Video and audio file migration integrity verification method

Abstract

The invention relates to the technical field of integrity verification, in particular to a video and audio file migration integrity verification method, which comprises the following steps: acquiring a synchronous frame extraction pixel sequence, establishing mapping to generate feature data, extracting time, marking a stable section, extracting brightness residual error, judging the direction to be consistent, and analyzing chromaticity variance to generate the stable section. According to the method and the device, the brightness and chroma pixel value sequences are extracted from the synchronous frames in the program packaging file and the broadcast rendering frame, and the mapping relation is established, so that frame-by-frame accurate comparison can be realized, the synchronous state is identified in combination with the time sequence difference trend, the inter-frame comparison precision is improved, and the accuracy of the inter-frame comparison is improved. Through the conjoint analysis of the brightness channel residual direction and the chrominance channel variance distribution, the image anomaly recognition capability is enhanced, and the frame-level consistency result is extracted based on the multi-dimensional residual feature intersection, so that the video and audio integrity verification without depending on a unified decoder is realized, and the real-time performance and fine granularity of quality assurance are enhanced.

Description

Technical Field

[0001] This invention relates to the field of integrity verification technology, and in particular to a method for verifying the integrity of video and audio file migration. Background Technology

[0002] The field of integrity verification technology mainly involves technical means to verify the correctness and consistency of data during storage, transmission, or processing. Core aspects include data content consistency detection, error identification, and integrity assurance during transmission or migration. Methods such as constructing hash values, checksums, and comparison mechanisms are used to determine whether file data has been tampered with, missing, or corrupted, ensuring the reliability of data throughout the entire processing chain from source to destination. Traditional audio-visual file migration integrity verification methods refer to generating a single MD5 value for the entire file during the transmission of program files to the broadcast system or storage medium. The method of verification involves reading the complete file, generating a digest value, and comparing it with the source file to determine whether the file has changed during migration. However, since this method can only be executed after the file is fully received, it cannot be segmented for speed-up, resulting in low efficiency when dealing with emergency broadcast scenarios. Furthermore, in terms of file quality detection, it relies on software-based parsing methods, making it difficult to detect decoding anomalies such as black screens and stuttering caused by differences in decoders used by the broadcast server and the file technical review system. Traditional methods rely on manual monitoring and post-event sampling to assist in verifying file quality, but they are difficult to achieve real-time and comprehensive integrity assurance.

[0003] Verification methods based on full-file MD5 digest comparison can only be performed after the file is fully received, lacking adaptability to real-time processing scenarios and failing to meet the timeliness requirements of synchronous verification in rapid broadcast environments. As a global data identifier, the digest value cannot reflect whether there are audiovisual anomalies at the micro-level, and cannot provide an effective detection mechanism for issues such as segmented stuttering and localized black screens. Existing technologies rely on comparison after complete file decoding, lacking frame-by-frame verification capabilities, making it difficult to detect playback-side rendering anomalies when decoder differences exist. For example, a program file may preview normally through the technical review system, but issues such as black screens and audio-visual asynchrony on the broadcast server due to different decoder implementations cannot be identified before migration. Furthermore, quality issues require manual monitoring and post-event sampling, failing to achieve full-time, full-content detection coverage, increasing labor costs, and posing a risk of missed detections, impacting broadcast security and content assurance capabilities. Existing methods lack the ability to perform pixel-level residual analysis and time synchronization sequence modeling, making it difficult to effectively correlate the dynamic consistency between program content and broadcast performance, thus limiting the specificity and reliability of integrity verification results. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a method for verifying the integrity of video and audio file migration, comprising the following steps:

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for verifying the integrity of video and audio file migration, comprising the following steps:

[0006] S1: Obtain the synchronized frames in the program's MXF file and the decoder rendering frames of the broadcast server, extract the pixel value sequences of the luminance and chrominance channels, establish the corresponding pixel position mapping, and generate frame pixel feature comparison data.

[0007] S2: Based on the frame numbering relationship in the frame pixel feature comparison data, extract the timestamp and baseband signal sampling time from the program MXF file, construct two sets of time series, analyze the time difference change trend, mark the numbering segment with stable difference, and generate a synchronized aligned frame numbering sequence.

[0008] S3: Extract the luminance channel difference sequence from the synchronized aligned frame number sequence, analyze the trend of inter-frame difference changes, determine the similarity of residual directions, record the frame numbers that meet the conditions, and generate a set of residual direction matching frame segment numbers.

[0009] S4: Extract the corresponding chroma channel residual sequence from the residual direction matching frame segment number set, judge the continuous inter-frame variance change based on the intra-frame residual distribution characteristics, mark the frame segments whose change range meets the requirements, and generate a chroma stable frame segment number set.

[0010] S5: Take the intersection of the residual direction matching frame segment number set and the chroma stable frame segment number set, extract the consistent frame set, mark the comparison frame status, and generate a frame-level comparison consistency tag set.

[0011] As a further embodiment of the present invention, the frame pixel feature comparison data includes a luminance channel pixel value sequence, a chrominance channel pixel value sequence, and a pixel position mapping relationship; the synchronization alignment frame number sequence includes a program MXF encapsulation file frame number, a baseband signal acquisition frame number, and an inter-frame difference change trend marker; the residual direction matching frame segment number set includes a luminance channel residual value sequence, continuous inter-frame difference trend features, and a residual direction consistency marker; the chrominance stable frame segment number set includes a chrominance channel residual value sequence, intra-frame residual distribution features, and an inter-frame variance change range marker; and the frame-level comparison consistency label set includes an intersection frame number set, a comparison status label, and a frame-level consistency result.

[0012] As a further aspect of the present invention, the specific steps of S1 are as follows:

[0013] S101: Obtain the synchronized frame image data in the program's MXF encapsulation file and the broadcast server's rendered frame, extract the luminance channel value and chrominance channel value of each pixel in each frame image, form a luminance sequence and chrominance channel sequence according to the pixel position coordinates, and generate a luminance pixel sequence and chrominance channel sequence.

[0014] S102: Based on the brightness pixel sequence and the chroma channel sequence, call the coordinate information of each pixel, establish a correspondence between the brightness value and the two chroma channel values ​​according to the coordinates, and index and organize the mapping content to generate brightness and chroma mapping index pairs;

[0015] S103: Based on the brightness and chromaticity mapping index pairs, integrate the brightness values ​​and chromaticity channel values ​​at the same position coordinates, construct the feature combination structure of pixels in the frame image, output the comparison matrix between pixels, and generate frame pixel feature comparison data.

[0016] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0017] S201: Based on the frame number correspondence in the frame pixel feature comparison data, extract the timestamp associated with each frame number in the program MXF encapsulation file, and synchronously obtain the sampling time in the baseband signal acquisition frame image data. Arrange them according to the frame number to form two sets of time series, and generate a time stamp sequence group.

[0018] S202: Based on each pair of timestamps and sampling time values ​​in the time stamp sequence group, calculate the difference in sequence and arrange them in the order of frame number. Judge the trend of change of all time difference data, and divide the time difference sequence into segments according to the degree of change to obtain the stable segment of time difference.

[0019] S203: For the frame number range included in the stable time difference segment, filter the frame number data in consecutive numbering order, uniformly mark them as the corresponding numbers of the synchronization state, establish a frame number sequence structure, and generate a synchronization aligned frame number sequence.

[0020] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0021] S301: Select each group of frames in the synchronized alignment frame number sequence, extract the brightness channel data in the corresponding images of adjacent frames, calculate the gray value difference of the corresponding pixels of the brightness channel of each group of adjacent frames, arrange them in the order of frame number, establish the inter-frame brightness change sequence, and generate the inter-frame brightness difference sequence.

[0022] S302: Based on the direction of change of the difference between adjacent frames in the inter-frame brightness difference sequence, compare the positive and negative changes of the brightness difference in several consecutive frames in turn, identify the frame pair combination interval with the same direction of change, record all frame numbers in the interval, and obtain the residual direction consistency interval.

[0023] S303: For each number combination in the residual direction consistency interval, call the frame number data, remove the number segments whose direction change mutation number exceeds the specified direction consistency threshold in the order of consecutive frames, retain only the frame segment numbers that meet the consistency conditions, establish the number set, and generate the residual direction matching frame segment number set.

[0024] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0025] S401: Extract the number of each frame from the set of residual direction matching frame segments, obtain the image residual data of the corresponding frame in the chroma channel, extract the residual values ​​of all pixels in the chroma channel of each frame, and combine them into a sequence according to the frame number order to generate a chroma residual value sequence.

[0026] S402: Based on the pixel residual value corresponding to each frame in the chroma residual value sequence, calculate the chroma residual variance value of each frame in sequence, and then perform difference operation on the variance values ​​of adjacent frames to obtain the sequence of chroma residual variance variation amplitude values ​​between consecutive frames, and obtain the chroma variance variation amplitude value between frames.

[0027] S403: Based on the change amplitude value corresponding to each frame segment in the inter-frame chromaticity variance change amplitude value, compare all amplitude values ​​in the interval segment by segment to see if they are within the specified chromaticity stability threshold range, mark the frame segment number that meets the range condition, and combine the corresponding numbers to form a set to obtain the chromaticity stable frame segment number set.

[0028] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0029] S501: Based on the residual direction matching frame segment number set and the frame number of each frame in the chroma stable frame segment number set, perform a matching operation between the numbers, extract the frame numbers that exist in both sets at the same time, and obtain a set of frames with consistent numbers.

[0030] S502: Call the set of consistent numbered frames, determine whether there is an intersection based on the number correspondence between the set and the numbered set of the residual direction matching frame segments, mark the corresponding number status, and obtain the frame number comparison status sequence;

[0031] S503: Based on the frame number comparison status value of each frame in the status sequence, construct a tag matrix with consistent numbering order, summarize the tag status of all frames, and generate a frame-level comparison consistency tag set.

[0032] As a further aspect of the present invention, the program MXF file refers to a program file in the MaterialExchangeFormat format that conforms to broadcast industry standards. It encapsulates audio and video frame data arranged in chronological order and associated frame-level description information. The time information of the frames exists in the form of PresentationTimeStamp.

[0033] The broadcast server decoder rendering frame refers to a single frame of image data generated by the player or hardware decoding module inside the broadcast system after decoding the input program file. It is an image signal used to directly output video frames to the broadcast link.

[0034] The synchronization frame refers to the frame pair whose content has a corresponding relationship after a mapping relationship is established between the file end frame number and the broadcast end decoded frame number during the migration integrity comparison process;

[0035] The frame pixel feature comparison data refers to the data information set generated by establishing a mapping between the frame image data at the program file end and the frame image data at the broadcast server end at the same pixel position, including the correspondence between the luminance channel value, the chrominance channel value, and the frame number.

[0036] As a further aspect of the present invention, the synchronized alignment frame number sequence refers to the set of frame numbers with stable time fluctuations selected by analyzing the time difference change trend between the file end timestamp sequence and the acquisition end sampling time sequence.

[0037] The luminance channel difference sequence refers to the sequence of differences calculated between the corresponding pixel values ​​in the luminance channel of the source file frame and the broadcast frame;

[0038] The similarity of residual directions refers to the similarity of the directions of the residual vectors of the brightness channels in adjacent frames. The judgment is based on whether the angular relationship between the directions of the residual vectors is within an acceptable range, reflecting the consistency of the trend of image changes.

[0039] The residual direction matching frame segment number set refers to a set of multiple frame numbers that are determined to have a continuous and abrupt change trend after the residual direction consistency analysis of the luminance channel.

[0040] As a further aspect of the present invention, the chroma channel residual sequence refers to the set of chroma value differences of each corresponding pixel in the U and V chroma channels of the frame pair, reflecting the changes in color information between the two frames.

[0041] The variance variation refers to the degree of difference in the statistical distribution of the contrast chroma channel residual values ​​within or between frames;

[0042] The set of chroma-stable frame segments refers to the set of frame segments selected after evaluation of the variance change of the chroma channel residuals. The frame pairs in the set have the characteristics of stable color changes and strong regularity of residual fluctuations.

[0043] The number intersection processing refers to extracting the frame numbers that are common to the two number sets as common elements;

[0044] The frame-level alignment consistency tag set refers to the set after all the comparison frames are marked according to the results of the brightness direction consistency and color stability judgment. Each frame number corresponds to a tag value.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0046] In this invention, by extracting the luminance and chrominance pixel value sequences from the program encapsulation file and the synchronous frames in the broadcast rendering frame and establishing a mapping relationship, accurate frame-by-frame comparison can be achieved. By combining the time series difference trend to identify the synchronization status, the accuracy of inter-frame comparison is improved. Through joint analysis of the residual direction of the luminance channel and the variance distribution of the chrominance channel, the ability to identify image anomalies is enhanced. Based on the intersection of multi-dimensional residual features, frame-level consistency results are extracted, enabling audio and video integrity verification without relying on a unified decoder, and strengthening the real-time performance and fine granularity of quality assurance. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the steps of the present invention;

[0049] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0050] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0051] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0052] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0053] Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0054] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0055] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0056] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0057] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0058] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0059] Please see Figure 1 This invention provides a method for verifying the integrity of video and audio file migration, comprising the following steps:

[0060] S1: Obtain the synchronization frames in the MXF encapsulation file of the program and the rendering frames of the built-in decoder of the broadcast server, extract the pixel value sequences of the luminance and chrominance channels, establish a one-to-one mapping relationship at the corresponding pixel positions, and generate frame pixel feature comparison data.

[0061] The MXF encapsulated file of a program refers to a program file in the MaterialExchangeFormat format that conforms to broadcast industry standards. It encapsulates audio and video frame data arranged in chronological order and associated frame-level description information. The time information of the frames exists in the form of PresentationTimeStamp, which is used to identify the playback time order of each frame.

[0062] The built-in decoder rendering frame of the broadcast server refers to the single frame image data generated after the input program file is decoded by the player or hardware decoding module inside the broadcast system. It is a video frame that is used to directly output the image signal to the broadcast link, including the visual content corresponding to the source file but with anticipated format or timing differences.

[0063] Synchronization frames refer to the frames whose content corresponds to each other after a mapping relationship is established between the file end frame number and the broadcast end decoded frame number during the migration integrity comparison process. They are used as the starting data units for image and time verification.

[0064] Frame pixel feature comparison data refers to the data information set generated after the frame image data at the program file end and the frame image data at the broadcast server end are mapped to the same pixel position. It includes the correspondence between the luminance channel value, chrominance channel value and frame number, and is used for feature comparison and residual calculation.

[0065] S2: Based on the frame number correspondence in the frame pixel feature comparison data, extract the timestamp in the program MXF encapsulation file and the sampling time in the baseband signal acquisition frame image data, construct two sets of time series, determine the synchronization status according to the change trend of the two sets of time difference, mark the number segment with stable inter-frame difference change, and generate a synchronization aligned frame number sequence.

[0066] Baseband signal acquisition frame image data refers to the original signal frame images acquired by the video acquisition device during the broadcast process. The images originate from the baseband video signal output by the broadcast server to the monitoring equipment, and are obtained after SDI or HDMI signal conversion, serving as the actual physical output representation of the decoding result at the broadcast end.

[0067] Synchronous alignment frame number sequence refers to the set of frame numbers with stable time fluctuations selected by analyzing the time difference between the timestamp sequence at the file end and the sampling time sequence at the acquisition end, which is used to determine the valid frame pairs under time alignment.

[0068] S3: Select each group of frames in the synchronized alignment frame number sequence, extract the pixel difference sequence of the brightness channel, judge the similarity of the residual direction based on the difference change trend between consecutive frames, record the frame number that meets the condition, and generate a set of residual direction matching frame segment numbers.

[0069] The luminance channel pixel difference sequence refers to the sequence of differences calculated between corresponding pixel values ​​in the luminance channel (Y component) of the source file frame and the broadcast frame. It is used to describe the distribution of differences in visual luminance performance between frame pairs.

[0070] The similarity of residual directions refers to the degree of similarity of the directions of the residual vectors of the brightness channels in adjacent frames. The judgment is based on whether the angular relationship between the directions of the residual vectors is within an acceptable range, reflecting the consistency of the trend of image changes.

[0071] The residual direction matching frame segment number set refers to a set of multiple frame numbers that are determined to have a continuous and abrupt trend of directional change after the brightness channel residual direction consistency analysis. It is used to characterize image segments that maintain stable visual consistency.

[0072] S4: Extract frame numbers from the residual direction matching frame segment number set, obtain the residual value sequence of the corresponding chroma channel, judge the variance change between consecutive frames based on the distribution characteristics of the residual values ​​within the frame, mark the frame segments whose change range meets the requirements, and generate a chroma stable frame segment number set.

[0073] The chroma channel residual value sequence refers to the set of chroma value differences of each corresponding pixel in the U and V chroma channels of a frame pair. It reflects the changes in color information between two frames and serves as a key basis for detail consistency verification.

[0074] Variance variation refers to the degree of difference in the statistical distribution of the contrast chroma channel residual values ​​within or between frames. This is used to determine whether there are drastic fluctuations or abnormal interference at the color level.

[0075] The chroma-stable frame segment number set refers to the set of frame segment numbers selected after evaluating the variation of chroma channel residual variance. The frame pairs in the set have the characteristics of stable color changes and strong regularity of residual fluctuation.

[0076] S5: Perform numbering intersection processing on the residual direction matching frame segment number set and the color stable frame segment number set, extract the frame set with consistent comparison results, mark the comparison status of all frames, and generate a frame-level comparison consistency tag set.

[0077] Number intersection processing refers to extracting the common frame numbers of two number sets as common elements, which are used to filter the frame set that passes the verification in multiple feature dimensions.

[0078] The frame-level alignment consistency tag set refers to the set of all comparison frames marked according to the results of the luminance direction consistency and chromaticity stability judgment. Each frame number corresponds to a tag value, which is used to indicate whether it is judged to be consistent in the integrity check.

[0079] The frame pixel feature comparison data includes the luminance channel pixel value sequence, the chrominance channel pixel value sequence, and the pixel position mapping relationship. The synchronization alignment frame number sequence includes the program MXF encapsulation file frame number, the baseband signal acquisition frame number, and the inter-frame difference change trend marker. The residual direction matching frame segment number set includes the luminance channel residual value sequence, continuous inter-frame difference trend features, and residual direction consistency marker. The chrominance stable frame segment number set includes the chrominance channel residual value sequence, intra-frame residual distribution features, and inter-frame variance change range marker. The frame-level comparison consistency label set includes the intersection frame number set, comparison status label, and frame-level consistency result.

[0080] Please see Figure 2 The specific steps of S1 are as follows:

[0081] S101: Obtain the synchronized frame image data in the program's MXF encapsulation file and the broadcast server's rendered frame, extract the luminance channel value and chrominance channel value of each pixel in each frame image, form a luminance sequence and chrominance channel sequence according to the pixel position coordinates, and generate a luminance pixel sequence and chrominance channel sequence.

[0082] When extracting synchronized frame image data from the program's MXF encapsulation file and the broadcast server, image parsing tools such as FFmpeg are first used to extract the corresponding frame sequence images, ensuring that the timecodes or frame numbers are perfectly aligned, thus selecting image content at the same time node. During image data processing, the image's color data is converted from the original RGB format to luminance and chrominance components, i.e., the Y, Cb, and Cr channels. Each frame image traverses all pixel positions; for example, at a resolution of 1920×1080, there are 2,073,600 pixels. Each pixel's horizontal and vertical coordinates are obtained, such as (0, 0), (1, 0)...(1919, 1079), and the luminance and two chrominance values ​​at the corresponding positions are calculated according to color space transformation rules. In the generated data, the Y, Cb, and Cr values ​​of all pixels in each frame are recorded sequentially according to frame order, and combined with the coordinates to form a sequence structure of luminance and chrominance. If some frames are found to be missing during the extraction process, the missing frames are supplemented using time interpolation methods according to frame order, for example, by using data from the preceding and following valid frames for linear estimation. For video content at 60 frames per second, the luminance and chrominance component values ​​of a total of 124 million pixels in the 60 frames can be obtained after completion, forming a sequence data set arranged by pixel position.

[0083] S102: Based on the luminance pixel sequence and chrominance channel sequence, call the coordinate information of each pixel, establish a correspondence between the luminance value and the two chrominance channel values ​​according to the coordinates, and index and organize the mapping content to generate luminance and chrominance mapping index pairs;

[0084] Based on the sequential data of luminance and chrominance, the coordinates of each pixel are used as an index reference for data organization and mapping. During processing, the two-dimensional coordinates of all pixels in each frame are first read, determining their horizontal position from 0 to the image width minus 1, and their vertical position from 0 to the image height minus 1. Taking an image size of 1920×1080 as an example, the horizontal and vertical coordinate ranges are 0 to 1919 and 0 to 1079, respectively. Each coordinate position is combined with its corresponding luminance value and two chrominance values ​​to form a data item, which is then linked and bound using a two-dimensional array or hash structure, forming a data structure with pixel coordinates as keys and luminance and chrominance channel values ​​as content. Subsequently, a one-dimensional index is built for each pixel in a row-major order using linear numbering to facilitate rapid location during processing. This index numbering follows a left-to-right, top-to-bottom order; for example, the first row is 0 to 1919, the second row is 1920 to 3839, and so on. Each index position stores a three-channel data combination, meaning each pixel stores a set of three values ​​representing the current values ​​of Y, Cb, and Cr. For example, pixel number 500000 might be located in row 261, column 800, with a luminance value of 118 and chroma channel values ​​of 132 and 140. After all the data is integrated, an index structure with a length equal to the number of frames multiplied by the total number of pixels is formed, facilitating subsequent retrieval and calculation processing.

[0085] S103: Based on the luminance and chrominance mapping index pairs, integrate the luminance values ​​and chrominance channel values ​​at the same position coordinates, construct the feature combination structure of pixels in the frame image, output the comparison matrix between pixels, and generate frame pixel feature comparison data.

[0086] Using a luminance-chrominance mapping index structure, pixel channel values ​​at the same coordinates are combined. In each frame, the three-channel values ​​of all pixels are read, combined into a vector data structure, and stored using its position in the linear index as the index key. For each pixel, such as position 100000, its three channel values ​​in a frame are combined into a vector. After processing each frame, the channel combination vectors of all pixels are categorized and grouped according to frame number, forming a feature vector matrix per frame. For ease of comparison, the data of pixels with the same number in all frames are combined into a continuous time-series vector set, with each group containing the number of vector values ​​from each frame. For example, with 60 frames, each pixel will have 60 sets of luminance and chrominance three-channel combinations. For instance, the pixel at position 500,000 will have a set of three-channel values ​​from frame 1 to frame 60, such as 120, 130, and 140 in frame 1, 121, 132, and 138 in frame 2, and so on, forming a continuous comparison sequence. When processing large amounts of frame data, a compressed storage structure is needed to reduce memory pressure. For example, using a sparse matrix compressed storage method will only record the actual data items, avoiding duplication and redundant information. The feature vectors of all frames are combined to form a complete inter-frame comparison matrix, constituting a multi-frame feature data structure at the pixel level of the frame image.

[0087] Please see Figure 3 The specific steps of S2 are as follows:

[0088] S201: Based on the frame number correspondence in the frame pixel feature comparison data, extract the timestamp associated with each frame number in the program MXF encapsulation file, and synchronously obtain the sampling time in the baseband signal acquisition frame image data. Arrange them according to the frame number to form two sets of time series, and generate a time stamp sequence group.

[0089] Based on the frame number correspondence in the frame pixel feature comparison data, the image frame timestamps are read from the MXF encapsulation file. The frame parsing tool is called to extract the time code field from the metadata of each frame image. The result is recorded as a time series array, with the time unit uniformly in ms. For example, the frame numbers from 1 to 600 correspond to timestamps of 1000ms, 1016ms, 1033ms, etc., ensuring that the interval matches the set frame rate. At the same time, the sampling time information is synchronously obtained from the logs or metadata of the baseband image acquisition device. The sampling time is automatically recorded according to the acquisition clock, which also forms a sampling time series arranged in the order of frame numbers. The two time series correspond one-to-one in data structure and have the same length. Each item is bound to a unique frame number. After combination, a double time series group is generated, which represents the time mark of each frame image in the encapsulation file and in the actual sampling device. It is suitable for content processing tasks with a frame rate of 60 frames per second. When the sequence length reaches 600, it indicates a program segment with a processing duration of 10 seconds. After processing, two sets of time mark sequences with the same frame number dimension are formed.

[0090] S202: Based on each pair of timestamps and sampling time values ​​in the time stamp sequence group, calculate the difference in sequence and arrange them in the order of frame number. Determine the trend of change of all time difference data, and divide the time difference sequence into segments according to the degree of change to obtain the stable segment of time difference.

[0091] Based on two sets of time-stamped sequences, the time difference for each frame is calculated. The process starts with frame number 1, sequentially reading the MXF time value and the sampling time value, performing difference calculations, and storing the difference results sequentially into a new time difference array. This array has the same length as the original sequence, and the unit of measurement is milliseconds (ms). The difference value may be negative or positive, depending on the order of the encapsulation file and the sampling time. After calculation, the trend judgment stage begins. A sliding analysis window size is set, such as 10 frames, and the difference between the maximum and minimum differences within each window is calculated. If the difference is less than a specified stable value... A fixed threshold is marked as a stable window. The stable threshold can be 5ms or 10ms, depending on the specific device's synchronization error tolerance. The sliding window slides gradually backward from the first frame, repeating the judgment each time it slides one frame. If the difference between frame numbers 50 and 59 is between -3ms and -6ms, with a fluctuation range of 3ms, which is less than the 5ms threshold, the segment can be determined as a stable difference area. Repeated sliding judgment can obtain multiple stable segments with consecutive frame numbers, such as frame numbers 120 to 240, 301 to 410, etc. These segments are summarized into a set of stable time difference segments.

[0092] S203: For the frame number range included in the stable time difference segment, filter the frame number data in consecutive number order, uniformly mark it as the number corresponding to the synchronization state, establish a frame number sequence structure, and generate a synchronization aligned frame number sequence.

[0093] Extract the frame number range from all stable time difference segments. Perform a number continuity check on each segment. Starting from the segment's starting number, check sequentially whether the numbers are strictly increasing and without repeated skips. For example, after number 120, there should be 121, continuing sequentially to 240. If there are non-continuous numbers in the middle, such as skipping to 123, the segment needs to be removed or reorganized, retaining only the continuous and complete parts as valid frame segments. Construct a frame number sequence structure for each number set that passes the continuity check. The structure includes the starting number, ending number, and total number of frames for the segment. For segments numbered 120 to 240, the sequence length is 121 frames, and the number range is complete and continuous. Mark this segment as a set of synchronized state frames. If there are multiple stable segments, construct multiple frame number sequence structures for each segment and summarize them into a synchronized aligned frame number sequence with a total length equal to the sum of the number of frames in all stable segments. The output structure can be used for subsequent matching processing or image content synchronization analysis.

[0094] Please see Figure 4 The specific steps of S3 are as follows:

[0095] S301: Select each group of frames in the synchronized alignment frame number sequence, extract the brightness channel data in the corresponding images of adjacent frames, calculate the gray value difference of the corresponding pixels in the brightness channel of each group of adjacent frames, arrange them in the order of frame number, establish the inter-frame brightness change sequence, and generate the inter-frame brightness difference sequence.

[0096] Select each group of frames in the synchronized alignment frame number sequence. For each group of two adjacent frames, extract pixel grayscale data from the image's luminance channel. If the image format is YUV420, select the Y channel data as the luminance reference. Traverse the image's two-dimensional matrix and calculate the difference between pixels at the same position in the first and second frames. Each element in the resulting difference matrix represents the luminance variation at that position. The grayscale value range is set between 0 and 255. The difference for each pixel is calculated by subtracting the grayscale value of the second frame from the grayscale value of the first frame. For example, if the pixel at row 100, column 200 of the first frame is 120, the pixel at that position in the second frame... If the value is 130, the corresponding difference is -10. After traversing all pixels of the image, these difference data are statistically analyzed into a single-frame brightness difference vector. After sorting by frame number, multiple inter-frame difference sequences are formed. If the total number of frames is 600, the number of sequences is 599. Each item represents the brightness change between frame i and frame i+1. Each sequence is an aggregated representation of the overall brightness difference of the image. This process requires maintaining consistent pixel positions and image size alignment. Otherwise, image size normalization processing is required first, for example, unifying the image size to 1920×1080. Pixel-level difference generation is achieved through traversal, establishing inter-frame brightness change sequences, and generating inter-frame brightness difference sequences.

[0097] S302: Based on the direction of change of the difference between adjacent frames in the inter-frame brightness difference sequence, compare the positive and negative changes of the brightness difference in several consecutive frames in turn, identify the frame pair combination interval with the same direction of change, record all frame numbers in the interval, and obtain the residual direction consistency interval.

[0098] Based on the inter-frame luminance difference sequence, the positive and negative change directions of each difference sequence are analyzed. First, the luminance difference between the first and second frames in the difference sequence is read, and its positive and negative directions are determined. If the difference is positive, it is defined as an increase; if it is negative, it is a decrease; and if it is zero, it is defined as unchanged. This process continues, reading the difference between each subsequent frame and comparing their directions. Frame number segments with the same direction are counted. When a direction reversal occurs, the start and end frame numbers of the previous consecutive segment are recorded. For example, if the difference between frames 1 to 6 is positive, then numbers 1 to 6 are recorded as segments with the same direction. If the difference between frames 7 and 6 is negative, then the segment with the same direction is re-evaluated starting from frame 7. This evaluation process... Based on the sign change pattern of the brightness difference sequence, a sign extraction function and continuous segment search logic are implemented. Specifically, a state flag variable is set. If the current difference is consistent with the previous direction, the count is incremented; otherwise, the current segment is settled and a new segment is started. Each segment with consistent frame number is output, such as frames 1 to 6 being positive segments, frames 7 to 13 being negative segments, etc. If the total number of frames in the entire sequence is 599, there can be a maximum of 599 directional change points. However, by comparing the actual difference change direction, it can be compressed into fewer continuous consistent segments. The start and end positions of the frame number of the segment are recorded as the residual direction consistency interval.

[0099] S303: For each number combination in the residual direction consistency interval, call the frame number data, remove the number segments whose direction change mutation number exceeds the specified direction consistency threshold in the order of consecutive frames, retain only the frame segment numbers that meet the consistency conditions, establish the number set, and generate the residual direction matching frame segment number set;

[0100] For the residual direction consistency interval, the frame number combinations within it are evaluated for consistency. Specifically, the number of abrupt changes in the brightness difference direction is counted in each segment. A change is defined as a difference in direction between the current frame and the previous frame; for example, if the 10th frame is positive and the 11th frame is negative, this constitutes one change. The number of changes is recorded for all frame differences within the segment. A direction consistency threshold is set to determine whether to remove the segment. The threshold can be set to 10% of the total number of frames or a specific integer value. For example, if the direction consistency threshold is set to 3, and a segment contains 50 frames but has more than 3 changes, it is considered not to meet the direction consistency requirement and is removed from the candidate segments. Only segments with changes within the threshold are retained as valid segments. The threshold setting depends on the stability of the image content and the actual brightness change tolerance. It is recommended to set a reasonable range through image sample analysis. For example, according to test statistics, the average number of changes in stable segments in video samples is 1 to 3, so setting the threshold to 3 is more appropriate. The frame number segments that meet the conditions are uniformly set to generate a residual direction matching frame segment number set, which serves as a frame segment reference for subsequent processing stages.

[0101] Please see Figure 5 The specific steps of S4 are as follows:

[0102] S401: Extract the number of each frame from the residual direction matching frame segment number set, obtain the image residual data of the corresponding frame in the chroma channel, extract the residual values ​​of all pixels in the chroma channel of each frame, and combine them into a sequence according to the frame number order to generate a chroma residual value sequence.

[0103] The frame numbers are sequentially extracted from the residual direction matching frame segment number set. The corresponding video frame is located according to each number. After reading the frame image in YUV420 format, the U and V chroma channel processing flow is entered. The pixel values ​​of the two chroma channels are extracted separately. For each pixel position, the difference between the current frame and the previous reference frame is calculated. The difference is the chroma value of the current frame minus the chroma value of the reference frame. For example, if the pixel value in the 120th row and 160th column of the U channel is 110, and the value of the reference frame at the same position is 105, then the difference is 5. After calculating for all pixels, a complete residual matrix is ​​obtained. The matrix is ​​then flattened into a vector. The operation is repeated to extract the chroma residual vectors of all frames and sorted according to the frame number. For example, the chroma residual vectors composed of frames numbered 15, 16, 17, etc. are arranged in sequence and combined into a two-dimensional array. Each row corresponds to the chroma residual content of one frame. Finally, a chroma residual value sequence is formed to prepare the data foundation for subsequent chroma change analysis.

[0104] S402: Based on the pixel residual value corresponding to each frame in the chroma residual value sequence, calculate the chroma residual variance value of each frame in turn, and then perform difference operation on the variance values ​​of adjacent frames to obtain the sequence of chroma residual variance variation values ​​between consecutive frames, and obtain the chroma variance variation value between frames.

[0105] Based on the pixel residual data corresponding to each frame in the chroma residual value sequence, the variance of all pixel residuals is calculated frame by frame to obtain the dispersion of pixel changes within a single frame. The difference between all pixel values ​​in a chroma residual vector and its average value is taken, the squared difference is averaged, and the result is the chroma residual variance of that frame. If the chroma residual of a frame has multiple values ​​such as 3, 5, 4, 6, 2, the average is 4, the squared difference is 1, 1, 0, 4, 4, and the average result is 2. This value represents the degree of chroma change within the frame. Next, the variance calculation results of all frames are arranged in numerical order, and the difference between the variance values ​​of adjacent frames is calculated. For example, the variance of frame number 25 is 1.8, and the variance of frame number 26 is 2.3. The difference between the two is 0.5. All adjacent frame pairs are processed in this way to generate a new sequence. This sequence represents the fluctuation of chroma difference between consecutive frames. Each value in the sequence corresponds to the change range of chroma variance between two frames, which is used to reveal the stability or fluctuation trend of chroma residual between image frames.

[0106] S403: Based on the change amplitude value corresponding to each frame segment in the inter-frame chromaticity variance change amplitude value, compare all amplitude values ​​in the interval segment by segment to see if they are within the specified chromaticity stability threshold range, mark the frame segment number that meets the range condition, and combine the corresponding numbers to form a set to obtain the chromaticity stable frame segment number set.

[0107] The analysis uses the inter-frame chroma variance variation amplitude as the basis, iterating through the continuous variation amplitude values ​​of each frame segment to determine whether all values ​​are within the chroma stability threshold range. The threshold is set to 0 to 0.5, a range derived from statistical analysis of chroma residual variance of multiple typical video samples. In most scenes with static backgrounds and slow color changes, controlling chroma variance fluctuation within this range is reasonable. Each variation amplitude value within a given frame segment is checked individually; if any value exceeds the upper limit, the frame segment is marked as unstable. Otherwise, it is marked as a stable frame segment. For example, if the amplitude values ​​of segments numbered 40 to 48 are 0.2, 0.3, 0.4, 0.1, 0.2, 0.2, 0.3, 0.4, and 0.3, then all values ​​in this segment are within the threshold and meet the stability condition. If a value of 0.6 or greater appears in segments numbered 50 to 57, it is marked as unstable. All the frame segments that meet the conditions are organized by segment and integrated into a set of chroma stable frame segment numbers. The number set can be in the form of an array or a list structure, providing a basic frame order for subsequent video processing or frame filtering.

[0108] Please see Figure 6 The specific steps of S5 are as follows:

[0109] S501: Based on the residual direction matching frame segment number set and the frame number of each frame in the chroma stable frame segment number set, perform the matching operation between the numbers, extract the frame numbers that exist in both sets at the same time, and obtain the set of frames with consistent numbers.

[0110] Based on the residual direction matching frame segment number set and the frame numbers in the chroma stable frame segment number set, the two number sets must first be arranged in ascending order to ensure that the subsequent matching operations are performed in the same order. During the matching process, a cross-checking method is used to perform cross-checking between the numberings by comparing each item. For each number in set A (residual direction matching number set), a check is performed in set B (chroma stable frame segment number set) to see if there is a matching number. If there is, the number is recorded in the matching number frame set. For example, set A contains numbers 101, 102, 103, and 104, and set B contains... If the numbers are 100, 102, 104, and 106, then the set of matching number frames is 102 and 104. This operation can be performed using a set intersection function or an equality comparison method. Without relying on external function libraries, a nested loop approach can be used. The outer loop iterates through each number in set A, and the inner loop sequentially searches for completely equal numbers in set B. If a match is found, it is marked and stored in the output set. The extraction process of the matching number frame set can be stored in the form of a list to ensure that the data structure is scalable, such as the matching number frame set being [102, 104].

[0111] S502: Call the consistent number frame set, determine whether there is an intersection based on the number correspondence between the set and the residual direction matching frame segment number set, mark the corresponding number status, and obtain the frame number comparison status sequence;

[0112] After calling the consistent numbered frame set, it is necessary to perform a number intersection judgment operation based on the number correspondence between this set and the residual direction matching frame segment number set. First, a mapping table is established, with the residual direction matching frame segment number set as the main set. A number correspondence status dictionary structure is constructed, with each number initially in the "unmatched" state. Then, the number values ​​in the consistent numbered frame set are read one by one, and the main set is checked to see if the number exists. If it exists, the status mark of the number is updated to the "matched" state. If the number does not exist, the default "unmatched" state is maintained. For example, if the residual direction number set is [100, 101, 102, 103, 104], the consistent number is... If the set of numbered frames is [102, 104], then the comparison state sequence is constructed as follows: number 100 is in state 0, number 101 is in state 0, number 102 is in state 1, number 103 is in state 0, and number 104 is in state 1. Here, state 0 indicates a miss, and state 1 indicates a hit. The judgment process can use conditional judgment statements, such as "if number in matching numbered frame set", to search and update the state number by number. This operation is suitable for hash structures to speed up the judgment efficiency when processing a large number of data frames. The output frame number comparison state sequence is such as [0, 0, 1, 0, 1], indicating whether each frame has a corresponding number in the matching set.

[0113] S503: Based on the frame number comparison status value of each frame in the status sequence, construct a tag matrix with consistent numbering order, summarize the tag status of all frames, and generate a frame-level comparison consistency tag set.

[0114] To construct a label matrix with consistent numbering order, the state values ​​of each frame in the state sequence are compared based on their frame numbers. First, a two-dimensional data structure needs to be established, with each row corresponding to a state label for a numbered frame. Column attributes can be defined as two fields: the frame number and the state value. First, the list of numbers is bound to the state sequence one by one to generate label pairs. Then, this is converted into a matrix structure. For example, if the list of numbers is [100, 101, 102, 103, 104], and the corresponding state sequence is [0, 0, 1, 0, 1], then the label matrix rows are: the first row is 100 and state 0, the second row is 101 and state 0, the third row is 102 and state 1, and so on. The matrix generation process can be completed by constructing a list of tuples and then converting it into a two-dimensional array structure. To clarify the meaning of the labels, state 0 can be defined as "inconsistent" and state 1 as "consistent". At the same time, additional information fields such as timestamps and frame types can be added to the records, so that the label matrix has richer data structure support. For example, if a frame number is 102, state is 1, and frame type is a predicted frame, then the row in the label matrix is ​​[102, 1, P frame]. The label matrix forms a frame-level alignment consistency label set with a complete structure and consistent order, which helps to visualize the status of batch frames and further data comparison processing in the subsequent analysis stage.

[0115] To verify the effectiveness and stability of the proposed dual-dimensional luminance and chrominance collaborative analysis mechanism in the frame-level consistency determination process, an implementation method for verifying the collaborative effect was designed, and comparative tests were conducted using multiple sets of simulated abnormal video sequences. The test content included the following steps:

[0116] First, a simulated broadcast migration test environment was constructed, and video sequence samples containing different anomaly types were prepared. The sample types included: Type I (randomly inserted all-black frames), Type II (color block interference in local areas, simulating decoder errors), Type III (simultaneous discontinuous changes in luminance and chrominance channels), and Type IV (normal content without any human interference). Based on the algorithm framework of this invention, the following three types of analysis processes were executed respectively: (1) using the luminance channel residual direction analysis method alone (corresponding to the execution of steps S1 to S3, with output in S5 based on the results of S3); (2) using the chrominance channel residual variance analysis method alone (corresponding to the execution of steps S1, S2, and S4, with output in S5 based on the results of S4); (3) executing the complete collaborative analysis and intersection decision method of S1 to S5 of this invention. The three types of methods correspond to the different dimensions of the judgment process in this invention, forming a comparable experimental group.

[0117] Subsequently, the detection performance of the three methods was statistically analyzed by combining the abnormal frame detection rate and the normal frame false alarm rate. The results showed that method (1) performed well in handling brightness anomalies, but the detection rate dropped significantly when dealing with color information anomalies; method (2) was highly sensitive to chromaticity anomalies, but not sensitive to brightness structure variations, and was prone to omissions. Method (3), namely the dual-dimensional collaborative processing method proposed in this invention, significantly improved the ability to identify various abnormal frames by performing intersection processing on the two types of frame segment number sets of brightness direction and chromaticity stability (i.e., the operation described in S5). The average detection rate in the three types of abnormal test samples all exceeded 97%, and the false alarm rate in the normal sequence remained below 0.1%. This verification fully demonstrates that the joint comparison mechanism of the "residual direction matching frame segment number set" and the "chromaticity stable frame segment number set" described in this invention has high complementarity in frame-level consistency judgment, and can eliminate misjudgments caused by single-dimensional judgment through intersection operation.

[0118] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for verifying the integrity of video and audio file migration, characterized in that, Includes the following steps: S1: Obtain the synchronized frames in the program's MXF file and the decoder rendering frames of the broadcast server, extract the pixel value sequences of the luminance and chrominance channels, establish the corresponding pixel position mapping, and generate frame pixel feature comparison data. S2: Based on the frame numbering relationship in the frame pixel feature comparison data, extract the timestamp and baseband signal sampling time from the program MXF file, construct two sets of time series, analyze the time difference change trend, mark the numbering segment with stable difference, and generate a synchronized aligned frame numbering sequence. S3: Extract the luminance channel difference sequence from the synchronized aligned frame number sequence, analyze the trend of inter-frame difference changes, determine the similarity of residual directions, record the frame numbers that meet the conditions, and generate a set of residual direction matching frame segment numbers. S4: Extract the corresponding chroma channel residual sequence from the residual direction matching frame segment number set, judge the continuous inter-frame variance change based on the intra-frame residual distribution characteristics, mark the frame segments whose change range meets the requirements, and generate a chroma stable frame segment number set. S5: Take the intersection of the residual direction matching frame segment number set and the chroma stable frame segment number set, extract the consistent frame set, compare the frame tag status, and generate a frame-level comparison consistency tag set.

2. The method for verifying the integrity of video and audio file migration according to claim 1, characterized in that, The frame pixel feature comparison data includes a luminance channel pixel value sequence, a chrominance channel pixel value sequence, and a pixel position mapping relationship. The synchronization alignment frame number sequence includes the program MXF encapsulation file frame number, the baseband signal acquisition frame number, and an inter-frame difference change trend marker. The residual direction matching frame segment number set includes a luminance channel residual value sequence, continuous inter-frame difference trend features, and a residual direction consistency marker. The chrominance stable frame segment number set includes a chrominance channel residual value sequence, intra-frame residual distribution features, and an inter-frame variance change range marker. The frame-level comparison consistency label set includes an intersection frame number set, a comparison status label, and a frame-level consistency result.

3. The method for verifying the integrity of video and audio file migration according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the synchronized frame image data in the program's MXF encapsulation file and the broadcast server's rendered frame, extract the luminance channel value and chrominance channel value of each pixel in each frame image, form a luminance sequence and chrominance channel sequence according to the pixel position coordinates, and generate a luminance pixel sequence and chrominance channel sequence. S102: Based on the brightness pixel sequence and the chroma channel sequence, call the coordinate information of each pixel, establish a correspondence between the brightness value and the two chroma channel values ​​according to the coordinates, and index and organize the mapping content to generate brightness and chroma mapping index pairs; S103: Based on the brightness and chromaticity mapping index pairs, integrate the brightness values ​​and chromaticity channel values ​​at the same position coordinates, construct the feature combination structure of pixels in the frame image, output the comparison matrix between pixels, and generate frame pixel feature comparison data.

4. The method for verifying the integrity of video and audio file migration according to claim 3, characterized in that, The specific steps of S2 are as follows: S201: Based on the frame number correspondence in the frame pixel feature comparison data, extract the timestamp associated with each frame number in the program MXF encapsulation file, and synchronously obtain the sampling time in the baseband signal acquisition frame image data. Arrange them according to the frame number to form two sets of time series, and generate a time stamp sequence group. S202: Based on each pair of timestamps and sampling time values ​​in the time stamp sequence group, calculate the difference in sequence and arrange them in the order of frame number. Judge the trend of change of all time difference data, and divide the time difference sequence into segments according to the degree of change to obtain the stable segment of time difference. S203: For the frame number range included in the stable time difference segment, filter the frame number data in consecutive numbering order, uniformly mark them as the corresponding numbers of the synchronization state, establish a frame number sequence structure, and generate a synchronization aligned frame number sequence.

5. The method for verifying the integrity of video and audio file migration according to claim 4, characterized in that, The specific steps for S3 are as follows: S301: Select each group of frames in the synchronized alignment frame number sequence, extract the brightness channel data in the corresponding images of adjacent frames, calculate the gray value difference of the corresponding pixels of the brightness channel of each group of adjacent frames, arrange them in the order of frame number, establish the inter-frame brightness change sequence, and generate the inter-frame brightness difference sequence. S302: Based on the direction of change of the difference between adjacent frames in the inter-frame brightness difference sequence, compare the positive and negative changes of the brightness difference in several consecutive frames in turn, identify the frame pair combination interval with the same direction of change, record all frame numbers in the interval, and obtain the residual direction consistency interval. S303: For each number combination in the residual direction consistency interval, call the frame number data, remove the number segments whose direction change mutation number exceeds the specified direction consistency threshold in the order of consecutive frames, retain only the frame segment numbers that meet the consistency conditions, establish the number set, and generate the residual direction matching frame segment number set.

6. The method for verifying the integrity of video and audio file migration according to claim 5, characterized in that, The specific steps of S4 are as follows: S401: Extract the number of each frame from the set of residual direction matching frame segments, obtain the image residual data of the corresponding frame in the chroma channel, extract the residual values ​​of all pixels in the chroma channel of each frame, and combine them into a sequence according to the frame number order to generate a chroma residual value sequence. S402: Based on the pixel residual value corresponding to each frame in the chroma residual value sequence, calculate the chroma residual variance value of each frame in sequence, and then perform difference operation on the variance values ​​of adjacent frames to obtain the sequence of chroma residual variance variation amplitude values ​​between consecutive frames, and obtain the chroma variance variation amplitude value between frames. S403: Based on the change amplitude value corresponding to each frame segment in the inter-frame chromaticity variance change amplitude value, compare all amplitude values ​​in the interval segment by segment to see if they are within the specified chromaticity stability threshold range, mark the frame segment number that meets the range condition, and combine the corresponding numbers to form a set to obtain the chromaticity stable frame segment number set.

7. The method for verifying the integrity of video and audio file migration according to claim 6, characterized in that, The specific steps of S5 are as follows: S501: Based on the residual direction matching frame segment number set and the frame number of each frame in the chroma stable frame segment number set, perform a matching operation between the numbers, extract the frame numbers that exist in both sets at the same time, and obtain a set of frames with consistent numbers. S502: Call the set of consistent numbered frames, determine whether there is an intersection based on the number correspondence between the set and the numbered set of the residual direction matching frame segments, mark the corresponding number status, and obtain the frame number comparison status sequence; S503: Based on the frame number comparison status value of each frame in the status sequence, construct a tag matrix with consistent numbering order, summarize the tag status of all frames, and generate a frame-level comparison consistency tag set.

8. The method for verifying the integrity of video and audio file migration according to claim 1, characterized in that, The program MXF file refers to a program file in the MaterialExchangeFormat format that conforms to broadcast industry standards. It encapsulates audio and video frame data arranged in chronological order and associated frame-level description information. The time information of the frames exists in the form of PresentationTimeStamp. The broadcast server decoder rendering frame refers to a single frame of image data generated by the player or hardware decoding module inside the broadcast system after decoding the input program file. It is an image signal used to directly output video frames to the broadcast link. The synchronization frame refers to the frame pair whose content has a corresponding relationship after a mapping relationship is established between the file end frame number and the broadcast end decoded frame number during the migration integrity comparison process; The frame pixel feature comparison data refers to the data information set generated by establishing a mapping between the frame image data at the program file end and the frame image data at the broadcast server end at the same pixel position, including the correspondence between the luminance channel value, the chrominance channel value, and the frame number.

9. The method for verifying the integrity of video and audio file migration according to claim 1, characterized in that, The synchronized alignment frame number sequence refers to the set of frame numbers with stable time fluctuations selected by analyzing the time difference between the timestamp sequence at the file end and the sampling time sequence at the acquisition end. The luminance channel difference sequence refers to the sequence of differences calculated between the corresponding pixel values ​​in the luminance channel of the source file frame and the broadcast frame; The similarity of residual directions refers to the similarity of the directions of the residual vectors of the brightness channels in adjacent frames. The judgment is based on whether the angular relationship between the directions of the residual vectors is within an acceptable range, reflecting the consistency of the trend of image changes. The residual direction matching frame segment number set refers to a set of multiple frame numbers that are determined to have a continuous and abrupt change trend after the residual direction consistency analysis of the luminance channel.

10. The method for verifying the integrity of video and audio file migration according to claim 1, characterized in that, The chroma channel residual sequence refers to the set of chroma value differences of each corresponding pixel in the U and V chroma channels of the frame pair, reflecting the changes in color information between the two frames; The variance variation refers to the degree of difference in the statistical distribution of the contrast chroma channel residual values ​​within or between frames; The set of chroma-stable frame segments refers to the set of frame segments selected after evaluation of the variance change of the chroma channel residuals. The frame pairs in the set have the characteristics of stable color changes and strong regularity of residual fluctuations. The number intersection processing refers to extracting the frame numbers that are common to the two number sets as common elements; The frame-level alignment consistency tag set refers to the set after all the comparison frames are marked according to the results of the brightness direction consistency and color stability judgment. Each frame number corresponds to a tag value.

Images