WO2018196368A1 - Procédé de recompression de données multimédia - Google Patents

Procédé de recompression de données multimédia Download PDF

Info

Publication number
WO2018196368A1
WO2018196368A1 PCT/CN2017/113138 CN2017113138W WO2018196368A1 WO 2018196368 A1 WO2018196368 A1 WO 2018196368A1 CN 2017113138 W CN2017113138 W CN 2017113138W WO 2018196368 A1 WO2018196368 A1 WO 2018196368A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
block
multimedia data
similar
information
Prior art date
Application number
PCT/CN2017/113138
Other languages
English (en)
Chinese (zh)
Inventor
杨灿
魏玉凯
Original Assignee
华南理工大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华南理工大学 filed Critical 华南理工大学
Publication of WO2018196368A1 publication Critical patent/WO2018196368A1/fr

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/42—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation

Definitions

  • the commonly used methods for compressing multimedia data are to divide a pixel in an original image or an image frame into macroblocks according to a certain size, and correspondingly compress each macroblock data to perform corresponding compression.
  • the compressed data is stored in order for each compressed block data.
  • the compression of the internal pixels of the data block in the I-frame of the different GOPs of the data block divided into the image is compressed separately, and the compression between the data blocks is generally predicted by the neighborhood. In fact, there is a large amount of redundancy in non-adjacent areas in the airspace or time domain. There are a large number of similarly strong data blocks in these non-adjacent data blocks.
  • the compression of picture and video-like multimedia data is in human eye physiology.
  • the information loss within the perceptual threshold is acceptable. There is no absolute consistency requirement for the accuracy of the compressed data, and the similarly strong data blocks are replaced without causing excessive degradation of the picture and video quality. These similarly strong macroblocks still exist in the currently widely used compression methods, which results in a large number of unnecessary redundant information blocks in the compressed multimedia data.
  • the invention recompresses the current mainstream video compression standards such as MPEG, H.264 or H.265.
  • the recompression step of the data recompression terminal is as follows:
  • Step S1 decoding the multimedia data compressed by the block-based compression method
  • Step S2 performing block-by-blocking on the decoded multimedia data according to the block size of the original multimedia data compression method, performing similarity comparison on the block data, acquiring similar data block information, and dividing the similar block into similar block groups;
  • Step S3 According to the similar block group information, only the first macroblock data in the group is saved for the macroblock in each similar block group, and other macroblock data in the same similar block group is not saved, and only saved in the original location. Substituting the macroblock information to recompress the original compressed multimedia data;
  • step S1 the multimedia data type and the compression format are identified, and the corresponding decoding method is used to decode the multimedia data, and the luminance value and the chrominance value of each pixel of the original picture or the video I frame are decoded (YUV) Value or YCbCr value), using the pixel value to do the macroblock similarity comparison operation.
  • step S2 the similarity comparison is performed on the block data according to a certain judgment similarity criterion.
  • the similar data block location information group is saved as a temporary file, and a similar block group information is saved for each group of similar data blocks, and the similar block group information is compressed according to the original multimedia data corresponding compression method.
  • the relative position information of the similar block in the decoded multimedia data, wherein the first macroblock in the similar block group is used as the replacement macroblock, and the other macroblocks in the same similar block group are used as the replaced macroblock.
  • A is the final saved recompressed image data.
  • the recompression step of the data recompression terminal is as follows:
  • Step S11.3 According to the number value of the similar block group, only the pixel data of one reference block in the group is saved for the block in each similar block group, and the pixel data of other blocks in the same similar block group are not saved.
  • the block group number value of the replaced block is saved only in the original position, so that the original compressed multimedia data is recompressed to form a sequence consisting of block group number values;
  • the decoding is to restore the multimedia data recompressed by the above method to the multimedia data before recompression, and the decompressing step of the compressed data decompressing end is as follows:
  • the extended range or the global scope is determined according to the recompressed real application scenario or the non-real application scenario, and the real-time application scenario extension range is determined based on the codec delay indicator requirement;
  • the global scope is search matched.
  • step S11.1 the multimedia data type and the compression format are identified, the corresponding decoding method is used to decode the multimedia data, the original information is decoded, and the global data block similarity is performed after the decoding information is used. Comparison operation.
  • the original information is the YUV value or the YCbCr value of each pixel of the original picture or video I frame.
  • step S11.2 similar criteria are determined according to certainty (for example, SAD, PSNR or SSIM)
  • step S11.2 adaptively expanding data compression is performed, which is characterized in that: hierarchically adaptive multi-granularity division strategy is performed according to data blocks in the global content space of the multimedia content.
  • the original multimedia data is subjected to block processing according to the pixel position, and the similarity block is combined according to the basic unit block compressed by the original multimedia data using the compression method, and the blocks formed by the layer are used to perform hierarchical similarity.
  • the comparison operation and hierarchically performs the similar block number in the step S3 on the similar block group, and adds a number of bits in the number to identify the cell level in which the block group number is located.
  • the feature of the division strategy is: The higher the level of the division, the larger the division size, that is, the size of the division is larger, then smaller, and gradually split.
  • the specific coding process on the data compression side is characterized by:
  • S15.1 constructs layer 0 coding: the size of the single frame decoded by the original signal is the size of the block (MxN), and the similar block is found in the global domain; if there is a similar block, the similar block group number and layer are constructed. Number, and identifies the location of the similar block at the layer, and saves a reference pixel tile data for each similar block group of the layer for decoding to be called; if there is no similar block, go to step S15.2 ;
  • the specific decoding process at the data decompression end is characterized by:
  • step SU.3 the first recompressed file data organization manner is characterized by: according to the step S11.
  • step S11.3 the second recompressed file data organization manner is characterized by: generating a replacement block file A according to the similar block group information generated in the step S11.2, where the reference in each similar block group is saved.
  • the residual data relative to the reference block of the group is constructed for each block of the same similar block group.
  • DCT quantization, ZigZag scan, RLE, entropy coding
  • the decoding process is completed together. This step can be ignored if there is no fidelity requirement and only a high compression ratio is pursued.
  • the present invention is directed to a picture-based video compression method based on the block, and can compress the multimedia data again according to the currently widely used image and video compression algorithm, and divide the block in the algorithm into multiple units.
  • the similar block stores only the original data of one block, removes the redundant information in the multimedia data, and further increases the compression ratio of the multimedia data, so as to reduce the storage space occupied by the compressed multimedia data.
  • the invention has the function of compressing the multimedia data compressed by the corresponding data compression method again, and reduces the storage space and transmission bandwidth occupied by the compressed multimedia data.
  • the invention belongs to a universal universal recompression technology for an existing multimedia coding compression scheme.
  • the method provided by the present invention can obtain a specific implementation path.
  • the invention is an inherited innovation development of the existing video multimedia coding method, and has good compatibility and smooth expansion capability.
  • the invention further improves the compression ratio of the existing multimedia coding scheme in technical performance; and the invention can well protect the original investment, for example, expanding on the basis of H.264 to improve the compression ratio but does not need to adopt a brand new
  • the HEVC encoding method can also be applied to the recompression of H.265 to increase its compression ratio.
  • the present invention performs global or extended domain compression on the compressed data after decoding the local unit, and constructs a similar block index for the block-based image or video compression method.
  • the video compression algorithm can replace the data blocks in different GOPs, remove the redundant information in the multimedia data again, and further improve the compression ratio of the multimedia data to reduce compression.
  • the method can be recompressed on the basis of the existing compression method without changing the original data compression structure, can be compatible with the existing compression method, or can be integrated with the existing method, and thus has wide applicability.
  • FIG. 1 is a flow chart of a method of the present invention.
  • FIG. 2 is a schematic diagram of a picture after similarity comparison.
  • FIG. 3 is a schematic diagram of video comparison after similarity comparison.
  • FIG. 4 is a schematic diagram of a recompressed file obtained by the strategy a.
  • FIG. 5 is a schematic diagram of a recompressed file obtained by strategy b.
  • FIG. 8 is a flow chart of a specific implementation of video recompression at the data recompression end.
  • the data recompression terminal mainly includes the following steps:
  • P1 identifying a compression format that has been compressed by the block-based compression algorithm
  • P2 according to the compression format identified by the step PI corresponding to the decoding algorithm, decoding the brightness of each pixel of the picture And chromaticity values (for example: YUV value or YCbCr value);
  • step P4 selecting a certain standard to compare the similarity of the picture block data in step P3, and obtaining a copy file, wherein the temporary file stores each similar group of data blocks to store a similar macro block group information, similar block group The information stores the relative position information of the similar macroblocks in the decoded picture data according to the order of the data blocks in the compression algorithm of the picture data;
  • P5. Scan the original picture data sequentially according to the similar block group information. If the replaced macro block data is not scanned, copy the picture data to the recompressed picture file, and record the replacement macro block data offset when scanning to replace the macro block data. Moving the position and size, when scanning the replaced macroblock data, copying the replaced macroblock data into the recompressed image file, and inserting the replaced identification information, the identification information is corresponding to the replaced macroblock To replace the offset information and data size of the macroblock in the recompressed picture file, the offset position s of the replacement macroblock in the recompressed picture file is calculated as:
  • the replacement macroblock has an offset position of 1 in the original picture file
  • the number of all replaced macroblocks before the replacement macroblock in the original picture file is n
  • the total size is X
  • the replaced identification information is a fixed size d
  • the recompressed data decompression end mainly includes the following steps:
  • R1 the recompressed picture data obtained by sequentially scanning step P5, such as identifying the replaced identification information inserted in the recompressed data in step P5, performing the next step R2, if the replaced identification information is not recognized, going to the step R3, until processing to the end of the recompressed file;
  • the video data compressed by the block-based compression method is recompressed and decompressed, and the specifically implemented system includes two parts: a data recompressing end and a recompressing data decompressing end:
  • the data recompression terminal mainly includes the following steps: [0074] P1, identifying a compression format that has been compressed by the block-based compression algorithm; [0075] P2, corresponding to the decoding algorithm according to the compression format identified by the step PI, decoding each of each video I frame Pixel point YUV value;
  • P5 according to the obtained similar block group information, generate a replacement block file Filel (A), wherein the first data block in each group is stored in the original compressed video data corresponding compressed data and its file Filel (A) Starting offset position information;
  • P6 according to the obtained similar block group information, generate a recompressed file File2 (B), wherein the saved data is the original compressed video data, and all the block compressed data existing in the similar block group is removed, and The block original data position is inserted into the replacement information, and the replacement information is the position information of the first block data in the similar block group in the filel (A).
  • the resulting Filel (A) and File2 (B) are the recompressed files of the original compressed video file.
  • the recompressed data decompression end mainly includes the following steps:
  • the position and size information in (A), the replaced macroblock data is obtained from Filel (A), and additionally copied to the decompressed file;
  • R3 additionally copying the data in File2 (B) to the decompressed file, such as scanning to the end of file File2 (B), and going to step R1.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

L'invention concerne un procédé de recompression de données multimédia, comprenant les étapes suivantes : étape S1, décodage des données multimédia compressées au moyen d'un procédé de compression basé sur des blocs ; étape S2, re-blocage des données multimédia décodées selon une politique donnée, réalisation d'une comparaison de similarité sur les données de bloc, et obtention d'informations d'emplacement de blocs similaires de façon à obtenir des informations de groupe de blocs similaires ; étape S3, recompression, selon les informations de groupe de blocs similaires, des informations de données compressées par le procédé de compression correspondant. La présente invention a pour fonction de recompresser les données multimédia compressées par un procédé de compression de données correspondant, et présente des avantages tels que la réduction de l'espace de stockage occupé par les données multimédia compressées.
PCT/CN2017/113138 2017-04-28 2017-11-27 Procédé de recompression de données multimédia WO2018196368A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710291716.5A CN107087184B (zh) 2017-04-28 2017-04-28 一种多媒体数据再压缩方法
CN201710291716.5 2017-04-28

Publications (1)

Publication Number Publication Date
WO2018196368A1 true WO2018196368A1 (fr) 2018-11-01

Family

ID=59611766

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/113138 WO2018196368A1 (fr) 2017-04-28 2017-11-27 Procédé de recompression de données multimédia

Country Status (2)

Country Link
CN (1) CN107087184B (fr)
WO (1) WO2018196368A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021217623A1 (fr) * 2020-04-30 2021-11-04 深圳市大疆创新科技有限公司 Procédé et dispositif de traitement de données multimédias, et support d'enregistrement

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107087184B (zh) * 2017-04-28 2020-05-22 华南理工大学 一种多媒体数据再压缩方法
CN107682016B (zh) * 2017-09-26 2021-09-17 深信服科技股份有限公司 一种数据压缩方法、数据解压方法及相关***
CN109753572B (zh) * 2018-12-26 2021-03-23 维沃移动通信有限公司 一种多媒体数据处理的方法和装置
CN110087076B (zh) * 2019-04-09 2021-05-14 华南理工大学 一种基于分层块替换的多图像压缩方法
CN111526151A (zh) * 2020-04-28 2020-08-11 网易（杭州）网络有限公司 一种数据传输方法、装置、电子设备及存储介质
WO2022198483A1 (fr) * 2021-03-24 2022-09-29 深圳市大疆创新科技有限公司 Procédé et appareil de compression de données, plate-forme mobile et support de stockage
CN116415159B (zh) * 2023-06-12 2023-08-18 青岛市城市规划设计研究院 一种海绵城市数据管理方法及***

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1762161A (zh) * 2003-02-24 2006-04-19 Vixs***公司 视频数据代码转换的方法和***
CN103248893A (zh) * 2013-05-10 2013-08-14 四川大学 从h.264/avc标准到hevc标准的快速帧间转码方法及其转码器
CN105163122A (zh) * 2015-08-27 2015-12-16 西安空间无线电技术研究所 一种基于图像块相似性的图像压缩及解压缩方法
US20170048534A1 (en) * 2014-04-15 2017-02-16 Arm Limited Method of and apparatus for generating an encoded frame
CN107087184A (zh) * 2017-04-28 2017-08-22 华南理工大学 一种多媒体数据再压缩方法

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7054362B1 (en) * 2001-05-29 2006-05-30 Cisco Technology, Inc. Methods and apparatus for updating a reduction ratio

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1762161A (zh) * 2003-02-24 2006-04-19 Vixs***公司 视频数据代码转换的方法和***
CN103248893A (zh) * 2013-05-10 2013-08-14 四川大学 从h.264/avc标准到hevc标准的快速帧间转码方法及其转码器
US20170048534A1 (en) * 2014-04-15 2017-02-16 Arm Limited Method of and apparatus for generating an encoded frame
CN105163122A (zh) * 2015-08-27 2015-12-16 西安空间无线电技术研究所 一种基于图像块相似性的图像压缩及解压缩方法
CN107087184A (zh) * 2017-04-28 2017-08-22 华南理工大学 一种多媒体数据再压缩方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021217623A1 (fr) * 2020-04-30 2021-11-04 深圳市大疆创新科技有限公司 Procédé et dispositif de traitement de données multimédias, et support d'enregistrement

Also Published As

Publication number Publication date
CN107087184A (zh) 2017-08-22
CN107087184B (zh) 2020-05-22

Similar Documents

Publication Publication Date Title
WO2018196368A1 (fr) Procédé de recompression de données multimédia
JP5139416B2 (ja) ビデオ圧縮における重要度係数符号化のための方法およびシステム
JP5854439B2 (ja) 適応セグメンテーションを用いた動画符号化システムおよび方法
JP5231391B2 (ja) ビデオ圧縮に関連する情報を符号化および復号化する方法およびシステム
US20040179610A1 (en) Apparatus and method employing a configurable reference and loop filter for efficient video coding
TW202114421A (zh) 視訊寫碼之存取單元中之適應參數集合
US8306347B2 (en) Variable length coding (VLC) method and device
US9325997B2 (en) Signaling scalability information in a parameter set
JP2009531992A5 (fr)
CN107864379B (zh) 一种应用于视频编解码的压缩方法
CN117998085A (zh) 视频编码的方法、装置以及存储介质
EP4024872A1 (fr) Procédé et appareil de codage vidéo, procédé et appareil de décodage vidéo, dispositif électronique et support de stockage
WO2023065891A1 (fr) Procédé et appareil de traitement de données multimédia, dispositif, support de stockage lisible par ordinateur et produit de programme informatique
CN114786019A (zh) 图像预测方法、编码器、解码器以及存储介质
TW202203647A (zh) 跨層參考限制條件
CN117136540A (zh) 残差编码方法及设备、视频编码方法及设备、存储介质
JP2023508095A (ja) Ｎａｌユニット関連情報に基づく映像又はビデオコーディング
TW202143722A (zh) 在視訊解碼中的經解碼視訊序列起點存取單元
JP2007006390A (ja) 画像処理装置、画像処理方法及びプログラム
WO2022078339A1 (fr) Procédé et appareil de construction de liste de candidats de pixel de référence, dispositif, et support de stockage
WO2023272533A1 (fr) Procédé de codage et de décodage, codeur, décodeur et support de stockage
CN111212288B (zh) 视频数据的编解码方法、装置、计算机设备和存储介质
CN114731447A (zh) 用于视频编解码的方法和设备
EP1949700A2 (fr) Procede de codage/decodage et dispositif permettant d'ameliorer le masquage d'erreurs video
JP7494315B2 (ja) Ｇｄｒ又はｉｒｐａピクチャに対する利用可能スライスタイプ情報に基づく画像符号化／復号化方法及び装置、並びにビットストリームを保存する記録媒体

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17907147

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 10/02/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 17907147

Country of ref document: EP

Kind code of ref document: A1