CN-119597960-B - Video color ring multi-dimensional auditing and warehousing method and device, electronic equipment and medium

CN119597960BCN 119597960 BCN119597960 BCN 119597960BCN-119597960-B

Abstract

The invention discloses a multi-dimensional auditing and warehousing method, device, electronic equipment and medium for video color ring, which comprise the steps of preprocessing target video color ring to obtain media information, background sound files, a plurality of target I frames and target hash values corresponding to the target I frames, conducting picture scoring auditing on the target video color ring according to the media information, the target I frames and the target hash values, conducting picture re-auditing on the target video color ring according to the target hash values, conducting multi-stage wind control auditing on the target video color ring according to the background sound files, and conducting transcoding storage on the target video color ring according to the media information when the target video color ring passes the picture scoring auditing, the picture re-auditing and the multi-stage wind control auditing. The invention improves the efficiency and reliability of video color bell auditing and warehousing, and can be widely applied to the technical field of video processing.

Inventors

ZHANG YUMIN
CHEN YUANFENG
Lu guixing

Assignees

天翼爱音乐文化科技有限公司

Dates

Publication Date: 20260512
Application Date: 20241114

Claims (9)

1. The multi-dimensional auditing and warehousing method for the video color ring back tone is characterized by comprising the following steps of: Preprocessing target video color ring to obtain media information, background sound files, a plurality of target I frames and target hash values corresponding to the target I frames; Performing picture scoring auditing on the target video color ring according to the media information, the target I frame and the target hash value; performing picture review on the target video color ring according to the target hash value; Performing multi-stage wind control auditing on the target video color ring according to the background sound file; when the target video color ring passes through the picture scoring audit, the picture checking and re-checking and the multi-stage wind control audit, transcoding and storing the target video color ring according to the media information; the step of performing picture scoring auditing on the target video color ring according to the media information, the target I frame and the target hash value specifically comprises the following steps: Determining the resolution, code rate and frame rate of the target video color ring according to the media information; Determining the color number of each target I frame, and determining the average color value of the target video color ring according to the color number; Determining first Hamming distances of the target hash values corresponding to two adjacent target I frames, counting target adjacent target I frame logarithms of the first Hamming distances which are larger than or equal to a preset distance threshold, and further determining the front and rear frame change rate of the target video color ring according to the ratio of the target adjacent target I frame logarithms to the total logarithms of the adjacent target I frames; the resolution, the code rate, the frame rate, the average color value and the front and rear frame change rate are weighted and summed according to a preset weight coefficient to obtain the picture grading value of the target video color bell; when the picture grading value is larger than or equal to a preset picture grading threshold value, determining that the picture grading verification passes; and when the picture grading value is smaller than the picture grading threshold value, determining that the picture grading audit is not passed.
2. The method for multi-dimensional auditing and warehousing of video color ring back tones according to claim 1, wherein the preprocessing of target video color ring back tones to obtain media information, background sound files, a plurality of target I frames and target hash values corresponding to the target I frames specifically comprises: Using FFmpeg to read the media information of the target video color ring; Separating the background audio file of the target video using FFmpeg; separating a plurality of the target I frames of the target video using FFmpeg; converting the target I frames into gray level images, and generating pixel gray level matrixes of the target I frames according to the gray level images; Performing two-dimensional discrete cosine transform on the pixel gray matrix to obtain a pixel transformation matrix of each target I frame; intercepting the pixel transformation matrix according to a preset pixel window to obtain a pixel interception matrix of each target I frame; Determining the element average value of all elements in each pixel interception matrix, and judging whether each element is larger than the element average value; Assigning an element larger than the element average value to be 1, and assigning an element smaller than or equal to the element average value to be 0, so as to obtain a pixel assignment matrix of each target I frame; and sequentially reading each element of the pixel assignment matrix and splicing the elements in sequence to obtain the corresponding target hash value of each target I frame.
3. The method for multi-dimensional review and warehousing of video color ring back tones according to claim 1, wherein the method for performing picture review and re-review on the target video color ring back tones according to the target hash value specifically comprises the following steps: Determining first hash values corresponding to a plurality of I frames of each first video color ring in a preset video color ring library; determining the similarity between the target video color ring and each first video color ring according to the target hash value and the first hash value; when the first video color ring which has the similarity larger than or equal to a preset similarity threshold value does not exist, determining that the picture check and review pass; and when the first video color ring with the similarity larger than or equal to the similarity threshold exists, determining that the picture check and review do not pass.
4. The method for multi-dimensional auditing and warehousing of video color ring back tones according to claim 3, wherein the determining the similarity between the target video color ring back tone and each first video color ring back tone according to the target hash value and the first hash value specifically comprises: determining a second hamming distance between the target hash value and the first hash value; When the second Hamming distance is smaller than a preset distance threshold, determining that a similar I frame exists between the target video color ring and the corresponding first video color ring; And counting the similar I frame logarithm of the target video color ring and each first video color ring, and determining the similarity of the target video color ring and each first video color ring according to the ratio of the similar I frame logarithm to the total number of the target I frames.
5. The method for multi-dimensional auditing and warehousing of video color ring back tones according to claim 1, wherein the method for multi-stage wind control auditing the target video color ring back tones according to the background sound file specifically comprises the following steps: performing voice recognition on the background sound file to obtain target text information, and obtaining a video source file of the target video color ring; performing text auditing on the target text information; Content auditing is carried out on the video source file; And when the target text information passes the text audit and the video source file passes the content audit, determining that the multi-level wind control audit passes.
6. The method for multi-dimensional auditing and warehousing of video color ring back tones according to any one of claims 1 to 5, characterized in that the transcoding and storing of the target video color ring back tone according to the media information specifically comprises: Transcoding the video source file of the target video color ring into a plurality of video color ring files with preset formats and preset resolutions through FFmpeg according to the media information; and distributing the video color ring file to a preset video color ring library for storage.
7. A video color ring multi-dimensional auditing and warehousing device, which is characterized by being used for realizing the video color ring multi-dimensional auditing and warehousing method according to any one of claims 1 to 6, comprising the following steps: The preprocessing module is used for preprocessing the target video color ring to obtain media information, a background sound file, a plurality of target I frames and target hash values corresponding to the target I frames; The picture scoring auditing module is used for conducting picture scoring auditing on the target video color ring according to the media information, the target I frame and the target hash value; The picture checking and re-checking module is used for checking and re-checking the picture of the target video color ring according to the target hash value; The multi-stage wind control auditing module is used for conducting multi-stage wind control auditing on the target video color ring according to the background sound file; And the transcoding storage module is used for transcoding and storing the target video color ring according to the media information when the target video color ring passes through the picture scoring audit, the picture re-checking audit and the multi-stage wind control audit.
8. An electronic device comprising a memory, a processor, a program stored on the memory and executable on the processor, and a data bus for enabling a connection communication between the processor and the memory, the program when executed by the processor implementing the steps of the video color ring multi-dimensional audit warehousing method according to any of claims 1 to 6.
9. A storage medium, which is a computer readable storage medium, for computer readable storage, wherein the storage medium stores one or more programs, and the one or more programs are executable by one or more processors to implement the steps of the video color ring multi-dimensional auditing and warehousing method according to any one of claims 1 to 6.

Description

Video color ring multi-dimensional auditing and warehousing method and device, electronic equipment and medium Technical Field The invention relates to the technical field of video processing, in particular to a method, a device, electronic equipment and a medium for multi-dimensional auditing and warehousing of video color ring back tones. Background With the popularization of the 4G/5G network and the wide application of the smart phone and the mobile Internet, the video color ring serves as an emerging personalized communication mode, and a large number of users are rapidly attracted to be concerned and used. The video color ring not only can provide richer and lively incoming call experience for users, but also provides a platform for enterprises and individuals to display brand images and creative contents. In order to ensure the playing fluency of the video color ring, the attractive effect of the visual effect and the compliance of the content, it is particularly important to develop a set of efficient and intelligent video color ring auditing system. The existing video color ring auditing system has only wind control auditing and video transcoding processes, and whether video pictures are clear or not and whether the pictures are rich or not are judged subjectively by editing auditing, and in addition, the on-line video color ring of the library memory is more, repeated video materials are easy to purchase during purchasing, so that the auditing is needed, however, taking 20 ten thousand video color rings as an example, the playing time of one video color ring is about 25 seconds, if the auditing is carried out by manpower, 200000 x 25/60/60/8 approximately 170 days are needed for the light viewing of the videos, and the workload is huge. In addition, in the existing video color ring auditing system, most of video preprocessing of each part is independently carried out. The ultra-high definition video color ring has high definition and large occupied memory, and if each link needs to process the video, more time and memory space are spent. If the preprocessing flow is to be modified, each link is to be modified independently, and the development workload is large. In summary, the existing video color ring auditing system has the following disadvantages: 1) Whether the video pictures are clear or not and whether the pictures are rich or not are judged subjectively by editing, checking and checking, and error is easy to occur; 2) The number of the stored videos is large, and the manual duplicate checking efficiency is low; 3) Each link independently preprocesses video, the same preprocessing flow needs to be executed for a plurality of times, and the time is long; 4) Each link independently preprocesses the video, each link independently stores the preprocessing result, and the occupied space is large; 5) Each link is independently preprocessed, the bottom layer is not uniform, and each system needs to be independently modified if the auditing standard is to be modified, so that the development workload is high. The above problems need to be solved. Term interpretation: Video color ring, wherein the video color ring is a special service of VoLTE, and the "listened color ring" is evolved into the "watched color ring". When a user makes a voice call or video call under the VoLTE network, a piece of video content that is seen before the call is put through is called a "video color bell". FFmpeg-a set of open source computer programs that can be used to record, convert digital audio, video, and convert it into streams. Resolution, also known as resolution, generally refers to the ability of a measurement or display system to resolve details, determines how fine a bitmap image is, and is typically used as a measure of detail and sharpness displayed in an image, video, or display device. The code rate, also called Bit rate (variable R), refers in the field of telecommunications and computing to the number of bits (bits) transmitted or processed per unit time, in bps (Bit per second). Frame rate (FRAME RATE) is the frequency (rate) at which bitmap images in frames are called units appear continuously on the display. The term applies equally to film and video cameras, computer graphics and motion capture systems. The frame rate may also be referred to as the frame frequency and is expressed in hertz (Hz). H264, H.264, also part ten of MPEG-4, is a highly compressed digital Video codec standard proposed by the Joint Video Team (JVT, joint Video Team) consisting of the ITU-T Video Coding Experts Group (VCEG) and the ISO/IEC Moving Picture Experts Group (MPEG). I-frames, also known as intra pictures (I-frames), are typically the first frame of each GOP (a video compression technique used by MPEG) that is moderately compressed to serve as a reference point for random access and can be regarded as pictures. Gray scale, also known as gray scale. The logarithmic relationship between white and black is divided into s