Cloud platform data processing method, device, medium and computer equipment

CN118338042BActive Publication Date: 2024-11-05中亿(深圳)信息科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410761259.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-06-13
Publication Date
2024-11-05
Estimated Expiration
2044-06-13

AI Technical Summary

Technical Problem

[0002]随着视频分辨率和帧率的提高,视频数据的处理呈指数级增长,但是由于处理视频平台的处理能力是有限的,处理数据量过大会导致平台崩溃或者运行缓慢,且一般在进行压缩过程中是采用同个压缩方式对整个视频进行压缩,如果压缩方式精度不高会导致视频质量整体降低且整体压缩后的视频可能还会占据部分存储空间,进而导致压缩的无效性

Benefits of technology

[0059] Based on the operating capabilities of the platform and the quality requirements for the video, the ability to perform similar analysis on adjacent frame images is determined, and the first frame and the remaining frames in the continuous video are separately encoded to ensure the effective processing of the platform As well as the video quality of the compressed video, ensuring the effectiveness of the compression.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118338042B_ABST
    Figure CN118338042B_ABST
Patent Text Reader

Abstract

The present application proposes a cloud platform data processing method, device, medium and computer equipment, which belongs to the field of data processing technology. The method includes: performing similarity analysis on adjacent frame images in sequence, setting similar symbols to each frame target image, determining the current specification of the first frame image under the continuous video segment, and combining the number of frames and symbol differences to determine the maximum and minimum operating capabilities required for the corresponding continuous video segment, and then obtaining the best operating capability based on the current operating capability of the cloud platform; synchronously splitting several continuous video segments, matching 1 encoding channel to each split result, and determining the compression method of the first frame image according to the video quality requirements of the target video; compressing and encoding the first frame image according to the corresponding compression method, and calling encoding and special encoding for the remaining frame images in the same continuous video segment. Ensure the effective processing of the platform and the video quality of the compressed video, and ensure the effectiveness of compression.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a cloud platform data processing method, device, medium and computer equipment, and belongs to the technical field of data processing. Background Art

[0002] As video resolution and frame rate increase, video data processing grows exponentially. However, due to the limited processing capacity of video processing platforms, processing too much data will cause the platform to crash or run slowly. In addition, the same compression method is generally used to compress the entire video during the compression process. If the compression method is not accurate enough, the overall video quality will be reduced and the compressed video may occupy some storage space, which will lead to the ineffectiveness of the compression.

[0003] Therefore, the present invention provides a cloud platform data processing method, device, medium and computer equipment. Summary of the invention

[0004] In view of the deficiencies of the above-mentioned prior art, the purpose of the present invention is to provide a cloud platform data processing method, device, medium and computer equipment, which determines the ability to perform similarity analysis on adjacent frame images based on the platform's operating capabilities and video quality requirements, and by encoding the first frame and the remaining frames in a continuous video segment separately, ensures the effective processing of the platform and the video quality of the compressed video, thereby ensuring the effectiveness of the compression.

[0005] The present invention provides a cloud platform data processing method, comprising:

[0006] Step 1: Obtain the current operating capacity of the cloud platform and the video quality requirements for the target video, wherein the target video includes multiple frames of target images;

[0007] Step 2: Perform similarity analysis on adjacent frame images in sequence, set similarity symbols to each target frame image, and determine continuous video segments;

[0008] Step 3: construct a pixel matrix of the first frame image under each continuous video segment, determine the current specification of the corresponding first frame image, and determine the maximum operating capacity and the minimum operating capacity required for the corresponding continuous video segment according to the number of frames contained in the corresponding continuous video segment and the symbol difference of similar symbols under the corresponding continuous video segment;

[0009] Step 4: Comprehensively analyze all maximum operating capabilities and minimum operating capabilities under the target video according to the current operating capabilities of the cloud platform to obtain the optimal operating capability of the cloud platform;

[0010] Step 5: Synchronously splitting a number of continuous video segments based on the optimal operation capability to obtain N split results, and matching one encoding channel to each split result, and determining a compression method for the first frame image in each continuous video segment based on the video quality requirements of the target video;

[0011] Step 6: In the N coding channels, the first frame image in the same continuous video segment under the corresponding coding channel is compressed and encoded in turn according to the corresponding compression method, and the remaining frame images in the same continuous video segment are called and encoded and specially encoded to obtain the compression result based on the corresponding coding channel.

[0012] In a possible implementation, performing similarity analysis on adjacent frame images in sequence and setting a similarity symbol to each frame of target image includes:

[0013] Overlapping a first frame image and a second frame image in adjacent frame images, and obtaining non-overlapping pixel points of the first frame image and non-overlapping pixel points of the second frame image, wherein the first frame image is located in front of the second frame image;

[0014] Compare the brightness values ​​of non-overlapping pixels of the first frame image and non-overlapping pixels of the second frame image at the same position, and obtain the brightness difference of each position respectively;

[0015] Determine the initial similarity value of the adjacent frame images according to a first ratio of the number of overlapping pixels to the total number of pixels of the frame image and a second ratio of the number of brightness differences that meet the difference range standard to the number of non-overlapping pixels;

[0016] Respectively acquiring picture description information of non-overlapping pixel points of the first frame image and picture description information of non-overlapping pixel points of the second frame image to obtain adjustment similarity coefficients of the adjacent frame images;

[0017] Determine a final similarity value based on the initial similarity value and the adjusted similarity coefficient and in combination with the variance under all brightness differences;

[0018] A similarity symbol matching the final similarity value is obtained from the value-symbol comparison table.

[0019] In a possible implementation manner, determining continuous video segments includes:

[0020] If the first frame image is the first frame image of the target video, assigning two identical similar symbols to the first frame image;

[0021] If the first frame image is not the first frame image of the target video, assigning a similar symbol under which the first frame image is the first image in adjacent frame images and a similar symbol under which the first frame image is the second image in adjacent frame images to the first frame image;

[0022] If the second frame image is the last frame image of the target video, assigning two identical similar symbols to the second frame image;

[0023] If the second frame image is not the last frame image of the target video, assigning a similar symbol under which the second frame image is the first image in the adjacent frame images and a similar symbol under which the second frame image is the second image in the adjacent frame images to the second frame image;

[0024] According to the time sequence and the consistency of the symbol combination of each frame of the target image, the preliminary division points are determined to obtain a number of continuous video segments, and each continuous video segment contains at least one frame of the target image.

[0025] In a possible implementation, determining the maximum operating capacity and the minimum operating capacity required for the corresponding continuous video segments includes:

[0026] Determine a first difference quantity and a first difference distance according to a symbol combination of each frame of the target image in the corresponding continuous video segment;

[0027] Locking a first symbol corresponding to a maximum similarity value and a second symbol corresponding to a minimum similarity value under corresponding continuous video segments, and determining a second difference quantity and a second difference distance between the first symbol and the second symbol;

[0028] At the same time, the number of repetitions of each similar symbol in the corresponding continuous video segments is obtained;

[0029] Allocating a difference conversion coefficient to a corresponding frame target image according to the first difference amount, the second difference amount, the first difference distance, and the second difference distance;

[0030] According to the number of repetitions and the difference conversion coefficient, a total conversion coefficient corresponding to the repetition symbol is obtained;

[0031] All total conversion coefficients are summed to obtain the overall conversion coefficients of the corresponding continuous video segments;

[0032] Extracting a maximum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a first conversion coefficient, and at the same time, extracting a minimum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a second conversion coefficient;

[0033] The first conversion coefficient and the second conversion coefficient are optimized and adjusted respectively based on the overall conversion coefficient to obtain a maximum operating capacity and a minimum operating capacity.

[0034] In one possible implementation, obtaining the optimal operating capability of the cloud platform includes:

[0035] Obtaining first operation information of the cloud platform in each historical period, wherein the first operation information includes the duration of processing of each received historical video by the cloud platform and the amount of resources called at each processing moment;

[0036] Extracting the maximum operation processing volume, the second largest operation processing volume, and the duration period under the maximum operation processing volume and the second largest operation processing volume from the first operation information to obtain the historical processing capacity based on the corresponding historical period;

[0037] The historical processing capacity is plotted according to the chronological order of historical time periods to obtain a capacity curve;

[0038] Obtaining a reference curve according to the position of the current operating capacity in the capacity curve;

[0039] Performing a first adjustment on the reference curve according to the curvature of each convex point on the reference curve;

[0040] At the same time, performing a second adjustment on the reference curve according to the curvature of each concave point on the reference curve;

[0041] Based on the first attenuation area under the first adjustment curve and the second attenuation area under the second adjustment curve, and combined with the closed area under the reference curve, a first adjustment coefficient for maximum operating capacity and a second adjustment coefficient for minimum operating capacity are obtained, thereby obtaining the optimal operating capacity.

[0042] In one possible implementation, N split results are obtained, including:

[0043] Taking the maximum operating capacity required for each continuous video segment as a benchmark and the optimal operating capacity as a boundary, constructing a capacity constraint condition and obtaining a number of split points;

[0044] A plurality of continuous video segments are synchronously split according to the splitting points to obtain N splitting results.

[0045] In a possible implementation, the compression method of the first frame image in each continuous video segment is determined according to the video quality requirement of the target video, including:

[0046] Determine the image information amount of the first frame image in each continuous video segment respectively, and obtain the optimal video requirement of the corresponding first frame image in combination with the maximum operation capacity of the corresponding continuous video segment;

[0047] Determining the effective video quality of the corresponding first frame image according to the video quality requirement and the optimal video requirement;

[0048] A compression method matching the effective video quality is obtained from a quality-compression comparison table.

[0049] The present invention provides a cloud platform data processing device, comprising:

[0050] A preliminary data acquisition module, used to acquire the current operating capability of the cloud platform and the video quality requirements for the target video, wherein the target video includes multiple frames of target images;

[0051] A symbol setting module is used to perform similarity analysis on adjacent frame images in sequence, set similar symbols to each frame of target image, and determine continuous video segments;

[0052] A capability determination module is used to construct a pixel matrix of the first frame image under each continuous video segment, determine the current specification quantity of the corresponding first frame image, and determine the maximum operating capability and the minimum operating capability required for the corresponding continuous video segment according to the number of frames contained in the corresponding continuous video segment and the symbol difference of similar symbols under the corresponding continuous video segment;

[0053] A comprehensive analysis module, used to comprehensively analyze all maximum operating capabilities and minimum operating capabilities under the target video according to the current operating capabilities of the cloud platform, to obtain the optimal operating capability of the cloud platform;

[0054] A compression method determination module is used to synchronously split a plurality of continuous video segments to obtain N split results based on the optimal operation capability, and to match one encoding channel to each split result, and to determine a compression method for the first frame image in each continuous video segment according to the video quality requirements of the target video;

[0055] The encoding module is used to compress and encode the first frame image in the same continuous video segment under the corresponding encoding channel in N encoding channels in turn according to the corresponding compression method, and to call encoding and special encoding for the remaining frame images in the same continuous video segment to obtain a compression result based on the corresponding encoding channel.

[0056] The present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor is enabled to execute the steps of any one of the methods.

[0057] The present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of any one of the methods.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] The ability to perform similarity analysis on adjacent frames is determined based on the platform's operating capabilities and the quality requirements for the video. The first frame and the remaining frames in a continuous video segment are encoded separately to ensure effective processing of the platform and the quality of the compressed video, thereby ensuring the effectiveness of compression. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a flow chart of the cloud platform data processing method;

[0061] Figure 2 This is a structural diagram of the cloud platform data processing device. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of this application.

[0063] It should be noted that when an element is referred to as being "fixed on" or "set on" another component, it can be directly on the other component or indirectly set on the other component; when a component is referred to as being "connected to" another component, it can be directly connected to the other component or indirectly connected to the other component.

[0064] It should be understood that the orientation or position relationship indicated by terms such as "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or component referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0065] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, "multiple" and "several" mean two or more, unless otherwise clearly and specifically defined.

[0066] It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the conditions under which this application can be implemented. Therefore, they have no substantive technical significance. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical content disclosed in this application without affecting the effects and purposes that can be achieved by this application.

[0067] Embodiment 1:

[0068] The present invention provides a cloud platform data processing method, such as Figure 1 As shown, including:

[0069] Step 1: Obtain the current operating capacity of the cloud platform and the video quality requirements for the target video, wherein the target video includes multiple frames of target images;

[0070] Step 2: Perform similarity analysis on adjacent frame images in sequence, set similarity symbols to each target frame image, and determine continuous video segments;

[0071] Step 3: construct a pixel matrix of the first frame image under each continuous video segment, determine the current specification of the corresponding first frame image, and determine the maximum operating capacity and the minimum operating capacity required for the corresponding continuous video segment according to the number of frames contained in the corresponding continuous video segment and the symbol difference of similar symbols under the corresponding continuous video segment;

[0072] Step 4: Comprehensively analyze all maximum operating capabilities and minimum operating capabilities under the target video according to the current operating capabilities of the cloud platform to obtain the optimal operating capability of the cloud platform;

[0073] Step 5: Synchronously splitting a number of continuous video segments based on the optimal operation capability to obtain N split results, and matching one encoding channel to each split result, and determining a compression method for the first frame image in each continuous video segment based on the video quality requirements of the target video;

[0074] Step 6: In the N coding channels, the first frame image in the same continuous video segment under the corresponding coding channel is compressed and encoded in turn according to the corresponding compression method, and the remaining frame images in the same continuous video segment are called and encoded and specially encoded to obtain the compression result based on the corresponding coding channel.

[0075] In one embodiment, the current operating capacity is based on the difference between the factory-set capacity of the cloud platform and the capacity currently used for operation.

[0076] In one embodiment, the video quality requirement is that the video should be clear after compression and decoding, such as reaching 270P, 480P, 720P, 1080P, etc. That is, the requirement is set subjectively by the user, and it only needs to meet the clarity requirement.

[0077] In one embodiment, the target video refers to a video that needs to be compressed later, and the video can be any video that the user needs to compress, without limiting the type.

[0078] In one embodiment, similarity analysis refers to determining similarity between adjacent frame images to set similarity symbols.

[0079] For example, adjacent frame images include: frame image 1 and frame image 2. At this time, the similarity between frame image 1 and frame image 2 is determined, and the obtained similarity value is obtained from the value-symbol comparison table to obtain the similarity symbol to attach to frame image 1 and frame image 2.

[0080] For example, the similarity symbol between frame image 1 and frame image 2 is #1, and the similarity symbol between frame image 2 and frame image 3 is ¥1. At this time, the symbol group attached to frame image 2 is [#1, ¥1], where frame image 2 and frame image 3 are also adjacent frame images.

[0081] In one embodiment, continuous video segments are divided based on the consistency of similar symbols, for example: frame image 1--[r1,#1], frame image 2--[#1,¥1], frame image 3--[¥1,¥1], frame image 4--[¥1,¥1], frame image 5--[¥1,r2]--frame image 6--[r2,r2]. At this time, frame image 1 is regarded as a continuous video segment, frame images 2 to 5 are regarded as a continuous video segment, and frame image 6 is regarded as a continuous video segment.

[0082] In one embodiment, the pixel matrix actually represents the image in the form of a matrix, and each element in the pixel matrix corresponds one-to-one to each pixel in the first frame image, and the current specification quantity is obtained based on the cumulative sum of the information content of the pixel points, for example, the specification quantity of the first frame image is 1.0M.

[0083] In one embodiment, the maximum operating capacity and the minimum operating capacity refer to the capacity range required by the cloud platform when processing the continuous video segments.

[0084] In one embodiment, the determination of the best operating capacity is to provide a basis for subsequent synchronization splitting.

[0085] In one embodiment, if there are N split results, there are N encoding channels.

[0086] In one embodiment, the compression method of each first frame image may be different, because both the compression quality and the effective operation of the platform must be guaranteed, wherein the compression method is related to the compression frame rate and the compression format.

[0087] In one embodiment, a coding channel adopts a compression method. For example, there are frame image 1, frame image 2, and frame image 3. At this time, frame image 1 is compressed as the first frame image according to the compression method. Thereafter, frame image 2 is called for coding, and the called coding represents the compression result of calling and using frame image 1. Special coding refers to the coding after compressing inconsistent image information between frame image 2 and frame image 1, and in the subsequent decoding process, frame image 2 can be effectively restored by calling coding and special coding, that is, in the process of compressing frame image 2, only inconsistent image information needs to be compressed, and the remaining information can be called from frame image 1, and the restoration is the restoration of the same position point, which effectively reduces the space occupied by compression and can also ensure the quality of the frame image.

[0088] The beneficial effects of the above technical solution are: based on the operating capabilities of the platform and the quality requirements for the video, the ability to perform similarity analysis on adjacent frame images is determined, and by encoding the first frame and the remaining frames in a continuous video segment separately, the effective processing of the platform and the video quality of the compressed video are guaranteed, thereby ensuring the effectiveness of compression.

[0089] Embodiment 2:

[0090] The present invention provides a cloud platform data processing method, which performs similarity analysis on adjacent frame images in sequence and sets similarity symbols to each frame target image, including:

[0091] Overlapping a first frame image and a second frame image in adjacent frame images, and obtaining non-overlapping pixel points of the first frame image and non-overlapping pixel points of the second frame image, wherein the first frame image is located in front of the second frame image;

[0092] Compare the brightness values ​​of non-overlapping pixels of the first frame image and non-overlapping pixels of the second frame image at the same position, and obtain the brightness difference of each position respectively;

[0093] Determine the initial similarity value of the adjacent frame images according to a first ratio of the number of overlapping pixels to the total number of pixels of the frame image and a second ratio of the number of brightness differences that meet the difference range standard to the number of non-overlapping pixels;

[0094] Respectively acquiring picture description information of non-overlapping pixel points of the first frame image and picture description information of non-overlapping pixel points of the second frame image to obtain adjustment similarity coefficients of the adjacent frame images;

[0095] Determine a final similarity value based on the initial similarity value and the adjusted similarity coefficient and in combination with the variance under all brightness differences;

[0096] A similarity symbol matching the final similarity value is obtained from the value-symbol comparison table.

[0097] In one embodiment, the overlapping process refers to aligning the first frame image with the second frame image at the same position point. If the information of the two pixels at the position point is completely consistent, they are regarded as overlapping points, otherwise they are regarded as non-overlapping points.

[0098] In one embodiment, the brightness difference=|brightness values ​​of non-overlapping pixels of the first frame image−brightness values ​​of non-overlapping pixels of the second frame image|.

[0099] In one embodiment, the adjustment similarity coefficient=sim(My1, My2), wherein My1 is the picture description information of non-overlapping pixel points of the first frame image, and My2 is the picture description information of non-overlapping pixel points of the second frame image.

[0100] In one embodiment, .

[0101] In one embodiment, the final similarity value=initial similarity value×adjusted similarity coefficient+variance×similarity conversion coefficient.

[0102] In one embodiment, the value-symbol comparison table includes similar symbols that match different final similarity values. For example, the similarity value of 0.9 corresponds to the similarity symbol #1, where the same similar symbol corresponds to a value range, for example, the value range corresponding to #1 is 0.9~0.93.

[0103] The beneficial effect of the above technical solution is: by overlapping adjacent frame images, a similarity value is preliminarily determined based on the ratio of overlapping points and the ratio of non-overlapping points, and the similarity is re-determined in combination with the picture description information, thereby ensuring the reliability of the final similarity value acquisition and providing a basis for matching similar symbols.

[0104] Embodiment 3:

[0105] The present invention provides a cloud platform data processing method for determining continuous video segments, comprising:

[0106] If the first frame image is the first frame image of the target video, assigning two identical similar symbols to the first frame image;

[0107] If the first frame image is not the first frame image of the target video, assigning a similar symbol under which the first frame image is the first image in adjacent frame images and a similar symbol under which the first frame image is the second image in adjacent frame images to the first frame image;

[0108] If the second frame image is the last frame image of the target video, assigning two identical similar symbols to the second frame image;

[0109] If the second frame image is not the last frame image of the target video, assigning a similar symbol under which the second frame image is the first image in the adjacent frame images and a similar symbol under which the second frame image is the second image in the adjacent frame images to the second frame image;

[0110] According to the time sequence and the consistency of the symbol combination of each frame of the target image, the preliminary division points are determined to obtain a number of continuous video segments, and each continuous video segment contains at least one frame of the target image.

[0111] In one embodiment, when it is the last frame image or the first frame image, the two symbols in the corresponding symbol group are the same.

[0112] The beneficial effect of the above technical solution is: by determining the position of the frame image, the effective setting of the frame image symbol combination is achieved, and then the division points are determined according to the consistency to obtain several continuous video segments, which provides a basis for subsequent analysis and setting of the compression method.

[0113] Embodiment 4:

[0114] The present invention provides a cloud platform data processing method, which determines the maximum operating capacity and the minimum operating capacity required for corresponding continuous video segments, including:

[0115] Determine a first difference quantity and a first difference distance according to a symbol combination of each frame of the target image in the corresponding continuous video segment;

[0116] Locking a first symbol corresponding to a maximum similarity value and a second symbol corresponding to a minimum similarity value under corresponding continuous video segments, and determining a second difference quantity and a second difference distance between the first symbol and the second symbol;

[0117] At the same time, the number of repetitions of each similar symbol in the corresponding continuous video segments is obtained;

[0118] Allocating a difference conversion coefficient to a corresponding frame target image according to the first difference amount, the second difference amount, the first difference distance, and the second difference distance;

[0119] According to the number of repetitions and the difference conversion coefficient, a total conversion coefficient corresponding to the repetition symbol is obtained;

[0120] All total conversion coefficients are summed to obtain the overall conversion coefficients of the corresponding continuous video segments;

[0121] Extracting a maximum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a first conversion coefficient, and at the same time, extracting a minimum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a second conversion coefficient;

[0122] The first conversion coefficient and the second conversion coefficient are optimized and adjusted respectively based on the overall conversion coefficient to obtain a maximum operating capacity and a minimum operating capacity.

[0123] In one embodiment, the symbol combination [$01 $21], the first difference quantity at this time is calculated according to the distance editing formula, the 0 in "$01" does not correspond to the 2 in "$21", and at this time, the first difference quantity is 1.

[0124] Get the final similarity values ​​corresponding to $01 and $21 respectively. At this time, , where D1 is the first difference distance, The set distance value of $01; for The set distance value; for The final similarity value under ; for The final similarity value under ; e represents the exponential function symbol.

[0125] In one embodiment, the maximum similarity value is the maximum value of all final similarity values ​​in the same continuous video segment, and the minimum similarity value is the minimum value of all final similarity values ​​in the same continuous video segment.

[0126] In one embodiment, the principles of the second difference quantity and the second difference distance are similar to those of the first difference quantity and the first difference distance, and are not described in detail here.

[0127] In one embodiment, .

[0128] In one embodiment, the total conversion coefficient=the number of repetitions×the average value of all difference conversion coefficients under all symbol combinations involving repeated symbols.

[0129] In one embodiment, the overall conversion coefficient is the sum of all total conversion coefficients under the continuous video segment.

[0130] In one embodiment, the first conversion coefficient=maximum conversion coefficient×number of frames;

[0131] The second conversion coefficient=the minimum conversion coefficient×the number of frames.

[0132] Maximum operating capacity = overall conversion coefficient × capacity standardization coefficient × ln(2 + |overall conversion coefficient-maximum conversion coefficient| / max(overall conversion coefficient, maximum conversion coefficient)).

[0133] Minimum operating capacity = overall conversion coefficient × capacity standardization coefficient × ln((e-1) + minimum conversion coefficient / (overall conversion coefficient + minimum conversion coefficient)), where e is a constant with a value of 2.7.

[0134] The beneficial effects of the above technical solution are: determining the number and distance of differences through symbol combinations, and determining another number and distance of differences based on symbols corresponding to the maximum similarity value and the minimum similarity value. The combination of the two can effectively determine the difference conversion coefficient, providing a basis for calculating operating capabilities. The existence of differences means that the platform needs to perform redundant operations, and by determining the total conversion coefficient, the overall conversion coefficient, the first conversion coefficient and the second conversion coefficient of repeated symbols, the maximum operating capacity and the minimum operating capacity of the platform can be effectively obtained, providing a basis for determining the best, ensuring the effective operation of the platform and effective compression of image quality.

[0135] Embodiment 5:

[0136] The present invention provides a cloud platform data processing method to obtain the optimal operation capability of the cloud platform, comprising:

[0137] Obtaining first operation information of the cloud platform in each historical period, wherein the first operation information includes the duration of processing of each received historical video by the cloud platform and the amount of resources called at each processing moment;

[0138] Extracting the maximum operation processing volume, the second largest operation processing volume, and the duration period under the maximum operation processing volume and the second largest operation processing volume from the first operation information to obtain the historical processing capacity based on the corresponding historical period;

[0139] The historical processing capacity is plotted according to the chronological order of historical time periods to obtain a capacity curve;

[0140] Obtaining a reference curve according to the position of the current operating capacity in the capacity curve;

[0141] Performing a first adjustment on the reference curve according to the curvature of each convex point on the reference curve;

[0142] At the same time, performing a second adjustment on the reference curve according to the curvature of each concave point on the reference curve;

[0143] Based on the first attenuation area under the first adjustment curve and the second attenuation area under the second adjustment curve, and combined with the closed area under the reference curve, a first adjustment coefficient for maximum operating capacity and a second adjustment coefficient for minimum operating capacity are obtained, thereby obtaining the optimal operating capacity.

[0144] In one embodiment, the operating processing capacity is determined by the total amount of called resources at each moment, so the maximum operating processing capacity and the second largest operating processing capacity can be obtained. For example, the maximum operating processing capacity is at time t1. At this time, t1 is taken as the center to expand the left and right time of t1 so that the operating processing capacity on its left and right boundaries is consistent with the second largest operating processing capacity. At this time, the obtained segment is the required duration segment.

[0145] In one embodiment, the historical processing capacity is obtained by averaging the processing capacity at each moment in the required duration period. The maximum operating processing capacity does not appear at only one moment, but at many moments. Therefore, the historical processing capacity of each required duration period is obtained separately, and the capacity curve can be drawn in chronological order.

[0146] In one embodiment, the position of the current operating capacity in the capacity curve is used to lock the leftmost and rightmost positions, so as to obtain a reference curve.

[0147] In one embodiment, the adjustment of the curve is based on a curvature-adjustment amount comparison table to obtain the downward adjustment amount related to the curvature of the convex point and the upward adjustment amount related to the curvature of the concave point, so as to obtain the adjusted curve. First, the first adjustment curve is obtained after adjustment only according to the curvature of the convex point, and the closed area of ​​the reference curve - the closed area of ​​the first adjustment curve = the first attenuation area. Similarly, the closed area of ​​the second adjustment curve - the closed area of ​​the reference curve = the second attenuation area.

[0148] In one embodiment, the first adjustment coefficient=min(second attenuation area, first attenuation area) / enclosed area.

[0149] The second adjustment coefficient = the second attenuation area + the first attenuation area / (2×enclosed area).

[0150] In one embodiment, the optimal operating capacity=(maximum operating capacity×(1-first adjustment coefficient)+minimum operating capacity×(1+second adjustment coefficient)) / 2.

[0151] The beneficial effect of the above technical solution is: obtaining the historical processing capacity of several historical periods to draw the capacity curve, and combining the occurrence position to obtain the reference curve to narrow the analysis scope, and then adjusting the curve upward and downward according to the curvature to obtain the attenuation area, providing guarantee for capacity adjustment and facilitating obtaining the optimal operating capacity.

[0152] Embodiment 6:

[0153] The present invention provides a cloud platform data processing method, which obtains N splitting results, including:

[0154] Taking the maximum operating capacity required for each continuous video segment as a benchmark and the optimal operating capacity as a boundary, constructing a capacity constraint condition and obtaining a number of split points;

[0155] A plurality of continuous video segments are synchronously split according to the splitting points to obtain N splitting results.

[0156] In one embodiment, the capacity constraint condition is: sum{maximum operating capacity of a number of consecutive video segments}<optimal operating capacity.

[0157] In one embodiment, synchronous splitting refers to synchronously splitting the target video based on the splitting points.

[0158] The beneficial effects of the above technical solution are: constructing capability constraints to achieve synchronous splitting of the video, saving splitting time, and ensuring the rapidity of subsequent encoding.

[0159] Embodiment 7:

[0160] The present invention provides a cloud platform data processing method, which determines the compression method of the first frame image in each continuous video segment according to the video quality requirement of the target video, including:

[0161] Determine the image information amount of the first frame image in each continuous video segment respectively, and obtain the optimal video requirement of the corresponding first frame image in combination with the maximum operation capacity of the corresponding continuous video segment;

[0162] Determining the effective video quality of the corresponding first frame image according to the video quality requirement and the optimal video requirement;

[0163] A compression method matching the effective video quality is obtained from a quality-compression comparison table.

[0164] In one embodiment, the image information volume is the information contained in the first frame of the image, for example, 1.9M.

[0165] In one embodiment, the optimal video requirement is obtained based on a matching table of quantity-capability-requirement, wherein the table includes the image information quantity and the maximum operating capacity under different combinations, and the requirements matched therewith.

[0166] In one embodiment, the effective video quality is based on matching the required quality in the optimal video requirement with the quality requirement range in the video quality requirement to obtain the closest quality, which is regarded as the effective video quality.

[0167] In one embodiment, the quality-compression comparison table includes different effective video qualities and pre-set compression methods that match them.

[0168] The beneficial effect of the above technical solution is: by preliminarily determining the optimal video requirements for the first frame image, and comparing and analyzing with the video quality requirements to determine the effective video instructions, it is easy to obtain the compression method and effectively ensure the quality of video compression.

[0169] Embodiment 8:

[0170] The present invention provides a cloud platform data processing device, such as Figure 2 As shown, including:

[0171] A preliminary data acquisition module, used to acquire the current operating capability of the cloud platform and the video quality requirements for the target video, wherein the target video includes multiple frames of target images;

[0172] A symbol setting module is used to perform similarity analysis on adjacent frame images in sequence, set similar symbols to each frame of target image, and determine continuous video segments;

[0173] A capability determination module is used to construct a pixel matrix of the first frame image under each continuous video segment, determine the current specification quantity of the corresponding first frame image, and determine the maximum operating capability and the minimum operating capability required for the corresponding continuous video segment according to the number of frames contained in the corresponding continuous video segment and the symbol difference of similar symbols under the corresponding continuous video segment;

[0174] A comprehensive analysis module, used to comprehensively analyze all maximum operating capabilities and minimum operating capabilities under the target video according to the current operating capabilities of the cloud platform, to obtain the optimal operating capability of the cloud platform;

[0175] A compression method determination module is used to synchronously split a plurality of continuous video segments to obtain N split results based on the optimal operation capability, and to match one encoding channel to each split result, and to determine a compression method for the first frame image in each continuous video segment according to the video quality requirements of the target video;

[0176] The encoding module is used to compress and encode the first frame image in the same continuous video segment under the corresponding encoding channel in N encoding channels in turn according to the corresponding compression method, and to call encoding and special encoding for the remaining frame images in the same continuous video segment to obtain a compression result based on the corresponding encoding channel.

[0177] The beneficial effects of the above technical solution are: based on the operating capabilities of the platform and the quality requirements for the video, the ability to perform similarity analysis on adjacent frame images is determined, and by encoding the first frame and the remaining frames in a continuous video segment separately, the effective processing of the platform and the video quality of the compressed video are guaranteed, thereby ensuring the effectiveness of compression.

[0178] Embodiment 9:

[0179] The present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor is enabled to execute the steps of any one of the methods.

[0180] Embodiment 10:

[0181] The present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of any one of the methods.

[0182] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A cloud platform data processing method, characterized in that: include: Step 1: Obtain the current operating capacity of the cloud platform and the video quality requirements for the target video, wherein the target video includes multiple frames of target images; Step 2: Perform similarity analysis on adjacent frame images in sequence, set similarity symbols to each target frame image, and determine continuous video segments; Step 3: construct a pixel matrix of the first frame image under each continuous video segment, determine the current specification of the corresponding first frame image, and determine the maximum operating capacity and the minimum operating capacity required for the corresponding continuous video segment according to the number of frames contained in the corresponding continuous video segment and the symbol difference of similar symbols under the corresponding continuous video segment, wherein the current specification is obtained based on the cumulative sum of the information content of the pixel points; Step 4: Comprehensively analyze all maximum operating capabilities and minimum operating capabilities under the target video according to the current operating capabilities of the cloud platform to obtain the optimal operating capability of the cloud platform; Step 5: Synchronously splitting a number of continuous video segments based on the optimal operation capability to obtain N split results, and matching one encoding channel to each split result, and determining a compression method for the first frame image in each continuous video segment based on the video quality requirements of the target video; Step 6: in the N coding channels, the first frame image in the same continuous video segment under the corresponding coding channel is compressed and encoded in turn according to the corresponding compression mode, and the remaining frame images in the same continuous video segment are called and encoded and specially encoded to obtain the compression result based on the corresponding coding channel; Among them, calling the code means calling and using the compression result of the first frame image, and the dedicated code refers to the code after compressing the image information that is inconsistent with the first frame image, and in the subsequent decoding process, the remaining frame images can be effectively restored by calling the code and the dedicated code; The maximum operating capacity and the minimum operating capacity refer to the capacity range required by the cloud platform when processing the continuous video segment. The determining of the maximum operating capacity and the minimum operating capacity required for the corresponding continuous video segments includes: Determining a first difference quantity and a first difference distance according to a symbol combination of each frame of the target image in the corresponding continuous video segment, wherein the symbol combination of each frame of the target image is composed of similar symbols of each frame of the target image; Locking a first symbol corresponding to a maximum similarity value and a second symbol corresponding to a minimum similarity value under corresponding continuous video segments, and determining a second difference quantity and a second difference distance between the first symbol and the second symbol; At the same time, the number of repetitions of each similar symbol in the corresponding continuous video segments is obtained; Allocating a difference conversion coefficient to a corresponding frame target image according to the first difference amount, the second difference amount, the first difference distance, and the second difference distance; According to the number of repetitions and the difference conversion coefficient, a total conversion coefficient corresponding to the repetition symbol is obtained; All total conversion coefficients are summed to obtain the overall conversion coefficients of the corresponding continuous video segments; Extracting a maximum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a first conversion coefficient, and at the same time, extracting a minimum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a second conversion coefficient; Based on the overall conversion coefficient, respectively, the first conversion coefficient and the second conversion coefficient are optimized and adjusted to obtain a maximum operating capacity and a minimum operating capacity; The maximum similarity value is the maximum value of all final similarity values ​​under the same continuous video segment, and the minimum similarity value is the minimum value of all final similarity values ​​under the same continuous video segment; The similarity analysis is performed on adjacent frame images in sequence to set a similarity symbol to each frame target image, including: Overlapping a first frame image and a second frame image in adjacent frame images, and obtaining non-overlapping pixel points of the first frame image and non-overlapping pixel points of the second frame image, wherein the first frame image is located in front of the second frame image; Compare the brightness values ​​of non-overlapping pixels of the first frame image and non-overlapping pixels of the second frame image at the same position, and obtain the brightness difference of each position respectively; Determine the initial similarity value of the adjacent frame images according to a first ratio of the number of overlapping pixels to the total number of pixels of the frame image and a second ratio of the number of brightness differences that meet the difference range standard to the number of non-overlapping pixels; Respectively acquiring picture description information of non-overlapping pixel points of the first frame image and picture description information of non-overlapping pixel points of the second frame image to obtain adjustment similarity coefficients of the adjacent frame images; Determine a final similarity value based on the initial similarity value and the adjusted similarity coefficient and in combination with the variance under all brightness differences; Obtaining a similarity symbol matching the final similarity value from a value-symbol comparison table; The overlapping process refers to aligning the first frame image and the second frame image at the same position point. If the information of two pixels at the position point is completely consistent, they are regarded as overlapping points, otherwise they are regarded as non-overlapping points.

2. The cloud platform data processing method according to claim 1, characterized in that: Determine continuous video segments, including: If the first frame image is the first frame image of the target video, assigning two identical similar symbols to the first frame image; If the first frame image is not the first frame image of the target video, assigning a similar symbol under which the first frame image is the first image in adjacent frame images and a similar symbol under which the first frame image is the second image in adjacent frame images to the first frame image; If the second frame image is the last frame image of the target video, assigning two identical similar symbols to the second frame image; If the second frame image is not the last frame image of the target video, assigning a similar symbol under which the second frame image is the first image in the adjacent frame images and a similar symbol under which the second frame image is the second image in the adjacent frame images to the second frame image; According to the time sequence and the consistency of the symbol combination of each frame of the target image, the preliminary division points are determined to obtain a number of continuous video segments, and each continuous video segment contains at least one frame of the target image.

3. The cloud platform data processing method according to claim 1, characterized in that: Obtain the optimal operational capabilities of the cloud platform, including: Obtaining first operation information of the cloud platform in each historical period, wherein the first operation information includes the duration of processing of each received historical video by the cloud platform and the amount of resources called at each processing moment; Extracting the maximum operation processing volume, the second largest operation processing volume, and the duration period under the maximum operation processing volume and the second largest operation processing volume from the first operation information to obtain the historical processing capacity based on the corresponding historical period; The historical processing capacity is plotted according to the chronological order of historical time periods to obtain a capacity curve; Obtaining a reference curve according to the position of the current operating capacity in the capacity curve; Performing a first adjustment on the reference curve according to the curvature of each convex point on the reference curve to obtain a first adjustment curve; At the same time, performing a second adjustment on the reference curve according to the curvature of each concave point on the reference curve to obtain a second adjustment curve; Based on a first attenuation area under the first adjustment curve and a second attenuation area under the second adjustment curve, and in combination with a closed area under a reference curve, a first adjustment coefficient for maximum operating capacity and a second adjustment coefficient for minimum operating capacity are obtained, thereby obtaining an optimal operating capacity; The running processing capacity is determined by the total amount of resources called at each moment; The closed area of ​​the reference curve minus the closed area of ​​the first adjustment curve = the first attenuation area; The closed area of ​​the second adjustment curve minus the closed area of ​​the reference curve=the second attenuation area.

4. The cloud platform data processing method according to claim 1, characterized in that: Get N split results, including: Taking the maximum operating capacity required for each continuous video segment as a benchmark and the optimal operating capacity as a boundary, constructing a capacity constraint condition and obtaining a number of split points; A plurality of continuous video segments are synchronously split according to the splitting points to obtain N splitting results.

5. The cloud platform data processing method according to claim 1, characterized in that: The compression method of the first frame image in each continuous video segment is determined according to the video quality requirements of the target video, including: Determine the image information amount of the first frame image in each continuous video segment respectively, and obtain the optimal video requirement of the corresponding first frame image in combination with the maximum operation capacity of the corresponding continuous video segment; Determining the effective video quality of the corresponding first frame image according to the video quality requirement and the optimal video requirement; A compression method matching the effective video quality is obtained from a quality-compression comparison table.

6. A cloud platform data processing device, characterized in that: The device is used to apply the cloud platform data processing method according to any one of claims 1 to 5, comprising: A preliminary data acquisition module, used to acquire the current operating capability of the cloud platform and the video quality requirements for the target video, wherein the target video includes multiple frames of target images; A symbol setting module is used to perform similarity analysis on adjacent frame images in sequence, set similar symbols to each frame of target image, and determine continuous video segments; A capability determination module is used to construct a pixel matrix of the first frame image under each continuous video segment, determine the current specification quantity of the corresponding first frame image, and determine the maximum operating capacity and the minimum operating capacity required for the corresponding continuous video segment according to the number of frames contained in the corresponding continuous video segment and the symbol difference of similar symbols under the corresponding continuous video segment, wherein the current specification quantity is obtained based on the cumulative sum of the information content of the pixel points; A comprehensive analysis module, used to comprehensively analyze all maximum operating capabilities and minimum operating capabilities under the target video according to the current operating capabilities of the cloud platform, to obtain the optimal operating capability of the cloud platform; A compression method determination module is used to synchronously split a plurality of continuous video segments to obtain N split results based on the optimal operation capability, and to match one encoding channel to each split result, and to determine a compression method for the first frame image in each continuous video segment according to the video quality requirements of the target video; The encoding module is used to compress and encode the first frame image in the same continuous video segment under the corresponding encoding channel in N encoding channels in turn according to the corresponding compression mode, and to call and encode the remaining frame images in the same continuous video segment, so as to obtain the compression result based on the corresponding encoding channel; The calling code refers to the calling and using of the compression result of the first frame image, and the dedicated code refers to the code after the remaining frame images are compressed with the inconsistent image information in the first frame image, and in the subsequent decoding process, the calling code and the dedicated code can effectively restore the remaining frame images; The maximum operating capacity and the minimum operating capacity refer to the capacity range required by the cloud platform when processing the continuous video segment. The capability determination module is used to: Determining a first difference quantity and a first difference distance according to a symbol combination of each frame of the target image in the corresponding continuous video segment, wherein the symbol combination of each frame of the target image is composed of similar symbols of each frame of the target image; Locking a first symbol corresponding to a maximum similarity value and a second symbol corresponding to a minimum similarity value under corresponding continuous video segments, and determining a second difference quantity and a second difference distance between the first symbol and the second symbol; At the same time, the number of repetitions of each similar symbol in the corresponding continuous video segments is obtained; Allocating a difference conversion coefficient to a corresponding frame target image according to the first difference amount, the second difference amount, the first difference distance, and the second difference distance; According to the number of repetitions and the difference conversion coefficient, a total conversion coefficient corresponding to the repetition symbol is obtained; All total conversion coefficients are summed to obtain the overall conversion coefficients of the corresponding continuous video segments; Extracting a maximum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a first conversion coefficient, and at the same time, extracting a minimum conversion coefficient from all difference conversion coefficients and combining it with a corresponding number of frames to obtain a second conversion coefficient; Based on the overall conversion coefficient, respectively, the first conversion coefficient and the second conversion coefficient are optimized and adjusted to obtain a maximum operating capacity and a minimum operating capacity; The maximum similarity value is the maximum value of all final similarity values ​​under the same continuous video segment, and the minimum similarity value is the minimum value of all final similarity values ​​under the same continuous video segment; Among them, the symbol setting module is used to: Overlapping a first frame image and a second frame image in adjacent frame images, and obtaining non-overlapping pixel points of the first frame image and non-overlapping pixel points of the second frame image, wherein the first frame image is located in front of the second frame image; Compare the brightness values ​​of non-overlapping pixels of the first frame image and non-overlapping pixels of the second frame image at the same position, and obtain the brightness difference of each position respectively; Determine the initial similarity value of the adjacent frame images according to a first ratio of the number of overlapping pixels to the total number of pixels of the frame image and a second ratio of the number of brightness differences that meet the difference range standard to the number of non-overlapping pixels; Respectively acquiring picture description information of non-overlapping pixel points of the first frame image and picture description information of non-overlapping pixel points of the second frame image to obtain adjustment similarity coefficients of the adjacent frame images; Determine a final similarity value based on the initial similarity value and the adjusted similarity coefficient and in combination with the variance under all brightness differences; Obtaining a similarity symbol matching the final similarity value from a value-symbol comparison table; The overlapping process refers to aligning the first frame image and the second frame image at the same position point. If the information of two pixels at the position point is completely consistent, they are regarded as overlapping points, otherwise they are regarded as non-overlapping points.

7. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 5.

8. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method according to any one of claims 1 to 5.