Classifications

• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
  • G06F1/26—Power supply means, e.g. regulation thereof
  • G06F1/32—Means for saving power
  • G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
  • G06F1/3234—Power saving characterised by the action undertaken
  • G06F1/3296—Power saving characterised by the action undertaken by lowering the supply or operating voltage
• • 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
  • H04N19/117—Filters, e.g. for pre-processing or post-processing
• • 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
  • H04N19/132—Sampling, masking or truncation of coding units, e.g. adaptive resampling, frame skipping, frame interpolation or high-frequency transform coefficient masking
• • 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
  • H04N19/154—Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion
• • 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
  • H04N19/156—Availability of hardware or computational resources, e.g. encoding based on power-saving criteria
• • 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
  • H04N19/157—Assigned coding mode, i.e. the coding mode being predefined or preselected to be further used for selection of another element or parameter
  • H04N19/159—Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
• • 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
  • H04N19/44—Decoders specially adapted therefor, e.g. video decoders which are asymmetric with respect to the encoder
• • 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
  • H04N19/80—Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation
  • H04N19/82—Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation involving filtering within a prediction loop
• • 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
  • H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
  • H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
  • H04N19/51—Motion estimation or motion compensation
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
  • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
  • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
  • H04N21/443—OS processes, e.g. booting an STB, implementing a Java virtual machine in an STB or power management in an STB
  • H04N21/4436—Power management, e.g. shutting down unused components of the receiver
• • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
  • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
  • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Definitions

• the invention relates to a decoding device, a method and a system for decoding multimedia data at a terminal while managing the power consumption necessary for decoding and restitution of a multimedia data stream.
• the invention applies to video decoders, for terminals using the H.264 / AVC standards for Advanced Video Coding or H.265 / HEVC, for High Efficiency Video Coding.
• Today electronic devices such as mobile phones, smartphones, tablets, etc., offer more and more features, such as video, audio, GPS positioning system (Global Positioning System) and various connectivity, for example multi-radio systems with Wi-Fi, Bluetooth, Universal Mobile Telecommunications System (UMTS), High Speed Packet Access (HSPA), LTE-Advanced network technology, which means that energy of these embedded systems is today very sought.
• GPS positioning system Global Positioning System
• UMTS Universal Mobile Telecommunications System
• HSPA High Speed Packet Access
• LTE-Advanced network technology which means that energy of these embedded systems is today very sought.
• UMTS Universal Mobile Telecommunications System
• HSPA High Speed Packet Access
• LTE-Advanced network technology LTE-Advanced network technology
• the company Morphbius offers a subsampled encoder in which the encoding of low quality data is performed.
• Figure 1 is a reminder of the principle implemented for managing the energy of a mobile terminal.
• the standard proposes to exchange metadata from an encoder 1 to one or more final decoders 2 (a single decoder is shown in the figure for reasons of simplification).
• the metadata M d is multiplexed with the conventional data stream F, to obtain a data flow F m .
• This data flow F m is then transmitted to a receiver 2 through a communication network, for example.
• the type of network used depends on the application or system.
• the protocol put in place is different. An example of this is television broadcasting, better known as "broadcast” or streaming or “streaming”.
• the insertion of the metadata depends on the protocol or mode of transport.
• the transmission frequency of the metadata can be done at several levels, for example, at the frame level, the image group level or GOP (in English Group of Pictures - 17 frames for example).
• the decoder will test in the associated field in a frame the presence or absence of the metadata for the sequence to be decoded.
• the encoder can make a decision to change the metadata broadcast frequency when parameters evolve as an evolution of content (scene change) or compression type (quantization parameter).
• a preprocessor 101 analyzes the source content F s and a video encoder 102 encodes the contents of an input video.
• the data stream containing the coded data and the metadata is transmitted to the receiver and decoded by a video decoder 106 which transmits the decoded information to a display module 109.
• the metadata Md is extracted by means of an analyzer 104 at the video encoder 102 and in this case the metadata is encapsulated in a stream conforming to the format of the video encoder, or at the level 103 of the preprocessor 101 and the metadata is then multiplexed with the stream.
• the metadata M d is used by the final decoder 2 to reduce the power required for decoding and presenting the data.
• the video decoder 106 analyzes, 107, the stream F m containing the multimedia stream F and the metadata M d , and transmits the metadata M d to a power or energy control module 108.
• energy control module will, for example, decode the metadata M d and then apply energy consumption reduction operations for the decoding and display of video streams, P c for example. The same energy control module is present at the encoder.
• FIG. 2 shows a known architecture of the video encoder comprising a video decoder 20.
• the video input (compressed data) is transmitted to a transformation module 11 whose output is in connection with a coding module 1 and a quantization module and inverse transformation 1 13.
• the encoded multimedia data F c are transmitted to a stream generation module 1 14 which will also receive motion estimation information from a motion estimation module 1 15 which receives, on the one hand the video input E v , and on the other hand an image resulting from the application of a loop filter 1 16 (known to those skilled in the art) receiving the parameters or vectors of motion compensation 1 17 to apply to the image parameters and the image parameters obtained by the application of the inverse inverse quantization inverse transform.
• a loop filter 1 16 known to those skilled in the art
• An output of the loop filter generates a video output S v .
• the stream generation module produces a compressed data stream that will be transmitted to the decoder.
• a buffer 1 19 makes it possible to store the video images before transmission to the motion compensation module and to the motion estimation module 1 15.
• US patent application 2002080874 describes a system for decoding a data stream representative of a video content and discusses the capacity for complexity decoders of various architectures to decode an incident flow.
• the terminal word designates, for example, a mobile phone, a tablet or any connected device for receiving multimedia streams.
• the term multimedia stream may designate various programs, such as movies, music (video clip), which are distributed through communication networks.
• the object of the present invention relates to a method and a system for managing the energy needed to decode information, multimedia data, video data. It also simplifies the complexity of the filters used, the consumption of filters representing a significant portion of the consumption of a decoder.
• the invention relates to a device for decoding a multimedia data stream at a terminal by managing the energy necessary for the decoding comprising a decoder of said multimedia data and characterized in that said decoder comprises at least one decoding string " low-power circuit comprising an activation module adapted to activate a first low consumption loop decoder or filter line and a low power interpolation chain as a function of at least one parameter representative of user operation constraints and / or user energy and / or the mobile terminal and one or more metadata M d associated with a maximum degradation of quality, decoding complexity and / or energy required for decoding.
• decoder comprises at least one decoding string " low-power circuit comprising an activation module adapted to activate a first low consumption loop decoder or filter line and a low power interpolation chain as a function of at least one parameter representative of user operation constraints and / or user energy and / or the mobile terminal and one or more metadata M d associated with a maximum degradation of quality, decoding complexity and / or energy required for decoding.
• the invention also relates to a system for decoding a multimedia data stream at a terminal by managing the energy required for the decoding, characterized in that it comprises at least the following elements:
• a multimedia data encoder comprising at least one module adapted to generating one or more metadata M d associated with a decoding complexity and / or a decoding energy of the multimedia data stream decoded by the decoder,
• a decoder comprising at least one standard decoding chain, at least one metadata extraction module M d contained in the multimedia data stream, an activation module connected to a switch adapted to activate a first decoding or filtering channel; of a low-power loop and a switch to activate a low power interpolation chain according to at least one parameter representative of energy constraints and one or more metadata M d contained in the multimedia data stream .
• the encoder comprises a modified decoder comprising a low consumption loop filter, an energy estimation module, a standard decoder comprising an energy estimation module, a metadata M d is a pair of values (maximum degradation, potential gain in energy) determined from the multimedia data E v , data S V 2 decoded by the standard decoder, an estimated energy value E 2 by the standard decoder, decoded data Svi by the modified decoder, the estimated energy Ei by the modified decoder.
• PSN RDecoderReference 10 * logi 0 (d 2 / MSE), d corresponding to the maximum amplitude of the pixels, for example 255 for pixels coded on 8 bits or 1023 for 10 bits coded pixels, MSE corresponding to the mean squared error (E v - Sv 2 ) 2 between the reference data stream E v and the data stream decoded by the standard decoder 10 * logi 0 (d 2 / MSE) with MSE corresponding to the mean squared error (E v - Svi) 2 between the reference data stream E v and the data stream decoded by the modified decoder
• the metadata can be a pair of values (maximum degradation, potential gain in complexity).
• the system considers as energy constraint a remaining energy level measured for the battery of a mobile terminal.
• the low power filters are finite impulse response filters.
• a metadata M d can be determined from the values of maximum degradation, potential gain, and "low-power" filter coefficient values.
• the system may also include a module for predicting the type or types of frames present in the multimedia stream, the module activation of the simplified filtering strings being activated according to a frame type or their position in the multimedia data stream.
• the decoder is, for example, an H.264 / AVC decoder or an H.265 / HEVC decoder and the video image data stream.
• the invention also relates to a method for decoding data from a multimedia data stream at a terminal by managing the energy required for the decoding, characterized in that it comprises at least the following steps:
• the method generates a metadata corresponding to a pair of values (maximum degradation, potential gain in energy), these values being calculated during a coding step of the multimedia data taking into account the multimedia data E v , decoded data S V 2 by a standard decoder, an estimated energy value E 2 by the standard decoder, decoded data Svi by a modified decoder, the estimated energy E- ⁇ by the modified decoder.
• a metadata is calculated in the following manner:
• PSN RDecoderReference 10 * logi 0 (d 2 / MSE), d corresponding to the maximum amplitude of the pixels, MSE corresponding to the mean squared error (E v - S V 2) 2 between the reference data stream E v and the data stream decoded by the standard decoder Sv2,
• the method takes as an energy constraint an energy level measurement for the terminal's battery.
• the method is, for example, implemented in an H.264 / AVC or H.265 / HEVC decoder and in that the data stream consists of video images.
• FIG. 2 a video coder diagram according to the prior art
• FIG. 3 a block diagram of the system according to the invention
• FIG. 4 a diagram showing the generation of metadata according to the invention
• FIG. 5 a detail of the integration of the modified decoder into the structure of the encoder
• FIGS. 6A and 6B two curves representing the energy saving achieved
• FIGS. 7A and 7B two curves comparing the distortion between the HEVC standard and the modified HEVC standard according to the invention.
• the description will be given by way of example in order to illustrate the principle implemented by the invention for a video stream that a user wishes to download on his terminal in the case of the HEVC standard.
• the final decoder will adapt the best decoding strategy to its means, by deducting metadata M d associated with energy management at the decoding level, the possible gains in energy consumption that can be achieved through the use of modified filtering functions, in particular by simplifying the filters present in the video decoding chain.
• a metadata can represent a gain in decoding complexity which makes it possible to deduce a reduction in energy or instantaneous power by the decoder.
• the metadata M d in the present invention is a distortion / economy pair between the low consumption chains and a standard chain.
• FIG. 3 is a block diagram of the operation of the system according to the invention according to user constraints 301 and / or according to a battery level of a mobile device 302.
• the system receives a stream F m composed of compressed data F and M d metadata generated by a modified encoder detailed in Figure 4, the total stream F m is transmitted to a demultiplexer 303 to generate the compressed data stream and metadata.
• the metadata M d is transmitted to an activation decision module 305 of the filtering chain of a modified video decoder 300.
• the decision module 305 evaluates the opportunity to activate the modified filtering string of the video decoder according to for example, the battery level of the terminal measured by an appropriate module, indicating whether or not it is necessary to make energy gains by how much, and / or user constraints and / or operating constraints, indicating by for example, maximum allowable degradation, user decoding quality requirements, quality degradation, decoding complexity, or energy required for decoding.
• the decision to activate the modified filters, command C d is sent to the modified video decoder 300.
• the compressed data stream F transmitted to the modified video decoder 300 is then decoded by the final video decoder by passing either through a conventional processing chain or a simplified processing chain according to the invention, an example of which is given in FIG. 5.
• the activation decision can be modified when a new metadata is sent for the following sequence.
• One of the operating hypotheses of the system according to the invention is as follows: under the compromise of reducing the quality of the decoded video, significant gains in energy consumption can be made. These gains are achieved in particular through the use of modified filtering functions ( Figure 5, for example).
• FIG. 4 illustrates an exemplary embodiment of a modified video encoder 400 according to the invention making it possible in particular to generate metadata M d used at the level of the modified final decoder.
• the modules known to those skilled in the art and having functions identical to those introduced in FIG. 4 bear the same references and will not be detailed.
• the video encoder 400 comprises for example in parallel a standard video encoder module comprising a reference decoder 402 and a modified encoder module comprising a modified decoder 401. This structure makes it possible to generate the metadata or metadata that will be used at the decoder for the activation of the low-power filter chain.
• the standard video decoder 402 and the modified video decoder 401 each comprise a module making it possible to estimate the energy necessary for the decoding of the video stream and its restitution, for example, a probe of energy or complexity of the flux, 403, 404 .
• the modified encoder module comprising the decoder 401 there is a first modified video output S v i, and a first estimated energy value E i.
• the standard decoder module 402 At the output of the standard decoder module 402 there is a second video output reference S V 2, and a second value of estimated decoding energy E 2 .
• the metadata or metadata are transmitted to a multiplexer and multiplexed by means of a multiplexer 406 with the compressed data stream.
• the stream comprising the stream of compressed data and the metadata or metadata is transmitted to a modified decoder according to the invention.
• FIG 4 shows a video decoder detail of the modified encoder.
• the loop filter is modified, 1 1 6b, as well as the motion compensation module which integrates a module estimating the energy 404.
• the modified loop filter 1 1 6b is, for example, composed of of low consumption filters an anti-block filter or in English “deblocking filter” low consumption 41 0, followed by a filter introducing an offset known as the English abbreviation "SAO" (abbreviation "Sample”) -adaptive Offset ”) 41 1 detailed in Figure 5. It also includes an energy estimation module 403.
• the estimated metadata or metadata can be a GEDQ quality degradation energy gain type metadata, this metadata represents the percentage slope of energy gain per dB of signal-to-noise ratio of the curves shown in FIG. 6A and FIG. 6B, for example , corresponding to the current image or image sequence.
• the metadata can also be a gain in complexity by degradation in GCDQ quality. The complexity is easily calculated thanks to the number of calculations necessary for the decoding.
• the metadata depends on the decoding hardware architecture, including the processor.
• the metadata generator also estimates the decoding degradation introduced by the modified decoder. For that, it can use as metric the power-on-noise ratio or PSNR YU v on the three components Y (Luminance) and (U, V) (Chrominance).
• PSNR YU v the power-on-noise ratio
• Y Luminance
• U, V Chrominance
• dB decibels
• Differences consider, for example, videos, pixel by pixel, ranging from 0 to 255.
• PSNR of the sequence is the average of PSNR on all the images of the video sequence considered.
• PSN R S v2 from the decoded image by the reference decoder 401
• PSN Rsvi from the decoded image by the modified decoder 402.
• the metadata GEDQ (Energy Gain by Degradation in Quality) can be calculated in the following way: MGE / (PSNR S v2 - PSN Rsvi).
• Other metrics measuring the subjective quality of the image can be used as the average opinion score (MOS) in addition to or alternatively to the PSNR metric.
• the previous examples were given by expressing the normalized PSNR as a percentage. This percentage represents the gain between the standard schema and the proposed level. More generally, the coding may use a dedicated scheme to represent this gain on N bits in the message / protocol. Energy gains can follow a linear, logarithmic scale or be coded on a reduced scale to limit the impact on system throughput.
• FIG. 5 represents an example of a modified decoder according to the invention.
• the decoder also includes a low-power filter chain which is activated or not according to the constraints of a user and / or according to the battery level of the mobile terminal and an indicator of activation issued by the activation module receiving at least the metadata.
• One of the principles implemented in the method according to the invention is to use at the level of the final decoder, so-called “low power” filters in addition to the standard video decoder scheme.
• Apply “Low power” loop filtering consists either of not filtering the current data stream, an image, or of using finite impulse response filters whose length is smaller than the filters specified in the standard.
• filter length is meant the degree of complexity of a filter, its window of observation.
• the activation or not of the "low consumption” filters can be decided according to a pattern corresponding to an activation profile or according to a more complex scheme.
• the activation decision module can take activation choices for the "low consumption” filters according to numerous criteria. It can, for example, consider the remaining energy in the system (battery level still available), the real-time decoding constraints, the Dynamic Voltage and Frequency Scaling (DVFS) parameters of the processor concerned to adapt the operating frequency to the maximum. near the situation and rendering quality criteria of the decoded video.
• DVFS Dynamic Voltage and Frequency Scaling
• N smoothing factor set at 9.5 (as in the state of the art).
• the generated coefficients are converted into an integer value with the following formula:
• Filter m (a) PartParty ⁇ Filter m (a) .2 a )
• a is a coefficient that defines the number of bits on which the quantization is performed, the value of a is chosen lower than the quantization coefficient of the norm, currently equal to 6.
• a filter of the standard for the interpolation uses a quantization on six bits, and one obtains a filter including the following coefficients:
• Filter-i / 2 5b (0, 2, -6, 20, 20, -6, 2, 0); the generation of zero coefficients makes it possible to reduce the number of operations performed at the level of the filter and therefore the complexity of the filter.
• the change of the quantifications shows zero coefficients that will be ignored during the filter calculation. In fact, this amounts to reducing the number of coefficients to be processed and therefore reducing the complexity of the filter.
• reducing the implementation complexity consists in rounding the coefficients towards their rounding in basis of 2 the closest.
• This variant embodiment can be implemented regardless of the quantization coefficient used, six for the current standard, or less than six. All the coefficients of the filter or at least the majority of the coefficients are rounded to the power of 2 nearest their value.
• the filter-i / 2 reference luminance interpolation
• Filter 1/2 base2 (-2 °, 2 2 , -2 3 , 2 5 , 2 5 , -2 3 , 2 2 2 , -2 °).
• the compensation will be obtained by the value (2 6 / ⁇ Filter-i / 2 _base2), or more generally by (2 a / ⁇ Filter_ b ase2) with ⁇ Filter_ b ase2 representing the sum of the coefficients of the filter.
• the compressed data stream F coming from the demultiplexer 551 is transmitted to a first entropy decoding module 501, the decoded stream is transmitted to a reverse quantization and inverse transformation module 502, and to a decoding module of the Inter / intra frame type, 503.
• the reconstructed data Fr is transmitted to a loop filter 504 which includes a first filter chain 505 composed of a reference deblocking filter 506 followed by a reference ODS filter. 507 whose output is linked to a video output 508, and a second low power filtering chain 509 which comprises a first low consumption deblocking filter 510 followed by a second low consumption SAO filter 51 1.
• a first switch 512 makes it possible to direct the reconstructed data through the first standard filter chain or the second low power filter chain according to an activation flag C d generated by the activation module 520, a second switch Sending the filtered data to the video output. Without departing from the scope of the invention, it is possible to consider replacing the low-power filters by a line having no filter and ensuring only the transmission of data.
• the video output S v is connected to a not shown display device and a part of the images is also stored in a buffer memory 521 whose output is connected to a motion compensation module 530.
• the motion compensation module comprises a switch 538 receiving the activation flag C d in order to choose the filter chain to use.
• the compensation module 530 comprises a first standard chain 531 comprising a first luminance interpolation module 532 followed by a first chrominance interpretation module 533 whose output is connected to a motion vector compensation module 534.
• a second low-power filter chain 535 comprises, for example, a low power interpolation module 536 followed by a chroma interpolation module 537.
• a first switch 538 receives an activation command C d and allows the passage of the video images. stored either in the low power filter chain or in the standard chain, and a second switch 539 connecting either the standard chroma interpolation module 531 or the low power chroma interpolation module 535 to the motion compensation module.
• the output of the compensation module is transmitted to the intra / inter mode selection module to be applied for the reconstruction of the data.
• the switches of the low-power decoding chain can be controlled independently, the low-power loop filter chain can be activated independently of the low-power string of the interpolation filter. It is also possible to envisage a scheme in which the low consumption filters of the loop filter are replaced by a data transmission line without any action on the data, this depending on metadata or metadata and / or operating constraints set by the user.
• the activation level specifies the proportion of images for which "low power" filtering is applied. For example, an activation rate of 25% means that one in four decoded images are decoded with the "low power" filtering mode. These images are evenly distributed over a period of time. For example, by choosing a period of 12 images, image indexes from 1 to 12 and an activation rate of 25%, images 2, 6 and 10 will be decoded with the "low power” filtering mode and the images 1, 3, 4, 5, 7, 8, 9, 1 1 and 12 will be decoded with the reference filtering mode of HEVC, for example.
• the activation pattern thus defined can be written [2, 6, 10]. In this example, the patterns [1, 5, 9], [3, 7, 1 1] and [4, 8, 12] are also valid.
• FIG. 6B shows that the effect of the activation level on the signal-to-noise ratio of the decoded video depends on the compression ratio of the video.
• FIGS. 7A, 7B illustrate the consequence of the activation level of the low-power filters on the signal-to-noise ratio of the decoded video.
• the activation level is used to select the low power filters to apply. A finer adjustment of the filtering parameters is possible, for example by choosing the images on which we will apply or not the low power filtering. For this, the decision module receives information on the type of frames contained in the current data stream. A decision will therefore be made at the frame level to decide whether the method applies unmodified filters or modified filters.
• image types I (intra), P (predicted) and B (bidirectional)
• I intra
• P predicted
• B bidirectional
• the information 550 of the decoded picture type is transmitted to the activation decision block by the entropy decoding.
• Interpolation and "low power" loop filters can be activated at the same activation level and activation pattern. It is also possible to independently activate interpolation filters and loop filters. In this case, a different pattern and / or profile is used for each filter.
• the functional module is adapted to decide when and which frames need to be subjected to simplified filters.
• the following table indicates frame numbers on which the simplified filter method applies:
• the system according to the invention allows an adjustment of the energy power consumed with prior knowledge of the video quality, without increasing the complexity of the decoder side.
• the data part of the stream or bitstream remains unchanged from the state of the art.
• all the decoders reading this stream can adapt their decoding to their battery level.
• the filters of the standard part of the encoder are not modified, unlike an implementation which consists in using the same filters for the encoder side and the decoder side.

Landscapes

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

Applications Claiming Priority (3)

Publications (1)

Family

ID=52779728

Family Applications (1)

Country Status (5)

Families Citing this family (1)

Family Cites Families (6)

• 2014
  • 2014-10-17 FR FR1460043A patent/FR3027480B1/fr not_active Expired - Fee Related
• 2015
  • 2015-02-10 FR FR1551085A patent/FR3027481A1/fr not_active Withdrawn
  • 2015-10-16 CN CN201580056024.1A patent/CN107211127A/zh active Pending
  • 2015-10-16 EP EP15785076.9A patent/EP3207702A1/de not_active Withdrawn
  • 2015-10-16 WO PCT/EP2015/073964 patent/WO2016059196A1/fr active Application Filing
  • 2015-10-16 US US15/519,333 patent/US10531096B2/en not_active Expired - Fee Related

Non-Patent Citations (2)

Also Published As

Similar Documents

Legal Events