, obtains basic layer of quality of coded picture evaluation index d _{B_ssim}With time domain enhancement layer coding image quality evaluation index d _{Ts_ssim}, pass to the video coding control module;

6) code check real-time statistics module is carried out real-time statistics to the code check of basic layer encoding code stream and time domain enhancement layer coding code stream respectively, with the basic layer encoder bit rate r that obtains _bWith time domain enhancement layer coding code check r _Ts, and the code check r that always encodes _e, pass to video coding control module and video packets and send control module;

7) according to basic layer quality of coded picture evaluation index d _{B_ssim}, time domain enhancement layer coding image quality evaluation index d _{Ts_ssim}, basic layer encoder bit rate r _b, time domain enhancement layer coding code check r _TsWith total coding code check r _eThe video coding control module is calculated the hierarchical control parameter of passing to the hierarchical control module; Pass to the basic layer quality of coded picture grade adjustment parameter of basic layer coding module, pass to the time domain enhancement layer coding image quality level adjustment parameter of time domain enhancement layer coding module;

8) video packets is sent control module sending basic layer encoding code stream to network communicating system when; Based on basic layer encoder bit rate, the time domain enhancement layer coding code check and the code check of always encoding, confirm the time domain enhancement layer coding code stream flow that in basic layer encoding code stream, inserts.

2. the holographic numerical map method for video coding of a kind of electrical network as claimed in claim 1, it is characterized in that: in the said step 6), code check r always encodes _eBe basic layer encoder bit rate r _bWith time domain enhancement layer coding code check r _TsSum.

3. the holographic numerical map method for video coding of a kind of electrical network as claimed in claim 1 is characterized in that: in the said step 5), the quality of coded picture evaluation index of basic layer and time domain enhancement layer all adopts the SSIM index.

4. the holographic numerical map method for video coding of a kind of electrical network as claimed in claim 2 is characterized in that: in the said step 5), the quality of coded picture evaluation index of basic layer and time domain enhancement layer all adopts the SSIM index.

5. like claim 1 or the holographic numerical map method for video coding of 2 or 3 or 4 described a kind of electrical networks, it is characterized in that: in the said step 7), the hierarchical control parameter is basic frame rate f _b, time domain enhancement layer frame per second f _TsWith frame-skipping frame per second f _Skip, basic layer quality of coded picture grade adjustment parameter is basic layer coded frame level quantization parameter q _b, time domain enhancement layer coding image quality level adjustment parameter is time domain enhancement layer coding frame level quantization parameter q _TsEvery coding finishes a two field picture, and these parameter values are all upgraded through following algorithm iteration by the video coding control module:

1) judges r _c＞=r _eWhether set up:

If r _c＞=r _e, get into step 2);

If r _c＜r _e, get into step 9);

2) judge f _Ts=f _In-f _bWhether set up:

If f _Ts=f _In-f _b, get into step 3);

If f _Ts≠ f _In-f _b, get into step 5);

3) f _Ts, f _bAnd f _SkipValue remain unchanged, and judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, get into step 4);

4) judge d _{Ts_ssim}Whether＜=0.8 set up:

If d _{Ts_ssim}＜=0.8, promote time domain enhancement layer coding picture quality, q _Ts=Max (q _b+ 2, q _Ts-1), q _bValue remain unchanged, get into step 12);

If d _{Ts_ssim}＞0.8, q _bAnd q _TsValue remain unchanged, get into step 12);

5) judge f _TsWhether=0 set up:

If f _Ts=0, get into step 6);

If f _Ts≠ 0, get into step 7);

6) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, q _bValue remain unchanged q _Ts=Min (51, q _b+ 2),

Calculate basic frame rate f _bMedian f ' _b,

And pass through Confirm f _Ts, pass through f again _b=Min (5, f ' _b) confirm f _bValue, and pass through f _Skip=f _In-(f _b+ f _Ts) confirm f _Skip, get into step 12);

7) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, q _b=Max (1, q _b-1), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If d _{B_ssim}＞0.85, get into step 8);

8) judge d _{Ts_ssim}Whether＜=0.8 set up:

If d _{Ts_ssim}＜=0.8, promote time domain enhancement layer coding picture quality, q _Ts=Max (q _b+ 2, q _Ts-1), f _Ts, f _b, f _SkipAnd q _bValue remain unchanged, get into step 12);

If d _{Ts_ssim}＞0.8, f _b, q _bAnd q _TsValue remain unchanged, and

$f_{\mathrm{ts}}=\mathrm{Min}(f_{\mathrm{in}}-f_b,\frac{(r_c-r_e)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}}),$

f _skip＝f _in-(f _b+f _ts)，

Get into step 12);

9) judge r _b＜=r _cWhether set up:

If r _b＜=r _c, get into step 10);

If r _b＞r _c, get into step 11);

10) judge d _{B_ssim}Whether＜=0.85 set up:

If d _{B_ssim}＜=0.85, promote basic layer quality of coded picture, f _bAnd q _TsValue remain unchanged q _b=Max (1, q _b-1),

$f_{\mathrm{ts}}=\frac{(r_c-r_b)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}},$

f _skip＝f _in-(f _b+f _ts)，

Get into step 12);

If d _{B_ssim}＞0.85, reduce basic layer encoder bit rate, q _b=Min (51, q _b+ 2), f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

11) reduce basic layer encoder bit rate, q _b=Min (51, q _b+ 2), and judge f _TsWhether=0 set up:

If f _Ts=0 sets up f _Ts, f _b, f _SkipAnd q _TsValue remain unchanged, get into step 12);

If f _Ts＞0, stop the time domain enhancement layer coding, f _Ts=0, f _Skip=f _In-f _b, f _bAnd q _TsValue remain unchanged, get into step 12);

12) finish.

6. the holographic numerical map method for video coding of a kind of electrical network as claimed in claim 5 is characterized in that: said frame-skipping frame per second f _SkipInitial value f _Skip0Be set at 0, basic frame rate f _bInitial value f _B0With time domain enhancement layer frame per second f _TsInitial value f _Ts0Then confirm by following formula respectively:

$f_{b0}=\mathrm{Min}(5,\frac{7864.32*r_c}{W*H})$

$f_{\mathrm{ts}0}=\mathrm{Min}((f_{\mathrm{in}}-f_{b0}),\mathrm{Max}(0,\frac{7864.32*r_c-f_{b0}*W*H}{\mathrm{\α}*W*H}))$

In the following formula, current available network bandwidth r _cUnit be kilobits/second (kbps), W presentation code picture traverse, H presentation code picture altitude, f _InThe holographic numerical map video acquisition of expression electrical network input frame rate, α representes the encoder bit rate proportioning parameter of the basic relatively layer of time domain enhancement layer, default value is made as 0.6.

7. like claim 1 or the holographic numerical map method for video coding of 2 or 3 or 4 or 6 described a kind of electrical networks; It is characterized in that: in the said step 8), video packets is sent the number N of the time domain enhancement layer coding image that control module confirms through following formula to insert between every adjacent two basic layers of coded image _Ts:

$N_{\mathrm{ts}}=\mathrm{Max}(0,\frac{(r_c-r_b)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}*f_b})$

In the following formula, r _cBe current available network bandwidth, r _bBe basic layer encoder bit rate, r _TsBe time domain enhancement layer coding code check, f _bBe basic frame rate, f _TsBe time domain enhancement layer frame per second.

8. the holographic numerical map method for video coding of a kind of electrical network as claimed in claim 5; It is characterized in that: in the said step 8), video packets is sent the number N of the time domain enhancement layer coding image that control module confirms through following formula to insert between every adjacent two basic layers of coded image _Ts:

$N_{\mathrm{ts}}=\mathrm{Max}(0,\frac{(r_c-r_b)*f_{\mathrm{ts}}}{r_{\mathrm{ts}}*f_b})$

In the following formula, r _cBe current available network bandwidth, r _bBe basic layer encoder bit rate, r _TsBe time domain enhancement layer coding code check, f _bBe basic frame rate, f _TsBe time domain enhancement layer frame per second.

9. the holographic numerical map video coding system of the electrical network of a realization such as each said method of claim 1～8, it is characterized in that: it comprises hierarchical control module, one basic layer of video packets transmission control module that coding module, a time domain enhancement layer coding module, a video encoding quality evaluation module, a video coding control module, a code check real-time statistics module and an output are connected with network communicating system that an input is connected with video acquisition system; First output of said hierarchical control module connects said basic layer coding module, and second output connects said time domain enhancement layer coding module, and the 3rd output connects said video encoding quality evaluation module; First output of said basic layer coding module connects said video encoding quality evaluation module and said time domain enhancement layer coding module, and second output connects said code check real-time statistics module and said video packets is sent control module; First output of said time domain enhancement layer coding module connects said video encoding quality evaluation module, and second output connects said code check real-time statistics module and said video packets is sent control module; The output of said video encoding quality evaluation module connects said video coding control module; First output of said video coding control module connects said basic layer coding module, and second output connects said time domain enhancement layer coding module, and the 3rd output connects said hierarchical control module; The output of said code check real-time statistics module connects said video coding control module and said video packets is sent control module.

10. the holographic numerical map video coding system of a kind of electrical network as claimed in claim 9, it is characterized in that: the video encoder in said basic layer coding module and the said time domain enhancement layer coding module all adopts H.264High Profile video encoder.