Transmission optimization method based on conference audiovisual

Abstract

The invention discloses a transmission optimization method based on conference audiovisual, relates to the technical field of picture transmission, and solves the problems that static and dynamic characteristics of a conference picture are not accurately identified, and adaptive adjustment of a coding rate is realized in combination with network bandwidth fluctuation. The stable transmission of the static picture ensures the clear presentation of contents such as document demonstration and blackboard writing, and the adaptive coding of the dynamic picture ensures the fluency and detail integrity of scenes such as spokesman action and multi-person interaction; through the combination of the two methods, unnecessary bandwidth consumption is reduced, information distortion caused by a transmission problem is avoided, the core requirements of low delay, high stability and excellent image quality in the conference process are finally met, and the immersion and communication efficiency of the remote conference are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of picture transmission, in particular to a transmission optimization method based on conference audio and video. BACKGROUND

[0002] In real-time audio and video interaction scenarios such as remote conferences and online collaboration, stable transmission and quality assurance of audio and video are the core requirements to improve communication efficiency. However, in the process of conference audio and video transmission, the existing technology often faces the following problems: on the one hand, the frame picture state recognition accuracy is insufficient.

[0003] Traditional methods often judge whether the picture is dynamic through a single dimension (such as simple frame difference comparison), which is easily disturbed by noise and light changes, resulting in static pictures (such as PPT demonstrations and fixed backgrounds) being misjudged as dynamic, or dynamic pictures (such as slowly moving figures and local area motion) being missed.

[0004] Such misjudgment will directly lead to mismatched encoding strategies - over-allocating bandwidth for static pictures (wasting resources) and insufficient encoding for dynamic pictures (causing blurred picture quality and loss of details).

[0005] On the other hand, the encoding rate adjustment lacks dynamic adaptation capability. In existing solutions, the encoding rate is often a fixed value or is only adjusted roughly based on bandwidth, without combining the dynamic characteristics of picture content (such as motion intensity and change range) with the bandwidth fluctuation rules (such as short-term bandwidth sudden rise and fall and long-term trend change) for collaborative optimization.

[0006] When the network bandwidth fluctuates, a fixed encoding rate can easily lead to two extremes: the encoding rate is too high to exceed the bandwidth carrying capacity, resulting in packet loss and lag; or the encoding rate is too low to meet the transmission needs of dynamic pictures, causing picture tearing and motion blur, which seriously affects the conference interaction experience. In addition, there is a lack of differentiated processing mechanism for static and dynamic pictures. Static pictures (such as documents and charts) have lower real-time requirements but need to ensure clear content; dynamic pictures (such as speaker actions and multi-person interactions) require higher fluency and detail integrity and need to match higher encoding resources. The existing technology does not develop fine encoding strategies for the characteristics of the two types of pictures, resulting in unreasonable allocation of bandwidth resources and difficulty in balancing transmission efficiency and audio and video quality.

[0007] Therefore, how to accurately identify the static and dynamic characteristics of conference pictures and achieve adaptive adjustment of the encoding rate in combination with network bandwidth fluctuations has become a key problem to improve the stability and quality of conference audio and video transmission. SUMMARY

[0008] To overcome the deficiencies of the prior art, the present application provides a transmission optimization method based on conference audio and video, which solves the problem of not accurately identifying the static and dynamic characteristics of conference pictures and achieving adaptive adjustment of the encoding rate in combination with network bandwidth fluctuations.

[0009] To achieve the above object, the present application is realized by the following technical scheme: the transmission optimization method based on conference audio and video, comprising the following steps:

[0010] Step one, confirm the adjacent frame pictures in the real-time interactive picture, and carry out the gray processing synchronously, confirm whether there is a change point or a dynamic area in the current frame picture from the processing result, based on the confirmation characteristics, mark the current frame picture as a static frame picture or a dynamic frame picture;

[0011] Preferably, in the step one, the specific way of carrying out the gray processing on the adjacent frame pictures is:

[0012] Confirm the frame picture associated with the current time as a standard frame picture, and record the two groups of frame pictures associated with the front end time of the standard frame picture as the frame pictures to be verified of the standard frame picture, and sort the frame pictures to be verified and the standard frame picture according to the time sequence to confirm the picture sorting set;

[0013] Carry out the gray processing on the three groups of frame pictures in the picture sorting set: confirm the RGB values associated with different pixel points in a single group of frame pictures, and adopt: 0.299×R+0.587×G+0.114×B=gray value to obtain the gray value associated with the corresponding pixel points, and carry out the gray processing on the associated frame picture based on the confirmed gray value, confirm the gray image associated with the corresponding frame picture, and confirm the gray image associated with the corresponding frame picture synchronously.

[0014] Preferably, in the step one, the specific way of confirming whether there is a change point in the current frame picture is:

[0015] Confirm the gray image associated with the standard frame picture based on the picture sorting set after the gray processing;

[0016] Confirm the different gray values associated with the same position pixel points from the adjacent gray images, carry out the difference processing on the two groups of gray values, confirm the gray difference value, and the gray difference value≥0, and evaluate whether the gray difference value satisfies: gray difference value≥Y1, if yes, mark the corresponding pixel point as a change point, if not, do not mark it;

[0017] Confirm the change points in the subsequent two groups of adjacent gray images in turn, and mark the change points at the same position as the same points from the confirmation process, then confirm the total number Zz of the change points at different positions from the three groups of gray images, and mark the total number of the pixel points in the standard frame picture as Zs, if (Zz÷Zs)≥0.3, mark the standard frame picture as a dynamic frame picture, otherwise, mark the standard frame picture as a static frame picture.

[0018] Preferably, the specific way of confirming whether the current frame picture has dynamic area in step one is as follows:

[0019] Confirming the gradient of different gray images in the picture sequence set, confirming the gray value associated with the corresponding pixel point from the corresponding gray image, and confirming the vertical gradient and vertical gradient associated with the specified gray point based on the Sobel algorithm, according to the confirmed vertical gradient and vertical gradient, confirming the comprehensive gradient associated with the corresponding pixel point, wherein: The pixel point satisfying the comprehensive gradient≥Y1 is recorded as a gradient point, otherwise, no marking is performed, wherein Y1 is a preset value;

[0020] Confirming a plurality of gradient points associated with a single set of gray images in sequence, connecting adjacent gradient points, confirming gradient lines, recording the confirmed gradient lines as gradient contours, recording the area contained by the gradient contours as gradient areas, and confirming the center point of the corresponding gradient area, confirming the gradient area existing in each gray image, and confirming whether there is a spatially overlapping gradient area from adjacent gray images, if there is, it is recorded as a similar dynamic area, if not, no marking is performed;

[0021] Confirming the similar dynamic area from three sets of gray images, and confirming the change vector of the center point of the corresponding gradient area in the three sets of gray images from the center point of the corresponding gradient area in the first gray image, and confirming the total number of pixel points XH passed by the change vector, if XH≥50, the standard frame picture is recorded as a dynamic frame picture, otherwise, the standard frame picture is marked as a static frame picture.

[0022] Step two, confirming the status of the current frame picture, if the current frame picture is a static frame picture, limiting the encoding rate in the transmission process of the current frame picture, if the current frame picture is a dynamic frame picture, confirming the best encoding rate in the transmission process according to the fluctuation characteristics of the bandwidth, completing the transmission process of the current frame picture;

[0023] Preferably, in step two, the execution encoding rate associated with the static frame picture is the preset encoding rate, and the encoding rate is adjusted in the transmission process of the static frame picture, so that the adjusted encoding rate is the execution encoding rate, and the transmission process of the current frame picture is completed.

[0024] Preferably, in step two, the way of confirming the bandwidth fluctuation characteristics of the dynamic frame picture is as follows:

[0025] Take the bandwidth at the current time as the reference bandwidth, and trace back the time from the current time, and the traced back unit time is not less than ten groups, confirm the historical bandwidth associated with the corresponding time from the running data of the traced back unit time, and confirm the change bandwidth associated between adjacent times according to the change process of the bandwidth data, the change bandwidth = | the change bandwidth of the next time - the change bandwidth of the previous time |, and select the minimum value and the maximum value from the confirmed several groups of change bandwidths to generate the change interval of the corresponding change bandwidth, and take the generated change interval as the fluctuation characteristic of the corresponding bandwidth.

[0026] Preferably, in step two, the specific way of confirming the optimal encoding rate in the transmission process of the dynamic frame is:

[0027] According to the confirmed fluctuation characteristic and the bandwidth D1 associated with the current time, confirm the associated bandwidth interval of the next time, and the fluctuation characteristic is [DBmin, DBmax], the associated bandwidth interval is [D1+DBmin, D1+DBmax], confirm the maximum encoding rate BM1 associated with the bandwidth (D1+DBmax) from the historical processing process data, and simultaneously confirm the maximum encoding rate BM2 associated with the bandwidth (D1+DBmin);

[0028] Confirm the dynamic characteristic from the dynamic frame, take the total number of change points confirmed in the dynamic frame as Zz, and use: DZ=ZzxC1 to confirm the dynamic characteristic associated with the dynamic frame, or confirm the total number of pixel points XH passed by the change vector in the dynamic frame, and use: Dz=XHxC2 to confirm the dynamic characteristic associated with the dynamic frame, wherein C1 and C2 are both preset fixed coefficient factors;

[0029] Confirm the encoding rate BM associated with the dynamic characteristic from the feature comparison table, and the feature comparison table is a preset table to identify whether the confirmed encoding rate BM is within the numerical range associated with BM1 and BM2:

[0030] If yes, directly use BM as the execution encoding rate, and execute the transmission process of the dynamic frame;

[0031] If no, determine the difference between BM and the numerical range associated with BM1 and BM2, if the difference exceeds the numerical range, use the maximum value of the numerical range as the execution encoding rate, and execute the transmission process of the dynamic frame, if the difference does not exceed the numerical range, use the minimum value of the numerical range as the execution encoding rate, and execute the transmission process of the dynamic frame.

[0032] The application provides a transmission optimization method based on conference audio and video. Compared with the prior art, the following beneficial effects are achieved:

[0033] The application determines the frame picture state from the pixel level change to the region level motion in two dimensions through gray processing, change point proportion analysis and dynamic region tracking, avoiding the misjudgment problem of single dimension judgment; the accurate distinction ensures that the subsequent encoding strategy can "treat the disease with the right medicine", and matches the most suitable processing mode for different types of pictures;

[0034] For static frame pictures, a preset encoding rate is used for stable transmission, which can meet the demand through lower bandwidth and reduce bandwidth resource waste; for dynamic frame pictures, the best encoding rate is determined by combining the bandwidth fluctuation characteristics and the picture dynamic characteristics, which avoids packet loss and lag caused by too high encoding rate exceeding the bandwidth capacity, and prevents dynamic picture blur caused by too low encoding rate, achieving the balance between "maximum bandwidth utilization" and "optimal picture quality experience".

[0035] The adaptive mechanism enhances the stability and anti-fluctuation ability of the conference transmission; the method generates a fluctuation interval by tracing back the historical bandwidth data, predicts the bandwidth range in advance, matches the encoding rate from the feature comparison table combined with the real-time characteristics of the dynamic picture, and finally restricts the encoding rate through the bandwidth interval, so that the encoding strategy can dynamically respond to network fluctuations and picture content changes; this "prediction + adaptation" mechanism effectively reduces the impact of network sudden fluctuations on conference experience, and is especially suitable for remote conference scenarios in complex network environments.

[0036] The overall strategy significantly improves the user experience of conference audio-visual interaction; the stable transmission of static pictures ensures the clear presentation of content such as document demonstration and board writing, and the adaptive encoding of dynamic pictures ensures the fluency and detail integrity of scenarios such as speaker actions and multi-person interaction; the combination of the two reduces unnecessary bandwidth consumption and avoids information distortion caused by transmission problems, ultimately achieving the core needs of "low latency, high stability, and optimal picture quality" in the conference process, and improving the immersion and communication efficiency of remote conferences. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 The figure is a flowchart of the method of the application. DETAILED DESCRIPTION

[0038] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0039] First embodiment

[0040] Please refer to Figure 1The application provides a conference audio and video transmission optimization method, comprising the following steps:

[0041] Step one, confirm adjacent frame pictures in real-time interactive pictures, and perform grayscale processing synchronously, confirm whether there are change points or dynamic areas in the current frame picture from the verification result, based on the confirmed features, mark the current frame picture as a static frame picture or a dynamic frame picture, specifically, in the corresponding video interactive picture, there are associated continuous frame pictures, and there are dynamic change points in the continuous frame pictures, when the corresponding dynamic change points exceed a certain proportion or there are dynamically moving objects in the corresponding frame picture, it means that the corresponding frame picture belongs to a dynamic frame picture, so it is necessary to effectively adjust the corresponding coding rate in the subsequent coding process to effectively transmit the corresponding dynamic frame picture, if it belongs to a static frame picture, it can be associated with the adjustment according to the specific coding rate to ensure the transmission stability in the actual interactive process of the frame picture;

[0042] The specific way of grayscale processing of adjacent frame pictures is as follows:

[0043] Confirm the frame picture associated with the current time as a standard frame picture, and record the two groups of frame pictures associated with the front end time of the standard frame picture as the frame pictures to be verified of the standard frame picture, and sort the frame pictures to be verified and the standard frame picture according to the time sequence to confirm the picture sorting set;

[0044] Grayscale processing of three groups of frame pictures in the picture sorting set (based on the RGB values associated with the corresponding pixel points, and giving different weights to the corresponding RGB three groups of values, the corresponding gray value can be confirmed, and the grayscale picture associated with the corresponding picture can be confirmed): confirm the RGB values associated with different pixel points in a single group of frame pictures, and use: 0.299×R+0.587×G+0.114×B=gray value to obtain the gray value associated with the corresponding pixel points, and based on the confirmed gray value, the grayscale processing of the associated frame picture is performed, the grayscale image associated with the corresponding frame picture is confirmed, and the grayscale image associated with the corresponding frame picture is confirmed synchronously;

[0045] The specific way of confirming whether there are change points in the current frame picture is as follows:

[0046] Confirm the grayscale image associated with the standard frame picture based on the picture sorting set after grayscale processing;

[0047] Confirm the different gray values associated with the same position (pixel points are sorted in the form of a nine-square grid, so the corresponding pixel points in the same position can be effectively confirmed in the adjacent images) pixel points in the adjacent gray images, and perform difference processing on the two sets of gray values to confirm the gray difference value, and the gray difference value ≥ 0 (that is, after difference processing, absolute value processing is required to ensure that the gray difference value is positive), and whether the gray difference value satisfies: gray difference value ≥ Y1, if it satisfies, the corresponding pixel point is marked as a change point, if it does not satisfy, it does not perform any marking;

[0048] Confirm the change points in the subsequent two sets of adjacent gray images in turn, and mark the same point position from the confirmation process. Then, confirm the total number of different position change points Zz from the three sets of gray images, and mark the total number of pixel points in the standard frame picture as Zs. If (Zz ÷ Zs) ≥ 0.3, the standard frame picture is marked as a dynamic frame picture, otherwise, the standard frame picture is marked as a static frame picture.

[0049] Specifically, in adjacent gray images, there are pixel points in the same position, so that the gray value between the pixel points will change with the change of the frame time. In the specific change process, the point position where the gray value changes can be confirmed from the continuous frames, and according to the specific proportion of the corresponding change point, whether the corresponding frame belongs to a dynamic frame can be confirmed, and the dynamic frame can be specifically verified to complete the specific determination process of the frame.

[0050] The specific way to confirm whether the current frame has a dynamic area is:

[0051] Confirm the gradient of different gray images in the picture sorting set, confirm the gray value associated with the corresponding pixel point from the corresponding gray image, and confirm the vertical gradient and vertical gradient associated with the specified gray point based on the Sobel algorithm. According to the confirmed vertical gradient and vertical gradient (pixel points are sorted in the form of a nine-square grid, and different weight factors are assigned to the nine groups of pixel points during the sorting process. Then, according to the weight factor and the gray value associated with the corresponding pixel point, convolution sum processing is performed to confirm the vertical gradient and vertical gradient associated with the corresponding pixel point. The Sobel algorithm is commonly used in the prior art to confirm the gradient value, so it is not described in detail here), confirm the comprehensive gradient associated with the corresponding pixel point, wherein: Mark the pixel points that satisfy: comprehensive gradient ≥ Y1 as gradient points, otherwise, do not perform any marking, wherein Y1 is a preset value, and its specific value is determined by the operator according to experience;

[0052] Confirming a plurality of gradient points associated with a single group of gray images in sequence, connecting adjacent gradient points, confirming gradient lines, recording the confirmed gradient lines as gradient contours, recording the area contained by the gradient contours as gradient areas, and confirming the center point of the corresponding gradient area (combining the gradient contours and the two-dimensional coordinate system, the two-dimensional coordinates associated with the contour points can be confirmed, and the mean value coordinates can be obtained by averaging a plurality of two-dimensional coordinates, thereby confirming the center point associated with the corresponding gradient area), confirming the gradient areas existing in each gray image, and confirming whether there are gradient areas with spatial position intersection from adjacent gray images, if there are, it is recorded as the same type of dynamic area, if not, no calibration is performed;

[0053] Confirming the same type of dynamic area from three groups of gray images, and confirming the change vector of the center point of the corresponding gradient area in the three groups of gray images from the center point of the corresponding gradient area in the first gray image, and confirming the total number of pixel points XH passed by the change vector, if XH≥50, the standard frame picture is recorded as a dynamic frame picture, otherwise, the standard frame picture is calibrated as a static frame picture;

[0054] Specifically, if the same gradient area exists in three groups of gray images A, B and C, then the center point of the corresponding gradient area in gray image A is A1, the center point of the corresponding gradient area in gray image B is B1, and the center point of the corresponding gradient area in gray image C is C1, then the change process associated with the center points is A1-B1-C1, that is, the change process corresponding to the change connection, according to this change process, the movement process of the corresponding dynamic area can be confirmed, thereby the calibration process of the corresponding dynamic frame picture can be effectively completed;

[0055] Step two, confirming the status of the current frame picture, if the current frame picture is a static frame picture, limiting the encoding rate in the transmission process of the current frame picture, if the current frame picture is a dynamic frame picture, confirming the best encoding rate in the transmission process according to the fluctuation characteristics of the bandwidth, completing the transmission process of the current frame picture;

[0056] Among them, the specific sub-steps for detailed confirmation of the current frame picture encoding rate are:

[0057] If the current frame picture is a static frame picture, the preset encoding rate is used as the execution encoding rate, the encoding rate is adjusted in the transmission process of the current frame picture, the adjusted encoding rate is the execution encoding rate, and the transmission process of the current frame picture is completed;

[0058] If the current frame is a dynamic frame, the bandwidth fluctuation characteristics are determined from the historical bandwidth operation data. The bandwidth at the current time is taken as the reference bandwidth, and the time is traced back from the current time. The unit time for tracing back is not less than ten groups. The historical bandwidth associated with the corresponding time is determined from the operation data of the traced back unit time. The change bandwidth between adjacent times is determined according to the change process of the bandwidth data. The change bandwidth = | the change bandwidth of the next time - the change bandwidth of the previous time |. The minimum value and the maximum value are selected from the determined several groups of change bandwidths to generate the change interval of the corresponding change bandwidth. The generated change interval is taken as the fluctuation characteristics of the corresponding bandwidth.

[0059] According to the determined fluctuation characteristics and the bandwidth D1 associated with the current time, the associated bandwidth interval of the next time (that is, the transmission time of the dynamic frame) is determined. The fluctuation characteristics are [DBmin, DBmax], the associated bandwidth interval is [D1+DBmin, D1+DBmax], the maximum coding rate BM1 associated with the bandwidth (D1+DBmax) is determined from the historical processing data, and the maximum coding rate BM2 associated with the bandwidth (D1+DBmin) is determined synchronously.

[0060] The dynamic characteristics are determined from the dynamic frame. The total number of change points in the dynamic frame is taken as Zz. The dynamic characteristics associated with the dynamic frame are determined by using DZ=ZzxC1 (the more the total number of change points, the greater the associated dynamic characteristics). Or the total number of pixel points XH passed by the change vector in the dynamic frame is determined, and the dynamic characteristics associated with the dynamic frame are determined by using Dz=XHxC2. C1 and C2 are both preset fixed coefficient factors, and their specific values are determined by the operator according to experience.

[0061] The coding rate BM associated with the dynamic characteristics is determined from the feature comparison table. The feature comparison table is a preset table, which is determined in advance by relevant personnel. Different coding rates associated with different dynamic characteristics are internally preset. In the feature comparison table, there is a numerical interval associated with the corresponding dynamic characteristics. When the corresponding dynamic characteristics are in the numerical interval, the coding rate associated with the corresponding numerical interval is selected. It is determined whether the determined coding rate BM is in the numerical range associated with BM1 and BM2:

[0062] If yes, BM is directly selected as the execution coding rate, and the transmission process of the dynamic frame is executed.

[0063] If not, the difference between BM and the numerical range associated with BM1 and BM2 is determined. If it exceeds the numerical range, the maximum value of the numerical range is selected as the execution coding rate, and the transmission process of the dynamic frame is executed. If it does not exceed the numerical range, the minimum value of the numerical range is selected as the execution coding rate, and the transmission process of the dynamic frame is executed.

[0064] Specifically, in the actual encoding process, in order to ensure the optimal transmission process of the corresponding dynamic frame, it is necessary to confirm the optimal encoding rate associated with the corresponding process, so according to the dynamic characteristics associated with the corresponding dynamic frame, the encoding characteristics associated with the corresponding dynamic characteristics are quickly locked, and the most suitable encoding rate is confirmed from the associated encoding characteristics for execution, which not only effectively guarantees the transmission process of the picture, but also effectively avoids the packet loss behavior (excessive encoding rate and low bandwidth, which is easy to cause packet loss).

[0065] Some data in the above formula are dimensionless numerical calculations, and the contents not described in detail in the specification all belong to the prior art known to those skilled in the art.

[0066] The above examples are only used to illustrate the technical method of the present application and are not limited. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. A method for optimizing transmission of audiovisual conferencing, characterized in that, The method comprises the following steps: Step one, confirm the adjacent frame pictures in real-time interactive pictures, and perform gray processing synchronously, confirm whether there is a change point or dynamic area in the current frame picture from the processing result, and label the current frame picture as a static frame picture or a dynamic frame picture based on the confirmed features; Step two, confirm the state to which the current frame picture belongs, if the current frame picture is a static frame picture, limit the encoding rate in the current frame picture transmission process, if the current frame picture is a dynamic frame picture, confirm the best encoding rate in the transmission process according to the fluctuation characteristics of the bandwidth, and complete the transmission process of the current frame picture.

2. The conference audiovisual based transmission optimization method of claim 1, wherein, In the step one, the specific way of gray processing of the adjacent frame pictures is as follows: Record the frame picture associated with the current time as a standard frame picture, record the two groups of frame pictures associated with the front end time of the standard frame picture as the frame pictures to be verified of the standard frame picture, sort the frame pictures to be verified and the standard frame picture according to the time sequence, and confirm the picture sorting set; Gray process the three groups of frame pictures in the picture sorting set: confirm the RGB values associated with different pixel points in a single group of frame pictures, and obtain the gray value associated with the corresponding pixel point by using 0.299×R+0.587×G+0.114×B=gray value, and gray process the frame picture associated with the confirmed gray value, confirm the gray image associated with the corresponding frame picture, and synchronously confirm the gray image associated with the corresponding frame picture.

3. The conference audiovisual based transmission optimization method of claim 2, wherein, In the step one, the specific way of confirming whether there is a change point in the current frame picture is as follows: Confirm the gray image associated with the standard frame picture based on the picture sorting set after gray processing; Confirm the different gray values associated with the same position pixel points from the adjacent gray images, and confirm the gray difference value by difference processing of the two groups of gray values, and the gray difference value is greater than or equal to 0, and evaluate whether the gray difference value satisfies gray difference value≥Y1, if yes, record the corresponding pixel point as a change point, if not, do not perform any labeling; Confirm the change points in the subsequent two groups of adjacent gray images in turn, record the same position change points from the confirmation process, confirm the total number Zz of different position change points from the three groups of gray images, record the total number of pixel points in the standard frame picture as Zs, if (Zz÷Zs)≥0.3, record the standard frame picture as a dynamic frame picture, otherwise, label the standard frame picture as a static frame picture.

4. The conference audiovisual based transmission optimization method of claim 2, wherein, In the step one, the specific way of confirming whether there is a dynamic area in the current frame picture is as follows: The gradient of different gray images in the picture ordering set is confirmed, the gray value associated with the corresponding pixel point in the corresponding gray image is confirmed, and the vertical gradient and vertical gradient associated with the specified gray point are confirmed based on the Sobel algorithm. According to the confirmed vertical gradient and vertical gradient, the comprehensive gradient associated with the corresponding pixel point is confirmed, wherein: The pixel points satisfying the comprehensive gradient >= Y1 are marked as gradient points, otherwise, no mark is made, wherein Y1 is a preset value; Confirm a plurality of groups of gradient points associated with a single group of gray images in turn, connect the adjacent gradient points, confirm the gradient connecting line, record the confirmed gradient connecting line as a gradient contour, record the area contained by the gradient contour as a gradient area, and confirm the center point of the corresponding gradient area, confirm the gradient area existing in each gray image, and confirm whether there is a spatial position intersecting gradient area in the adjacent gray image, if yes, record it as a same type dynamic area, if not, do not perform any labeling; From three groups of gray images in the same kind of dynamic area, and from the first gray image corresponding to the gradient region center point, the change vector of the three groups of gray images corresponding to the gradient region center point is confirmed, and the total number of pixel points XH passed by the change vector is confirmed. If XH≥50, the standard frame picture is marked as a dynamic frame picture, otherwise, the standard frame picture is marked as a static frame picture.

5. The conference audiovisual based transmission optimization method of claim 1, wherein, In step two, the execution encoding rate associated with the static frame picture is the preset encoding rate, and the encoding rate is adjusted during the transmission process of the static frame picture, so that the adjusted encoding rate is the execution encoding rate, and the transmission process of the current frame picture is completed.

6. The conference audiovisual based transmission optimization method of claim 1, wherein, In step two, the confirmation method of the bandwidth fluctuation characteristics of the dynamic frame picture is: Take the bandwidth at the current time as the reference bandwidth, and trace back from the current time, the unit time traced back is not less than ten groups, from the running data of the traced back unit time, confirm the historical bandwidth associated with the corresponding time, and according to the change process of the bandwidth data, confirm the change bandwidth associated between adjacent time, its change bandwidth = | the change bandwidth of the next time - the change bandwidth of the previous time |, and select the minimum value and the maximum value from the confirmed several groups of change bandwidth, to generate the change interval of the corresponding change bandwidth, and mark the generated change interval as the fluctuation characteristics of the corresponding bandwidth.

7. The conference audiovisual based transmission optimization method of claim 6, wherein, In step two, the specific method of confirming the optimal encoding rate in the transmission process of the dynamic frame picture is: According to the confirmed fluctuation characteristics and the bandwidth D1 associated with the current time, confirm the associated bandwidth interval of the next time, the fluctuation characteristics is [DBmin, DBmax], the associated bandwidth interval is [D1+DBmin, D1+DBmax], confirm the maximum encoding rate BM1 associated with the bandwidth (D1+DBmax) from the historical processing process data, and confirm the maximum encoding rate BM2 associated with the bandwidth (D1+DBmin) at the same time. Confirm the dynamic characteristics of the dynamic frame picture from the dynamic characteristics, and mark the total number of change points in the dynamic frame picture as Zz, use: DZ=Zz×C1 to confirm the dynamic characteristics associated with the dynamic frame picture, or confirm the total number of pixel points XH passed by the change vector in the dynamic frame picture, use: Dz=XH×C2 to confirm the dynamic characteristics associated with the dynamic frame picture, wherein C1 and C2 are both preset fixed coefficient factors. From the feature comparison table, confirm the encoding rate BM associated with the dynamic characteristics: If yes, directly select BM as the execution encoding rate and execute the transmission process of the dynamic frame picture; If not, determine the difference between BM and the numerical range associated with BM1 and BM2. If it exceeds the numerical range, select the maximum value of the numerical range as the execution encoding rate to execute the transmission process of the dynamic frame picture. If it does not exceed the numerical range, select the minimum value of the numerical range as the execution encoding rate to execute the transmission process of the dynamic frame picture.

Images