WO2013098446A1 - Systeme de traitement adaptatif d'images numeriques - Google Patents

Systeme de traitement adaptatif d'images numeriques Download PDF

Info

Publication number
WO2013098446A1
WO2013098446A1 PCT/ES2012/070826 ES2012070826W WO2013098446A1 WO 2013098446 A1 WO2013098446 A1 WO 2013098446A1 ES 2012070826 W ES2012070826 W ES 2012070826W WO 2013098446 A1 WO2013098446 A1 WO 2013098446A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
time
processing system
images
flexible
Prior art date
Application number
PCT/ES2012/070826
Other languages
English (en)
Spanish (es)
Inventor
Higinio MORA MORA
Jerónimo MORA PASCUAL
María Teresa SIGNES PONT
Original Assignee
Universidad De Alicante
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 Universidad De Alicante filed Critical Universidad De Alicante
Publication of WO2013098446A1 publication Critical patent/WO2013098446A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T1/00—General purpose image data processing
    • G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • 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
    • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • 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
    • H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
    • H04N19/162—User input
    • 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 invention relates to a digital image processing system that allows setting the digital image processing time in compression / decompression processes for application contexts with temporary restrictions.
  • the present invention provides a solution to the processing of digital images in those situations for which the available time is limited.
  • Digital image compression / decompression methods play an important role in numerous applications where resources available for viewing, storage and processing are scarce or limited.
  • a framework in which such limitations are vital is in the development of applications for internet and mobile devices such as phones or PDAs.
  • the development of interfaces directed towards the end user is strongly conditioned by the time of downloading and viewing multimedia data.
  • the present invention provides a novel aspect in the compression and decompression systems of digital images since it allows to establish restrictions on the moment of completion of the compression / decompression of the images and regulate the processing time of the system. In this way, inconveniences related to the non-predictability of the processing time of the digital images are overcome. With this system it is possible to set the compression / decompression time of the images regardless of their compression rate, size, number of colors and resolution.
  • the general operation scheme is as follows: the system receives a compressed or uncompressed digital image for processing and the maximum duration, in units of time, to carry it out.
  • the flexible digital image processing system executes the action by compressing / decompressing the image within the stipulated time.
  • the process of compressing / decompressing digital images has been investigated in order to fragment it into independent subtasks. Subsequently, the functioning of these subtasks has been adapted to provide them with temporary flexibility, incorporating parameters that calibrate their operation and influence their duration.
  • the proposed flexible processing system determines, according to the time available, the non-mandatory and optional tasks that may be totally or partially disregarded during the execution of said processing and will only be executed when the temporary restriction permits. With this scheme, even if the complete processing exceeds the time available, it will be possible to obtain results within the imposed term.
  • the flexible digital image processing system receives two inputs and produces two outputs.
  • the entries correspond to the following elements: • Digital image or compressed image: the digital image corresponds to the Uncompressed digital image and the compressed image corresponds to the binary file of the compressed image.
  • Line No_Avaliable informs if it is possible to compress or decompress the image with the temporary restriction imposed.
  • the proposed system consists of the following components:
  • Flexible digital image processor that is responsible for processing the digital images it receives to compress or decompress them, consisting of an image compressor module and a compressed digital image compressor module.
  • Time control unit that is responsible for calibrating the flexible digital image processor according to the time available.
  • the implementation of the time control unit maintains determinism in the response times of the entire system.
  • the module can be built by means of a combinational circuit embedded in the system or, by accessing a look-up table with the value of the parameters associated with each of the temporary restrictions that allow the algorithm to be completed in the established time.
  • the line No_Avaliable is activated at a high value.
  • the application can decide what action to take: reject the calculation or obtain the image that is available in the shortest time available.
  • the first parameter used to adjust the processing time is the DCT coefficients of the digital image. These coefficients are arranged in a two-dimensional matrix and their placement is related to the spectral frequency. The higher frequency coefficients represent those parts of the image that have greater detail. If the number of coefficients of the DCT is altered in the coding of the image, the degree of detail with which it is represented is therefore influenced, so that the more coefficients that disregard the less detail the reconstructed image will possess. However, generally the level of detail of some images is higher than the human eye can perceive given its resolution, so if they are removed, the loss of detail after decompressing the image will not be perceived by a human observer and can Be assumed by certain applications.
  • the amount of DCT coefficients that are calculated is proportional to the time cost of this stage of the algorithm. The less coefficients are calculated, the less time is required to perform the corresponding transformation. Therefore, this parameter is used to adjust the processing time of the system to existing restrictions. Depending on whether the temporary restrictions are at the compressor end or at the decompressor end, the following situations may occur in the calculation of the DCT transform coefficients:
  • the parameter that indicates the amount of coefficients that are calculated according to the available time is called d ⁇ ⁇ n 1 > for the DCT in the od ! ⁇ ⁇ n 2 > for the IDCT (Reverse of the Discrete Cosine Transform) in decompression.
  • the second parameter that the system uses to adapt the calculation time consists in the decomposition of the original image into a quantity (D) of sub-images formed from homogeneously distributed samples of the original image. Each of the resulting sub-images shows a reduced version of it and has a size D times smaller than the original image and a consequently shorter processing time.
  • the D value is a system design parameter that determines the ability to adjust to temporary constraints. The system has more flexibility to cover more demanding temporary requirements the higher the value of D. The particular case of considering the image as a whole corresponds to the value of D equal to the unit.
  • a quantity D of sub-images to be constructed has been established in advance, of which only a part of them will be used for compression or decompression of the entire image.
  • a parameter called d 2 ⁇ ⁇ Ni> is used that indicates the number of sub-images used in compression or decompression.
  • the proximity pixels are used. Therefore, the order of selection of the images that are available is providing information of the whole image using homogeneously spaced pixels. In the decompression process you can also indicate the amount of sub-images that are used to reconstruct the original image according to the time available for it.
  • the highest quality image is obtained using all sub-images that have been encoded in the compression process, but if necessary, approximations to it can be obtained using a smaller number of encoded sub-images.
  • the parameter d 2 ⁇ ⁇ N 2 > is used to indicate the number of sub-images used in decompression.
  • This technique consists in simplifying the mathematical operations involved in the compression / decompression process of the images through the application at the hardware level of the concepts of inaccurate computing. With this technique it is verified that the precision that the data take in the arithmetic operations involved in the execution of the instructions directly influences the time that the processor needs to perform said calculation.
  • FIG. 1 An outline of the image compressor system is shown in Figure 1. This figure represents the inputs and outputs provided by the described system.
  • Figure 2. This figure represents a higher level of detail of the flexible processing system.
  • FIG. 3 shows the compression scheme of the flexible digital image processing system. This figure represents in detail the elements that make up the compression process.
  • FIG. 4 shows the decompression scheme of the flexible digital image processing system. This figure represents in detail the elements that make up the decompression process.
  • the flexible digital image processing system (3) receives two inputs and produces two outputs.
  • the scheme of the system is described in Figure 1.
  • the entries it has are the following:
  • Input image (1) that will be a digital image or a compressed image.
  • the outputs that are generated are the following:
  • Resulting image (4) that will be a digital image or compressed digital image as appropriate.
  • Line No_Avaliable (5) that indicates whether it is possible to compress or decompress the image with the time restriction indicated.
  • the parts of the flexible digital image processing system (3) are shown in Figure 2.
  • the system consists of the following components:
  • Flexible digital image processor (7) it is the module that is in charge of processing the digital images that it receives to compress or decompress them.
  • Time control unit (6) it is the module that is responsible for calibrating the flexible digital image processor according to the available time indicated.
  • These parameters have the function of limiting the compression time according to the restrictions imposed by the application, determining the characteristics of the compression / decompression carried out and obligatorily determining the quality of the final result obtained.
  • the flexible digital image processor (7) is composed of the image compressor module (8) and the image decompressor module (9).
  • Figures 3 and 4 describe the detail of the flexible digital image processor in which the described calibration parameters come into play and show the complete diagrams for the compression and decompression phase and the information flows between each of the device parts. The following describes the embodiments of each module. Compression of an image:
  • the time control unit (6) determines the value of the calibration parameters that adjust the processing process. compression to obtain the resulting image (4) compressed in the available time. If this adjustment is not possible, it is indicated with the line No_Avaliable (5).
  • the image compressor module (8) is then started up, taking as input the digital image itself and the calibration parameters.
  • the compression process is executed when an uncompressed digital image is received as input.
  • the elements that make up this compressor module are the following: preprocessor (13), spatial subdivisor (14), converter (15), resampler (16), DCT calculator (17), quantizer Q (18) and entropic encoder (19).
  • Image compression is detailed in the scheme shown in Figure 3.
  • the calibration parameters configured by the time control unit (6) influence the following stages of the compression process by adjusting its processing time:
  • the objective of this stage is to prepare the image for subsequent subdivision into sub-images.
  • the preprocessor (13) adapts the size of the initial image ⁇ f ⁇ to the processes to be carried out subsequently related to compression and processing time management. With the resizing, the optimum size is sought for the spatial subdivision of the image, the creation of the blocks for the DCT and for the decimation of the chroma of the color model used. For this reason, in addition to the input image itself (1), it receives the calibration parameter ⁇ N ⁇ which indicates the number of sub-images in which the image is decomposed (10).
  • This stage of processing is mandatory and is fully executed regardless of the time constraints and time control parameters.
  • the resized digital image ⁇ f ⁇ (20) is obtained with the appropriate size for further processing.
  • the spatial subdivisor element (14) decomposes the resized digital image ⁇ f ⁇ (20) into a collection of sub-images ⁇ , f 2 , fO ⁇ set by the system.
  • the amount D of sub-images can be set according to the application and the versatility required.
  • the way in which the image is going to decompose consists in forming each sub-image with pixels distributed homogeneously from the original resized image, thus obtaining reduced versions of the original image.
  • the calibration parameter ⁇ N ⁇ number of sub-images in which the image is decomposed (10), with ⁇ ⁇ ⁇ D, indicates how many of these sub-images must be constructed to comply with the application's restrictions.
  • a sequence of ⁇ ⁇ sub-images ⁇ , f 2 , f N1 ⁇ (21) of size D times smaller than the size of the original resized image is obtained.
  • each of the sub-images is processed independently. To promote the progressive processing of sub-images, a sequential and segmented design has been carried out. Thus, it is not necessary to wait for all sub-images to pass through a block to start the execution of the next block.
  • the processing time of this stage is linear with the amount of sub-images ⁇ ⁇ that must be constructed, and its complexity is related to the number of pixels in the original image.
  • This stage of the processing is also mandatory, although, from now on, the rest of the stages are mandatory only for the ⁇ ⁇ sub-images indicated by the calibration parameter that indicates the number of sub-images in which the image is decomposed (10) and Optional for the rest of sub-images ⁇ f N i + i, ⁇ 'D ⁇ according to the remaining time available.
  • RGB red, green, blue
  • YIQ luminance, in-phase, quadrature
  • the converter element (15) of this module is responsible for the transformation of the color space to obtain the separate YIQ components (22) and the resampler (16) receives these separate YIQ components (22) and resamples them, producing a version of the reduced chroma components (23) with lower spatial cost.
  • the methods used are the traditional ones to perform these operations. These stages are mandatory for those sub-images that are processed.
  • the DCT calculator (17) executes the DCT independently to the reduced chroma components (23) of color Y, ⁇ and Q 'of each of the sub-images successively. Each component is subdivided into 8x8 blocks and the transformation is applied to these blocks.
  • a degree of flexibility in the delay is incorporated by controlling the number of required DCT coefficients of each 8x8 matrix and by the precision of the mathematical operations of addition and product.
  • This control is specified by the time control unit (6) using the ⁇ ri> and ⁇ p 1 > calibration parameters.
  • the value of the ⁇ rii> parameter indicates the number of coefficients (1 1) ⁇ n> of the DCT that must be calculated to meet the temporary restrictions of the application. This value can range from 1 to 64.
  • n value only the first n coefficients of the 8x8 matrix are found according to the known zigzag path. For this, the calculation of the DCT 1 D for the rows is maintained, and only the DCT of the necessary columns is processed to obtain the indicated coefficients.
  • ⁇ > indicates the precision of operands (12) in the form of the number of bits of the operands involved in the sum and product operations that are carried out in this process. The fewer bits used, the smaller the delay produced by the processing.
  • the coefficients (25) obtained are quantified and the possible redundancy incorporated in the data is eliminated, generating a non-redundant sequence of information (26) for each subimage.
  • the quantifier Q (18) is responsible for the quantification process and the entropic encoder (19) the element that eliminates redundancy.
  • the methods used are the traditional ones to perform these operations.
  • This stage is mandatory for those sub-images that are processed.
  • the resulting compressed image (4) is formed by the concatenation of the information generated for each subimage.
  • the time control unit (6) determines the value of the calibration parameters that adjust the decompression process to obtain the resulting image (4) in the indicated time. If this adjustment is not possible, it is indicated with the line No_Avaliable (5).
  • This image decompressor module (9) is then put into operation, taking as input the compressed digital image itself and the calibration parameters.
  • the decompression module is detailed in Figure 4.
  • the parts that make up this decompressor module are the following: entropic decoder (27), Q quantifier (28), IDCT calculator (29), resampler (30), converter (31), Image builder (32) and postprocessor (33).
  • the output is the decompressed digital image.
  • the entropic decoder (27) takes the compressed input image (1) composed of the set of ⁇ ⁇ sub-images compressed according to the above compression scheme and generates the decoded coefficient vectors (25) corresponding to each of them.
  • the quality of the generated image is limited by the information collected in the compression phase, therefore, the consequence of applying these restrictions depends on the information that was used at that stage. If the values of the temporal parameters of this phase are equal to or greater than those used to compress, then there is no additional loss of information, while if they are lower, those sub-images and components of the coefficient vector that do not give time to calculate. The sub-images you don't have are the last ones in the data stream of the compressed image, while the discarded coefficients are those of higher frequencies to maintain the maximum subjective quality that the available time allows. The positions of the coefficients not considered have zero value.
  • This step performs the inverse operation to that performed by the quantizer Q of the compression process following a known calculation method.
  • the quantifier Q (28) obtains the vectors of quantified coefficients (24).
  • the IDCT calculator (29) is taking groups of n elements and forming an 8x8 matrix with each group taking into account the zig-zag sorting followed by the compressor.
  • the reverse transformation is performed on each of these matrices following the same pattern as the one used for the direct transformation, that is, it is first calculated in the dimension of the rows and then the same is done in the dimension of the columns.
  • the calibration parameter ⁇ n 2 > indicates the number of coefficients (1 1) ⁇ n> of the DCT, indicating how many coefficients of each vector the processing should be performed at this stage, while the parameter ⁇ p 2 > indicates the precision required for the operations of addition and product that take place by the transformed one.
  • the IDCT block there are all 8x8 matrices that complete the reduced chroma components (23) Y, ⁇ and Q 'of each subimage.
  • the original image is composed from the sub-images ⁇ , f 2 , f N 2 ⁇ (21) that have been decompressed.
  • the image constructor (32) progressively incorporates subimage after subimage, obtaining more information each time and gradually improving the quality of the image. This process ends when the final image has been completed with all the sub-images that the available time has allowed to decompress.
  • the reconstruction of the image in case of not using all sub-images is carried out with the criterion of obtaining the best possible subjective quality. In this way, the open pixel gaps corresponding to non-generated sub-images are covered by neighboring pixels according to a proximity criterion. As the number of sub-images that are counted is known, the reconstruction process forms the original image by placing the pixels of the sub-images or their duplicates according to the case.
  • the sequential processing performed allows the digital image (20) to be reconstructed progressively as sub-images arrive at this stage. In this way, the image can be shown to the user to perceive the increase in quality with each sub-image obtained. This stage is mandatory to reconstruct the image from the sub-images considered.
  • the postprocessor removes the rows and columns initially added thus obtaining the original input image (1).

Landscapes

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

Abstract

La présente invention concerne un système pour le traitement d'images numériques qui permet d'ajuster de manière dynamique les temps de compression/décompression aux besoins de chaque champ d'application ou de chaque industrie. Le problème résolu par le système selon l'invention est celui de fournir une prédictibilité des temps de compression/décompression de l'image, ce qui permet d'établir ce temps de traitement indépendamment des caractéristiques de l'image et de l'environnement de fonctionnement. Le système est formé d'une carte matérielle dans laquelle sont intégrés les modules fonctionnels nécessaires pour exécuter le traitement de compression/décompression. La carte reçoit comme entrée, le paramètre temporel d'ajustement et la séquence de données qui composent l'image de départ ou comprimée. Cette carte peut être placée directement dans des ordinateurs, ou bien, le système qu'elle décrit peut être mis en oeuvre avec des critères de miniaturisation ASIC pour pouvoir être incorporée dans d'autres dispositifs d'acquisition tels que des appareils photographiques numériques ou des dispositifs de visualisation tels que des moniteurs ou des écrans.
PCT/ES2012/070826 2011-12-26 2012-11-26 Systeme de traitement adaptatif d'images numeriques WO2013098446A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
ES201101362A ES2415755B2 (es) 2011-12-26 2011-12-26 Sistema de procesamiento flexible de imágenes digitales.
ESP201101362 2011-12-26

Publications (1)

Publication Number Publication Date
WO2013098446A1 true WO2013098446A1 (fr) 2013-07-04

Family

ID=48696385

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/ES2012/070826 WO2013098446A1 (fr) 2011-12-26 2012-11-26 Systeme de traitement adaptatif d'images numeriques

Country Status (2)

Country Link
ES (1) ES2415755B2 (fr)
WO (1) WO2013098446A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS63120378A (ja) * 1986-11-10 1988-05-24 Hitachi Medical Corp 画像処理装置
US6483540B1 (en) * 1997-06-16 2002-11-19 Casio Computer Co., Ltd. Image data processing apparatus method and program storage medium for processing image data
EP1509044A2 (fr) * 2003-08-21 2005-02-23 Matsushita Electric Industrial Co., Ltd. Dispositif de traitement de signal vidéo numérique
US20070133017A1 (en) * 2004-02-12 2007-06-14 Hideyuki Kobayashi Image processing apparatus, photographing apparatus, image processing system, image processing method and program

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS63120378A (ja) * 1986-11-10 1988-05-24 Hitachi Medical Corp 画像処理装置
US6483540B1 (en) * 1997-06-16 2002-11-19 Casio Computer Co., Ltd. Image data processing apparatus method and program storage medium for processing image data
EP1509044A2 (fr) * 2003-08-21 2005-02-23 Matsushita Electric Industrial Co., Ltd. Dispositif de traitement de signal vidéo numérique
US20070133017A1 (en) * 2004-02-12 2007-06-14 Hideyuki Kobayashi Image processing apparatus, photographing apparatus, image processing system, image processing method and program

Also Published As

Publication number Publication date
ES2415755B2 (es) 2014-02-04
ES2415755A1 (es) 2013-07-26

Similar Documents

Publication Publication Date Title
JP7313330B2 (ja) 電子機器でのビデオ画像データ処理
AU2018373495B2 (en) Frequency component selection for image compression
US8754827B2 (en) Updating an image
ES2848525T3 (es) Aparato de decodificación y procedimiento de decodificación
JP5537293B2 (ja) 画像符号化装置及び画像復号化装置
JP7490726B2 (ja) 低レイテンシ動画符号化及び送信のレート制御
US8159606B2 (en) Method and apparatus for reversible, polynomial based image scaling
US20190236400A1 (en) Image processing method and apparatus
US20160234521A1 (en) Parallel decode of a progressive jpeg bitstream
TW201904294A (zh) 數位內容串流壓縮
CN114009027A (zh) 视频译码中的残差的量化
WO2020006291A1 (fr) Codage et transmission vidéo fondés sur la priorité
CN106210769B (zh) 用于产生和分配可艺术指导的连续动态范围视频的方法
US20050196056A1 (en) Image coding and decoding method
JP2012522285A (ja) データ及び三次元レンダリングを符号化するためのシステム及びフォーマット
EP4016460A1 (fr) Procédé et dispositif de codage d'image par ia, et procédé et dispositif de décodage d'image par ia
ES2415755B2 (es) Sistema de procesamiento flexible de imágenes digitales.
WO2018092147A1 (fr) Compression et décompression d'une image sur la base d'un point de regard
CN113228665A (zh) 用于处理配置数据的方法、设备、计算机程序和计算机可读介质
KR20210011407A (ko) 비디오 인코딩 시스템
KR20160072038A (ko) 비디오 데이터 처리 시스템
JP2017050752A (ja) 画像符号化装置及び方法、並びに、画像復号装置及び方法
CN114079823A (zh) 基于Flutter的视频渲染方法、装置、设备及介质
TW202218424A (zh) 用於網路傳輸的內容壓縮
US9444971B2 (en) Image processing system, image processing method, and image processing program

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: 12861133

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12861133

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 12861133

Country of ref document: EP

Kind code of ref document: A1