BR112018009921B1 - DEVICE AND METHOD FOR CONFIGURING A SEQUENCE OF FILTERS FOR VIDEO CODING - Google Patents

Abstract

VIDEO CODING DEVICE AND METHOD FOR CONFIGURING A SEQUENCE OF FILTERS FOR VIDEO CODING. The present invention relates to a video coding device, which comprises a sequence of filters that are configurable by one or more primary parameters and one or more secondary parameters, and a filter controller configured to adjust the one or more secondary parameters with based on one or more primary parameters and based on an intensity criterion of the filter sequence.

Description

FIELD OF TECHNIQUE

[001] The present invention relates to a video coding device and a method for configuring a sequence of filters for video coding.

[002] The present invention also relates to a computer-readable storage medium that stores a program code, the program code comprising instructions for executing a method for configuring a sequence of filters for video coding.

BACKGROUND

[003] Digital video communication and storage applications are implemented by a wide range of digital devices, for example, digital cameras, cellular radio telephones, laptops, broadcast system, video teleconferencing systems, etc. One of the most important and challenging tasks of these applications is video compression. The task of video compression is complex and constrained by two contradictory parameters: compression efficiency and computational complexity. Video coding standards, such as ITU-T H.264/AVC or ITU-T H.265/HEVC, provide a good trade-off between these parameters.

[004] Prior art video coding standards are mainly based on partitioning a source image into blocks. The processing of these blocks depends on their size, spatial position and a coding mode specified by an encoder. Coding modes can be classified into two groups according to the type of prediction: intra- and inter-prediction modes. Intraprediction modes use pixels from the same image to generate reference samples to calculate prediction values for the pixels in the block being reconstructed. Intraprediction could also be referred to as spatial prediction. The interprediction modes are designed for temporal prediction and use reference samples from previous or nearby images to predict the pixels of the current image block.

[005] Due to different types of redundancy, the prediction processes for intra- and intercoding are different. Intraprediction typically constructs a one-dimensional temporary store of reference samples. Interprediction typically uses subpixel interpolation of a two-dimensional reference pixel array. To improve prediction results, additional processing can be used for both intra- and inter-coding (e.g., smoothing reference samples for intra-prediction, sharpening reference blocks for inter-prediction).

[006] The recently adopted standard ITU-T H.265/HEVC (ISO/IEC 23008-2:2013, "Information technology - High efficiency coding and media delivery in heterogeneous environments - Part 2: High efficiency video coding", November 2013 ) declares a set of prior art video coding tools that provide a reasonable trade-off between coding efficiency and computational complexity.

[007] Similar to the ITU-T H.264/AVC video coding standard, the HEVC/H.265 video coding standard provides a division of the source image into blocks, for example, coding units (CUs). Each of the CUs can be further divided into or smaller CUs or prediction units (PUs). A PU can be intra or interpredicted according to the type of processing applied to the PU pixels. In the case of interprediction, a PU represents an area of pixels that is processed by motion compensation using a motion vector specified for a PU. For intraprediction, the PU specifies a prediction mode for a set of transform units (TUs). A TU can have different sizes (e.g. 4x4, 8x8, 16x16 and 32x32 pixels) and can be processed in different modes. For a TU transform a coding is being performed, that is, the prediction error is being transformed with a discrete cosine transform (DCT) and quantized. As a result, the reconstructed pixels contain quantization noise and blocking artifacts that can influence the prediction accuracy.

[008] In order to reduce this influence on intraprediction, a reference pixel filtering was adopted for HEVC/H.265. For interprediction, reference pixels are calculated using subpixel interpolation. A smoothing of reference pixels in the case of motion compensation can be combined with anti-discontinuity filtering of subpixel interpolation process.

[009] Adaptive mode intraprediction smoothing techniques were presented. Smoothing filtering depends on the selected intraprediction mode and a flag encoded in the video bitstream. Depending on the intraprediction mode defined for a block, reference samples could be either smoothed by a filter or used without being modified. For the case when the reference samples are smoothed, the selection of the smoothing filter can also be based on the intraprediction mode. Furthermore, this selection could be performed according to the value of a reconstructed flag from a bitstream.

[0010] The current HEVC/H.265 standard partially uses this technique. Specifically, filter smoothing is turned off for various intramode and block size combinations.

[0011] A reference sample adaptive filter (RSAF) has been suggested as an extension of the reference sample filter adopted for the HEVC/H.265 standard. This adaptive filter segments the reference samples before smoothing to apply different filters to different segments. Additionally, a data hiding procedure was used to flag a smoothing flag. A simplified version of the adaptive filter for reference samples was adopted for the Joint Exploration Model 1 (JEM1) which contains several other tools which use smoothing, including: - Four-lead intra-interpolation filter, - Boundary prediction filters, and/or - Multi-parameter Intraprediction (MPI), which can be replaced by Position Dependent Intraprediction Combination (PDPC)

[0012] Problems with the above methods include high signaling effort and over-smoothing of the video during encoding or decoding.

SUMMARY OF THE INVENTION

[0013] The object of the present invention is to provide a video coding device and a method for configuring a sequence of filters for video coding, wherein the video coding device and the method for configuring a sequence of filters overcome one or more of the problems mentioned above.

[0014] A first aspect of the invention provides a video coding device, which comprises: - a sequence of filters that are configurable by one or more primary parameters and one or more secondary parameters, and - a filter controller configured to adjust the one or more secondary parameters based on the one or more primary parameters and based on a filter sequence strength criterion.

[0015] The video coding device of the first aspect may adjust the one or more secondary parameters so that an intensity criterion of a filter sequence is met. For example, as outlined below, the intensity criterion may refer to a global smoothing and the filter controller may adjust the secondary parameters so that a global smoothing of the sequence of filtering steps is not too high and not too low. In other words, the filter controller may be configured to ensure that an intensity criterion of a filter sequence is within a predetermined range.

[0016] The video encoding device of the first aspect may be configured for video encoding and/or decoding.

[0017] The filter controller may be configured to only partially adjust one or more of the secondary parameters. For example, a secondary parameter may be read from a bitstream or a user setting and adjusted, for example, within a certain range, by the filter controller. In other embodiments, the filter controller may also be configured to ignore a value of a secondary parameter that was otherwise derived, for example, from a bit stream or a user setting.

[0018] In the prior art, non-harmonious working of video coding tools including RSAF can result in oversmoothing which - reduces overall coding performance as a next filter does not take into account the effects caused by previous filters, and/or - increases the overall computational complexity since all the above mentioned filters are constantly turned on.

[0019] This can be avoided with the video coding device of the first aspect. Furthermore, as secondary parameters do not need to be stored in a bit stream, in certain embodiments a signaling effort can be reduced.

[0020] In a preferred embodiment, the video coding device of the first aspect can solve the aforementioned problem of over-smoothing by adjusting the parameter of these filters of the sequence of filters that use smoothing. This adjustment can be achieved, for example, by introducing a flag or add conditions for filters that use smoothing. The flag and/or conditions can be used - to turn on and off the tools' smoothing mechanisms, and/or - to change the intensity of filter smoothing (for example, switching from strong filters to weak ones).

[0021] The filter controller of the video encoding device of the first aspect may be configured to control not just a single filter, but multiple filters. This can be considered as a mechanism to harmonize different filters that impact intraprediction results, for example, by smoothing. Specifically, the filter controller may be configured to make the following adjustments: - a sample processing mechanism may be changed as some filter modules may be turned off subject to a flag value and/or meeting some conditions; - a new filter module can be introduced to provide control over all filters that impact the intraprediction results by smoothing them.

[0022] In a first implementation of the video coding device according to the first aspect, the intensity criterion comprises: - a smoothing criterion, - a ratio of an amplification factor to a high frequency domain and a amplification is in the low frequency domain, and/or - a ratio of a contrast value before filtering and a contrast value after filtering.

[0023] This allows you to optimize the filter sequence for one or more of the above criteria. As outlined above, preferably adjustments by the filter controller can be carried out so that the intensity criterion of the filter sequence is within a certain range, for example, the predetermined range.

[0024] In a second implementation of the video encoding device according to the first aspect, the one or more primary parameters are predetermined parameters, specifically parameters that are predetermined from an encoded bit stream, a user setting and/or a parameter lookup loop in an encoding device.

[0025] Determining the secondary parameters from the primary parameters, where, for example, only the primary parameters are predetermined, has the advantage that signaling effort can be reduced. For example, a bitrate can be reduced if secondary parameters are not stored in the bitstream but can be derived from the primary parameters in the bitstream.

[0026] In a third implementation of the video coding device according to the first aspect, the filter sequence comprises one or more primary filters that are configurable by one or more primary parameters and one or more secondary filters that are configurable by a or more secondary parameters, wherein the one or more primary filters are located in the filter sequence before the one or more secondary filters.

[0027] Adjusting filter parameters in later stages has the advantage that an effect of earlier stage filters can possibly be undone or at least not further intensified. For example, if previous stage filters generated a certain amount of smoothing, it can be ensured that later filter stages do not increase this smoothing effect. For example, a smoothing flag of later filter stages may be turned off.

[0028] In other embodiments of the invention, the primary parameters may be related to later filter stages and the secondary parameters to earlier filter stages.

[0029] In a fourth implementation of the video coding device according to the first aspect, the filter sequence comprises: - a reference sample filter configured to adaptively filter one or more neighboring samples of a current video block to obtain one or more reference samples, and - an interpolation filter configured to predict one or more samples of the current video block using an interpolation of the one or more reference samples, wherein the one or more primary parameters comprise a reference parameter of the reference sample filter and the one or more secondary parameters comprise an interpolation filter selection parameter, wherein the interpolation filter is configured to use an interpolation method in accordance with the selection parameter.

[0030] The filter sequence of the video coding device of the fourth implementation may be, for example, a filter sequence for intraprediction.

[0031] Preferably, the filter controller is configured to determine the selection parameter based on the reference parameter. This proved to be an effective method for improving a global filter strength criterion.

[0032] In a fifth implementation of the video coding device according to the first aspect, the one or more primary parameters include a reference sample filter flag of a reference sample filter and the one or more secondary parameters include a filter strength parameter of an intraprediction interpolation filter.

[0033] Preferably, the filter controller is configured to determine the filter intensity parameter based on the filter flag of reference samples. This proved to be an effective method for improving a global filter strength criterion.

[0034] In a sixth implementation of the video coding device according to the first aspect, the filter sequence comprises a boundary smoothing filter which is configured to perform boundary smoothing for one or more transform units belonging to to one or more prediction units that meet a size constraint.

[0035] This has the advantage that a boundary smoothing filter reduces the visibility of blocking artifacts for reconstructed blocks. A prediction unit typically contains an image region of a certain type: edge, texture, smooth region, etc. However, for larger PUs the probability of a smooth region is higher. For smooth regions, blocking artifacts are more crucial and therefore, boundary smoothing for large PUs is more preferable than for smaller ones. Consequently, it is proposed to restrict the boundary smoothing with the size of PU (for example, by the size of 32x32 pixels). Using this constraint it is possible, on the one hand, to avoid undesirable blurring for non-smooth regions of smaller PUs and, on the other hand, to reduce blocking artifacts for larger PUs. This allows to improve both objective and subjective quality compared to the case when boundary smoothing is predefined for intraprediction.

[0036] In a seventh implementation of the video coding device according to the first aspect, the one or more primary parameters include a direction parameter of a predicted block filter and the one or more secondary parameters include a link parameter. turns off a boundary smoothing filter.

[0037] Preferably, the filter controller is configured to determine the on-off parameter of the boundary smoothing filter based on the predicted block filter direction parameter. This proved to be an effective method for improving a global filter strength criterion.

[0038] In an eighth implementation of the video coding device according to the first aspect, the filter sequence comprises: - a subpixel interpolation filter configured to adaptively filter samples of a reference block to obtain an interpolated reference block , and - a low-pass and/or high-pass filter configured to smooth and/or sharpen the interpolated reference block to obtain a filtered reference block, wherein the one or more primary parameters comprise an interpolation parameter of the interpolation filter and the one or more secondary parameters comprise a sharpening and/or smoothing filter selection parameter, wherein the derivation of the secondary parameters is determined by the interpolation filter parameter.

[0039] The filter sequence of the video coding device of the eighth implementation may be, for example, a filter sequence for interprediction.

[0040] Preferably, the filter controller is configured to determine the selection parameter of the sharpening and/or smoothing filters based on the interpolation parameter. This proved to be an effective method for improving a global smoothing criterion.

[0041] It is noted that a filter sequence smoothing criterion may also be position dependent. For example, a filter could introduce strong smoothing in one region and strong sharpening in another region. Thus, the filter controller could be configured to adjust different secondary parameters for different regions of one or more image frames of the video.

[0042] In a ninth video coding implementation of the video coding device of the first aspect, the filter sequence comprises an adaptive loop filter configured to utilize a selected codebook to indicate one or more filter coefficients in a stream of bits, wherein the filter controller is configured to select the codebook from among a plurality of codebooks based on the one or more primary parameters.

[0043] The adaptive loop filter coefficients being encoded by a plurality of codebooks takes advantage of prior information about the processing that has been applied and an adaptive loop filter input signal. If smoothing has already been applied to a signal being processed by the adaptive loop filter, it is known that the adaptive loop filter can only introduce high pass filtering. Therefore, some of the adaptive loop filter coefficient combinations have become unavailable. This property is used to maintain two or more or more codebooks accordingly for at least two cases: when an adaptive loop filter is applied to an already smoothed input signal and when no smoothing has been applied to the input of the adaptive loop filter.

[0044] For example, the plurality of codebooks may comprise a first and a second codebook, wherein the first codebook only comprises coefficients for both high-pass and low-pass filtering and the second codebook only comprises coefficients for for low pass filtering.

[0045] In a preferred embodiment, the plurality of codebooks comprises more than two codebooks, wherein preferably the different codebooks of the plurality of codebooks correspond to different filtering strengths of filters applied after ALF.

[0046] In a tenth implementation of the video coding device according to the ninth implementation of the first aspect, the filter sequence further comprises: - a deblocking filter configured to process vertical edges based on a vertical filter strength parameter and/or horizontal borders based on a horizontal filter intensity parameter, and - a sample adaptive offset filter, SAO, configured to classify the pixels and add offset values to the pixels according to an SAO class parameter, wherein the one or more primary parameters comprise the SAO class parameter and an SAO filter SAO type parameter and the one or more secondary parameters comprise the horizontal filter intensity parameter and the vertical intensity parameter wherein the filter controller is configured to derive secondary parameters based on the SAO class parameter and/or the SAO type parameter of the SAO filter, and/or the filter controller is configured to select the codebook from among the plurality of codebooks based on the SAO type parameter.

[0047] The filter sequence of the video coding device of the tenth implementation may be, for example, a looped filter sequence.

[0048] A further implementation of the video coding device of the first aspect refers to a video coding device of one of the previous implementations of the first aspect, wherein the secondary vertical and horizontal edge deblocking filter strength parameters are different and wherein a ratio of the vertical deblocking filter strength and the horizontal deblocking filter strength is adjusted based on said SAO class.

[0049] A further implementation of the video coding device of the first aspect refers to a video coding device of one of the previous implementations of the first aspect, wherein the filter sequence does not contain an adaptive loop filter or wherein the filter controller does not adjust the parameters of a looped filter.

[0050] A further implementation of the video coding device of the first aspect refers to a video coding device of one of the previous implementations of the first aspect, wherein the filter sequence does not contain a deblocking filter or wherein the filter controller does not adjust the parameters of an unblocking filter.

[0051] A second aspect of the invention relates to a method for configuring a sequence of filters for video coding, the method comprising: - adjusting one or more secondary parameters based on one or more primary parameters and based on a criterion intensity of the filter sequence, and - configure the filter sequence with the primary and secondary parameters.

[0052] In a first implementation of the method of the second aspect, the method further comprises an initial step of determining the one or more primary parameters of a bit stream.

[0053] The methods according to the second aspect of the invention can be performed by the video coding device according to the first aspect of the invention. Additional features or implementations of the method according to the second aspect of the invention can perform the functionality of the video coding device according to the first aspect of the invention and its different forms of implementation.

[0054] A third aspect of the invention relates to a computer-readable storage medium that stores a program code, the program code comprising instructions for executing the method of the third aspect or one of the implementations of the third aspect.

BRIEF DESCRIPTION OF THE DRAWINGS

[0055] To illustrate the technical characteristics of embodiments of the present invention more clearly, the accompanying drawings provided to describe the embodiments are briefly introduced below. The accompanying drawings in the following description merely show some embodiments of the present invention. Modifications of these embodiments are possible without departing from the scope of the present invention as defined in the embodiments.

[0056] Figure 1 is a block diagram illustrating a video encoding device in accordance with an embodiment of the present invention,

[0057] Figure 2 is a flowchart of a method for configuring a sequence of filters for video coding in accordance with a further embodiment of the present invention,

[0058] Figure 3 is a schematic structure of a sequence of filters for intraprediction according to an additional embodiment of the present invention,

[0059] Figure 4 is a flowchart of a method for configuring a sequence of filters for controlled reference sample adaptive filtering in accordance with a further embodiment of the present invention,

[0060] Figure 5 is a flowchart of a method for configuring a sequence of filters for intraprediction in accordance with a further embodiment of the present invention,

[0061] Figure 6 is a flowchart of a method for configuring a sequence of filters for boundary smoothing in accordance with a further embodiment of the present invention,

[0062] Figure 7 is a schematic structure of a sequence of filters for intraprediction according to an additional embodiment of the present invention,

[0063] Figure 8 is a flowchart of a method for configuring a sequence of filters for intraprediction in accordance with a further embodiment of the present invention,

[0064] Figure 9 is a flowchart of a method for intraprediction with a filter control module on the decoder side according to a further embodiment of the present invention,

[0065] Figure 10 is a schematic structure of a serial-parallel interprediction modality with a filter control module according to a further embodiment of the present invention,

[0066] Figure 11 is a flowchart of a method for configuring a filter sequence for interprediction with a decoder-side filter control module in accordance with a further embodiment of the present invention,

[0067] Figure 12 is a schematic structure of a loop filter chain according to a further embodiment of the present invention,

[0068] Figure 13 is a flowchart of a method for configuring a sequence of filters, wherein the unblocking filtering depends on one or more SAO parameters, in accordance with a further embodiment of the present invention,

[0069] Figure 14 is a flowchart of an additional method for configuring a sequence of filters, wherein the deblocking filtering depends on SAO parameters and wherein the filtering occurs on a decoder side, in accordance with a further embodiment of the present invention,

[0070] Figure 15 is a flowchart of a method for configuring a filter sequence, wherein the method adjusts one or more ALF parameters in accordance with one or more SAO parameters on a decoder side, in accordance with an additional embodiment of the present invention, and

[0071] Figure 16 is a flowchart of a configuration of a filter sequence, in which the method adjusts one or more ALF parameters in accordance with one or more SAO parameters on a decoder side, in accordance with an additional embodiment of the present invention.

DETAILED DESCRIPTION OF MODALITIES

[0072] Figure 1 shows a video coding device 100, which comprises a filter sequence 110 and a filter controller 120.

[0073] The filter sequence 110 is configurable by one or more primary parameters and one or more secondary parameters. For example, a first set of filters of a filter sequence may be configurable by the primary parameters and a second set of filters of a filter sequence may be configurable by the second set of filters. The first and second set of filters can be overlapped.

[0074] The filter controller 120 is configured to adjust the one or more secondary parameters based on the one or more primary parameters and based on a strength criterion of the filter sequence 110. Specifically, the filter controller 120 may be configured to adjust the one or more secondary parameters partially based on the one or more first parameters. For example, a value of a secondary parameter may be partially based on a predetermined value, for example, from a bit stream, and partially based on tuning based on the primary parameters.

[0075] Figure 2 shows a method 200 for configuring a sequence of filters for video encoding. The method 200 comprises a first step 210 of adjusting one or more secondary parameters based on one or more primary parameters and based on a strength criterion of the filter sequence. The method further comprises a second step 220 of configuring the filter sequence with the primary and secondary parameters.

[0076] An intraprediction procedure may be part of a hybrid video coding toolchain on an encoder side and/or on a decoder side. Similarly, an interprediction procedure may comprise a sequence of filters (e.g., interpolation filters and so-called prediction filters) that are potentially capable of causing either oversmoothing or oversharpening of a block used as a reference that, in fact, is a analogous to an intrapredict block for interprediction.

[0077] A filter sequence may comprise, for example, one or more of the following filters: - Reference sample smoothing (e.g. RSAF), - Interpolation filtering for intraprediction, - Intrapredict block filtering (e.g. MPI or PDPC), and/or - Border smoothing.

[0078] These filters can impact the smoothing intraprediction results.

[0079] Figure 3 illustrates a filter sequence 300 comprising a filter control module 360 for adjusting filtering parameters at different intraprediction stages. The 360 filter module controller is a filter controller.

[0080] Intraprediction parameters may include, but are not limited to: - a size of a prediction unit, - a size of a block being predicted, - an intraprediction mode, - multi-parameter intramode index, and/ or - a reference sample filtering flag.

[0081] One or more of the above parameters may be primary or secondary parameters.

[0082] Apart from the filter control module 360, the filter sequence 300 comprises a reference sample smoothing unit 310, an intraprediction unit 320, a predicted block filter unit 330, and the boundary 340. The reference sample smoothing unit 310 is configured to be provided with one or more neighboring samples 302 as input. This is configured to smooth and/or further process the one or more neighboring samples 302 to obtain one or more reference samples 312, which are provided to the intraprediction unit 320 as input. The intraprediction unit 320 comprises an interpolation filter 322. The intraprediction unit 320 provides its output 324 as input to the predicted block filter 330.

[0083] The predicted block filter 330 is configured to compute one or more predicted blocks 332, which are provided to the boundary smoothing unit 340 as input. The boundary smoothing unit 340 outputs 342 one or more intrapredicted blocks 350.

[0084] A video coding device comprising a sequence of filters 300 may be configured to use implicit or explicit filter signaling of reference samples selectively, that is, only for those TUs that meet specific conditions.

[0085] The filter control module 360 may be configured to read the intraprediction parameters 362 as primary parameters. This can be configured to derive secondary parameters based on these primary parameters.

[0086] Quadtree partitioning results can be used as an indication of reference sample filter selection using explicit or implicit signaling. Specifically, if the size of a PU is greater than a threshold value (e.g., 32x32), the reference sample filter flag is set to zero. This designation cancels the conditions of the prior art. If a PU size condition is true, only the "NO FILTER" and "APPLY WEAK FILTER" options could be selected according to the PU size and/or intramode conditions.

[0087] Figure 4 is a flowchart of a method 400 for configuring a sequence of filters for controlled reference sample adaptive filtering.

[0088] Method 400 comprises a first step 402 of evaluating a condition relative to a size of a prediction unit. If the condition evaluation is true, the method continues to step 404, where a reference sample filter flag is derived. If the evaluation of the condition relative to a size of a prediction unit is false, the method continues to step 406, in which a reference sample filter flag is set to false. Step 404 or step 406 is followed by a step 408, in which one or more conditions relating to an intramode and a block size are evaluated.

[0089] If the evaluation result of step 408 is false, the method continues to step 410, in which the reference sample filter flag is evaluated. If false, the method continues to step 414, where the reference sample adaptive filter is adjusted to not apply a filter. If in step 410 the flag evaluates to true, in step 416 a weak filter is applied. Alternatively, if the condition evaluation in step 408 evaluated as true, the method continues to step 412, in which the reference sample filter flag is evaluated. If the evaluation is false, in step 416 a weak filter is applied. If the reference samples filter flag in step 412 evaluates to true, a strong filter is applied in step 418.

[0090] The "APPLY WEAK FILTER" and "APPLY STRONG FILTER" stages can select a filter from a predefined filter set, as soon as the selection of a specific filter from the set is mapped to the intramode and a block size being predicted. This specific embodiment, which has only 3 filters, does not mean that the number of filters within the filter set cannot be extended to an arbitrary amount (for example, 5 states including the "NO FILTER" state, a "FILTER WEAK", a "STRONG FILTER" and two intermediate filters).

[0091] In directional intraprediction, a pixel value of the predicted block and a projection onto the left and top block boundaries is calculated. However, the projection can have a fractional position, that is, it can fall between real positions of reference samples on the boundary. A weighted sum of values from adjacent reference samples is calculated to determine a value of a sample from the intrapredicted block. This process is in fact a two-lead linear interpolation filter, which can be further extended to a four-lead interpolation filter.

[0092] Four-lead interpolation filters can be used to improve directional intraprediction accuracy. In HEVC, a two-lead linear interpolation filter was used to generate the intraprediction block in directional prediction modes (i.e., excluding Planar and DC predictors). Alternatively, two types of four-lead linear interpolation filters can be used: Cubic interpolation filters for 4x4 and 8x8 blocks, and Gaussian interpolation filters for 16x16 and larger blocks. The filter parameters are fixed according to the block size, and the same filter is used for all predicted pixels, in all directional modes.

[0093] In HEVC, after the intraprediction block has been generated for VER and HOR intramodes, the leftmost column and the topmost row of the prediction samples are additionally adjusted, respectively. This can be further extended to several diagonal intramodes, and boundary samples up to four columns or rows are additionally fitted using a two-lead (for intramode 2 & 34) or three-lead (for intramode 3-6 & 30-33) filter. .

[0094] Figure 4 and Figure 5 illustrate two modalities for synchronizing a selection of an interpolation filter type with a reference sample filtering process. Both modalities consider that two types of interpolation filter could be applied: weak and strong, for example, a Gaussian filter is used for predicted blocks 16x16 and larger and for other block sizes a Cubic filter is selected. For both modalities, an interpolation filter selection can be harmonized with a reference sample filtering process.

[0095] Figure 5 is a flowchart of a method 500 for configuring an interpolation filter for intraprediction.

[0096] Method 500 comprises a first step 502 of deriving a filter flag from reference samples. In step 504, the reference sample filter flag is evaluated. If this evaluates to true, the method continues to step 506 where a condition regarding a size of a transform block is evaluated. If the condition evaluates to false, the method continues to step 508, where the weak intra-interpolation filter is applied. Similarly, if the reference sample filter flag evaluates to false at step 504, the method also continues at step 508. If the condition regarding a size of a transform block evaluates at step 506 to true, the method continues in step 510, where the strong intra-interpolation filter is applied.

[0097] The embodiments of Figure 4 and Figure 5 have differences with respect to filter flag derivation from reference samples. In the embodiment of Figure 4, the reference sample filter flag is true if the block predicted according to the condition has an option for different reference sample filters. This filter selection can be signaled explicitly or implicitly (i.e., by mapping to prediction mode or using data hiding on quantized residual data). For the embodiment of Figure 4, the actually selected reference filter value is not considered when deriving the filter flag from reference samples. But if the reference samples filter selection is performed for the predicted block on the encoder side and the selection is signaled to the decoder explicitly or implicitly, the reference samples flag value is true. Otherwise, if a predicted block has a predefined reference filter or has no reference samples filtering the reference samples the flag value is false.

[0098] The embodiment of Figure 5 uses a value from a reference filter selection as a reference sample flag value. If a strong filter is selected for the reference samples (such as {1 2 1}, or 5-lead filter), a reference samples flag value is assigned to true. And similar to the first embodiment, the reference samples flag value is false for predicted blocks that have no reference filtering or a weak reference samples filter has been selected.

[0099] The beneficial effects of the above-described modalities are achieved by harmonizing reference sample filtering and intraprediction interpolation processes. It can be observed that these modalities prevent the predicted blocks from being greatly smoothed.

[00100] Figure 6 is a flowchart of a method 600 for configuring a sequence of filters for boundary smoothing.

[00101] Method 600 comprises a first step 602 of determining whether a predicted block filter is directional. If so, the method continues to step 604, in which a block size condition is evaluated. If the block size condition evaluates to false or the predicted block filter directionality evaluates to false, the method continues to step 606, in which an intramode condition is evaluated. If the intramode condition evaluates to true, the method continues to step 608, where a boundary smoothing is applied. Otherwise, and if the block size condition in step 604 evaluates to true, no boundary smoothing is applied.

[00102] A boundary smoothing can be applied if an intra-prediction mode is selected to be a DC, horizontal, vertical or diagonal. The proposed invention synchronizes boundary smoothing with the selection of a filter for a predicted block. Specifically, the directionality of a predicted block filter is used to make the decision on whether to apply a boundary smoothing or not. Directionality is a property of a spatial filter that indicates the difference in filtering intensity between filtering directions. For example, if a two-dimensional filter has the same intensity in both vertical and horizontal directions, this filter is non-directional. Specifically, for non-directional filters a boundary smoothing is not applied. Multi-parameter intraprediction can be an example of a predicted block filter. If this technique is used as a predicted block filter, the first condition in Figure 6 could be formulated as "MPI index is greater than one".

[00103] For the opposite case when a predicted block filter is directional, the invention considers another restriction. If a size of a block being filtered is smaller than 32 pixels, boundary smoothing is skipped for this block despite the directionality of the predicted block filter.

[00104] Figure 7 is a schematic structure of a sequence of filters 700 for intraprediction. The filter sequence 700 can be used to encode or decode a video.

[00105] The filter sequence 700 comprises a filter control module 760 that is configured to adjust the parameters of various filters. Specifically, the filter sequence 700 receives as input one or more neighboring samples 702. These neighboring samples 702 are provided as input to a reference sample adaptive filter, SAF, 710, which represents a first filter of the filter sequence 700. The adaptive reference sample filter 710 generates one or more reference samples 712, which are provided as an input to the intraprediction unit 720. The intraprediction unit 720 comprises a set of 4-lead interpolation filters 722, which are configurable by one or more interpolation filter parameters.

[00106] An output 724 of the intra-prediction unit is provided as input to the boundary prediction filters 730. An output 732 of the boundary prediction filters is provided as input to a multi-parameter intra-prediction / multi-parameter-dependent intra-prediction combination unit. position 740. Unit 740 generates as output 742 one or more intrapredicted blocks 750.

[00107] The reference sample adaptive filter 710, the intraprediction unit 720, the boundary prediction filters 730, and the multi-parameter intraprediction/position-dependent intraprediction combination unit 740 may each be configurable by one or more more parameters, which can be adjusted by the 760 filter control module.

[00108] Multi-parameter intraprediction (MPI) is a post-processing for intraprediction which invokes additional smoothing with a decoded boundary. This can be implemented as PMPI [i,j] = (aPHEVC [i, j] + βPMPI [i “ 1, j] + 1PMPI [i, j ~ 1]+ δPMPI [i “ 1, j ~ 1] + 4 ) >> 3

[00109] where outside the block P^ [z, j ] is equal to the reconstructed signal PMPI[i, j] = REC[z, j] if i < 0|| j < 0.

[00110] The intensity of this post-processing (parameters a + β+Y + δ = 8) can be controlled at a CU level and signaled with up to 2 bits.

[00111] Position Dependent Intraprediction Combination (PDPC), which can replace MPI, is a content adaptive post-processing for intraprediction that invokes the above-described intraprediction combination with an unfiltered boundary. This can be implemented as follows: pred(í,j) = (a • temp(i,j) + β • temp(í — 1,j) + y • temp(i,j — 1) + 4) >> 3

[00112] Here (i,y) indicates the sample location relative to the top left corner, temp(i,j) is equal to the Intraprediction described above for i>0,j>0 and equal to the unfiltered reference for i = — 1 ,j = —1 .

[00113] The intensity of this post-processing can be controlled by parameters α + β + y = 8. Different sets of {α, β, y} make up the dictionary summarized in Table 1. The intensity of post-processing smoothing is different for blocks coded as 2N*2N and NxN. The same post-processing can be applied to both luminance and chrominance blocks within the CU. Table 1: Dictionary of post-processing parameters in combined intraprediction

[00114] The Intra Combined Index which determines post-processing is signaled with 2 bits at a CU level. This syntax element is not flagged if the left or top boundary of CU is an image boundary. A zero value of this index indicates that no post-processing is used.

[00115] If each tool has a flag to turn it on and off and the RDO procedure is performed not for each tool separately but together for all tools that use smoothing, the oversmoothing problem can be overcome. However, this solution may have a disadvantage of redundant signaling which may decrease overall encoding performance.

[00116] Figure 8 is a flowchart of a method 800 for configuring a sequence of filters for intraprediction.

[00117] Method 800 comprises a first step 810 of determining whether reference sample adaptive filtering, RSAF, is in use. For example, this can be determined by evaluating an RSAF flag. If it is determined that RSAF is in use, in step 812 reference sample adaptive filtering is applied. Subsequently, in step 814, an interpolation using a 4-lead cubic filter is applied. This can be achieved by adjusting an interpolation mode parameter of a filter sequence interpolation filter for intraprediction for the 4-lead cubic filter.

[00118] If it is determined in step 810 that RSAF is not in use, in step 816, interpolation using a 4-lead filter set is applied. Specifically, this may be a predetermined 4-lead filter set. Adjusting the interpolation filter to use the 4-lead filter set can be achieved, for example, by setting an interpolation parameter of the interpolation filter to "4-lead filter set".

[00119] In a further step 820, an MPI index variable i is set to 0. Subsequently, in step 822, it is determined whether the variable is greater than 1. If so, in step 824 it is determined whether a size of current prediction unit, PU is greater than 32. If so, in step 826, a boundary prediction filtering is applied. A boundary prediction filtering is also applied if in step 822 it is determined that i is not greater than 1. If in step 824 it is determined that the current PU size is not greater than 32, the method continues in step 828 with combination of multi-parameter intraprediction/position-dependent intraprediction.

[00120] Subsequently, in step 830, a rate distortion cost, RD cost Jt for the current configuration is calculated. Specifically, the current setting may correspond to a current value of the MPI index i. In other embodiments, also other parameters are varied and an RD cost for the different parameter settings can be determined.

[00121] In step 832 it is determined whether an MPI index i is greater than or equal to 3. If not, the MPI index is increased by 1 in step 834 and the method continues in step 822. If the MPI index is greater than or equal to 3, the method continues to step 836 selecting a better MPI index. This can be achieved by choosing the MPI index which corresponds to the lowest RD Jt cost.

[00122] Figure 9 is a flowchart of a method 900 for configuring a sequence of filters for intraprediction with a filter control module on the decoder side.

[00123] An interprediction mechanism may comprise, for example, the following filters: - a regular filter based on Lagrangian interpolation, - a sharpening filter used in DCT-based interpolation (mainly around sharper edges), and/ or - a non-interpolating smoothing filter.

[00124] Preferably, only one of the filters can be selected. For quarter-pel interpolation, a sharpening filter can be enabled by default, that is, its parameters must be recovered by analyzing a bit stream on the decoder side without deriving any other flags and parameters. For half-pel interpolation, sharpening can be turned off. For int-pel, both sharpening and smoothing filters are enabled and can therefore be turned off if necessary. However, if either of these is on, its parameters must be retrieved from a bitstream before performing filtering.

[00125] Method 900 comprises a first step 910 of determining whether a reference sample adaptive filtering, RSAF, is in use. For example, this can be determined by evaluating an RSAF flag. If it is determined that RSAF is in use, in step 912 a reference sample adaptive filtering is applied. Subsequently, in step 914, an interpolation using a 4-lead cubic filter is applied. This can be achieved by adjusting an interpolation mode parameter of a filter sequence interpolation filter for intraprediction for the 4-lead cubic filter.

[00126] If it is determined in step 910 that RSAF is not in use, in step 916, an interpolation using a 4-lead filter set is applied. Specifically, this may be a predetermined set of 4-lead filters. Adjusting the interpolation filter to use the 4-lead filter set can be achieved, for example, by setting an interpolation parameter of the interpolation filter to "4-lead filter set".

[00127] In step 920, it is determined whether an MPI index is greater than 1. The MPI index may have been determined, for example, by analyzing or otherwise determining the MPI index value of a bit stream. If the MPI index is greater than 1, the method continues to step 924 evaluating whether a current PU size is greater than 32. If so, and if the MPI index is not greater than 1, the method continues to step 926 applying a boundary prediction filtering. If, however, the current PU size is not greater than 32, the method continues to step 928 with the combination of multi-parameter intraprediction/position-dependent intraprediction.

[00128] Figure 10 is a schematic structure of a filter sequence 1000, wherein the sequence 1000 has a serial/parallel structure.

[00129] The filter sequence 1000 is configured to process a block of samples 1002 to obtain a reference (predicted) block 040. The sequence 1000 comprises one or more subpixel interpolation filters 1010 which are configured to interpolate between the samples of blocks 1002. A result 1012 of the subpixel interpolation is provided as input to a smoothing filter 1020 and/or a sharpening filter 1030. Preferably, either the smoothing filter 1020 or the sharpening filter 1030 is used.

[00130] The output of the smoothing filter 1020 and/or sharpening 1030 is a reference (predicted) block 1040. The filter sequence 1000 is controlled by a filter control module 1050 which is configured to adjust the parameters of the subpixel interpolation filters 1010, smoothing filter 1020 and/or sharpening filter 1030.

[00131] Figure 11 is a flowchart of a method for configuring a sequence of filters for interprediction with a filter control module on the decoder side.

[00132] Method 1100 comprises a first step of determining whether a %-%-PEL interpolation is in use. If this is not the case, the method continues at step 1104 determining whether a ^-PEL interpolation is in use. If so, the method continues at step 1106 analyzing a bit stream to retrieve one or more values of a control flag. Subsequently, in step 1112, it is determined whether the sharpening filter is enabled. Specifically, this can be determined from the one or more control flag values determined in step 1106. Alternatively, if in step 1104, it has been determined that a ^- PEL interpolation is not in use, the method continues at step 1108 with analysis of a bitstream to retrieve one or more values of one or more control flags. Subsequently, in step 1110, it is determined whether a smoothing filter is enabled. Specifically, this may be determined from the one or more values of the one or more control flags.

[00133] If a smoothing filter is enabled, the method continues to step 1116, in which a bit stream is analyzed to recover one or more values of a smoothing filter strength parameter and filtering is performed accordingly. Alternatively, if in step 1110 it is determined, for example from the bitstream, that the smoothing filter is not enabled, the method continues to step 1112, where it is determined whether the sharpening filter is enabled. If the sharpening filter is enabled, the method continues to step 1114, in which the bitstream is analyzed to recover one or more values of the sharpening filter strength parameter, and filtering is performed accordingly.

[00134] A prior art video encoding also provides filtering steps in the final encoder and decoder stage. This filtering is referred to as loop filtering, as the output data from this processing is passed through the motion compensation loop.

[00135] Some filter sequences can be used in both an encoder and a decoder. Preferably, a first stage of the filter sequence is configured to remove blocking artifacts using an unblocking filter. Low-pass filters are applied to edges of TUs according to a set of pre-defined rules. These rules have parameters referred to as unblocking parameters that could be specified for the entire sequence or for each frame separately.

[00136] Preferably, a second stage is configured to remove quantization noise by using adaptive sample shifting. A frame can be subdivided into pixel areas with SAO parameters assigned to each of these areas. SAO parameters can comprise: • SAO type that controls the classifier type: o BO (band offset): this SAO type selects to add offsets to pixels that have values in the range specified by SAO category o EO (band offset). edge): This type of SAO selects to add different offsets to pixels depending on the SAO category • SAO Class: Specifies a pixel pattern that should be used to derive the SAO category • SAO Offsets: A lookup table that assigns each SAO category with an offset. According to the pixel category a corresponding offset must be added.

[00137] Preferably, one or more of these SAO parameters are derived on the encoded side and are encoded into the bit stream so that the decoder can analyze them.

[00138] The next stage is to apply an adaptive loop filter (ALF) which is very close to the Wiener filter. On the encoder side such filter coefficients are derived that provide a minimum mean squared error after filtering reconstructed pixels. These coefficients are additionally quantized and signaled to the decoder in the bitstream.

[00139] Preferably, a filter control module for matching filter intensity at different stages of the loop filtering chain is configured to adjust processing at one filtering stage according to parameter values in the other(s). s) internship(s).

[00140] Figure 12 is a schematic structure of a loop filter chain 1200, which is a sequence of filters.

[00141] The loop filter chain 1200 is configured to process a reconstructed frame 1202 to obtain a temporary reference image storage 1240. The reconstructed frame 1202 is provided as input to the deblocking filter 1210. An output of the deblocking filter is provided as input 1212 to an adaptive sample shift filter 1220. An output from the adaptive sample shift filter 1220 is provided as input 1222 to an adaptive loop filter 1230. An output from the adaptive loop filter 1230 is provided as input 1232 of reference image 1240.

[00142] The deblocking filter 1210, the adaptive sample shift filter 1220, and the adaptive loop filter 1230 are configurable with parameters that can be adjusted by the filter control module 1250. The filter control module 1250 is configured to determine these parameters based on input parameters, which include one or more deblocking filter parameters 1252, one or more SAO parameters 1254, and one or more ALF parameters 1256. For example, these parameters may be user-defined or obtained from a bit stream.

[00143] The deblocking filter 1210 may depend on one or more SAO parameters. Preferably, on the encoder side, if the SAO type is selected to be EO (edge shift) and the pixel pattern (SAO class) is aligned with the horizontal or vertical direction, the unblocking operation is disabled for the encoder side. edges that have a direction orthogonal to the selected pixel pattern direction.

[00144] Figure 13 is a flowchart of a method 1300 for configuring a sequence of filters, wherein the unblocking filtering depends on one or more SAO parameters.

[00145] Method 1300 comprises a first step 1302 of estimating one or more SAO parameters. SAO parameters can include an edge offset parameter and an SAO class.

[00146] In step 1304 it is determined whether the edge displacement parameter is set to true. If so, in step 1306 the SAO class parameter is evaluated. If the SAO class parameter is set to vertical, the method in step 1308 disables the unblocking of horizontal edges. If the SAO class parameter is set to horizontal in step 1310, vertical edge unlocking is disabled. If the SAO class is set to another value, or if in step 1304 the edge shift parameter is determined to be false, the method continues to step 1312 of applying the deblocking filter according to the above determined setting. Subsequently the method continues at step 1314 with adaptive sample displacement filtering and possibly additional filtering steps.

[00147] Figure 14 is a flowchart of an additional method 1400 for configuring a sequence of filters, wherein the deblocking filtering depends on SAO parameters and wherein the filtering occurs on a decoder side.

[00148] In a first step 1402, one or more SAO parameters are retrieved from a bit stream.

[00149] In step 1404 it is determined whether the edge displacement parameter is set to true. If so, in step 1406 the SAO class parameter is evaluated. If the SAO class parameter is set to vertical, the method in step 1408 disables the unblocking of horizontal edges. If the SAO class parameter is set to horizontal, in step 1410, vertical edge unlocking is disabled. If the SAO class is set to another value, or if in step 1404 the edge shift parameter is determined to be false, the method continues to step 1412 of applying the deblocking filter according to the above determined setting. Subsequently the method continues at step 1414 with adaptive sample displacement filtering and possibly additional filtering steps.

[00150] Figure 15 is a flowchart of a method 1500 for configuring a filter sequence, wherein the method 1500 adjusts one or more adaptive loop filter parameters in accordance with one or more SAO parameters on a decoder side.

[00151] Specifically, method 1500 comprises a first step 1502 of estimating the SAO parameters. In a second step 1504, it is determined whether an edge offset flag is set. If this is adjusted, in step 1506, smoothing filters in the adaptive loop filter, ALF, are disabled.

[00152] Subsequently, in steps 1508 to 1514, an unblocking, adaptive sample shift filtering, adaptive loop filter parameter estimation, and adaptive loop filtering are applied.

[00153] Compared to the methods illustrated for example in Figure 13 and Figure 14, the codebook selection for encoding and decoding ALF parameters is different specifically for edge-shift and band-shift SAO cases.

[00154] Figure 16 is a flowchart of a method 1600 for adjusting one or more ALF parameters in accordance with one or more SAO parameters on an encoder side.

[00155] In a first step 1602, SAO parameters are derived from a bit stream. In a second step 1604, it is determined whether an edge offset flag is set. If this is not set, in step 1606 a first ALF parameter codebook is selected. If the edge shift flag is not set, in step 1608, a second ALF parameter codebook is selected.

[00156] Subsequently, in steps 1610 to 1616, the ALF parameters are retrieved from a bit stream, an unblocking filter is applied, an adaptive sample shift filtering is applied, and an adaptive loop filtering is applied.

[00157] Embodiments of the invention may refer to the following additional aspects:

[00158] 1. A method of encoding and decoding video data, comprising an intraprediction process consisting of a. Configure filter parameters from steps given below according to intraprediction parameters b. Adaptively prepare reference samples by applying a filter to neighboring samples of the block being predicted. w. Calculate a predicted value for each sample in the block being predicted using interpolated values from reference samples d. Apply a filter to the predicted block e. Perform border smoothing.

[00159] 2. A method of aspect 1, where quadtree partitioning results are used as an indication of reference sample filter selection using explicit or implicit signaling.

[00160] 3. A method of aspect 1, wherein calculating the predicted value for each sample of the block being predicted is performed using an interpolation filter which is selected in accordance with a reference sample filtering process.

[00161] 4. A method of aspect 3, wherein interpolation filter selection is performed for the blocks being predicted that were calculated without implicit or explicit reference sample filter selection signaling.

[00162] 5. A method of aspect 3, wherein interpolation filter selection is performed for the blocks being predicted that were obtained from the reference samples filtered by a weak reference sample filter.

[00163] 6. A method of aspect 1, where a boundary smoothing is performed for TUs that belong to PUs that meet a size constraint.

[00164] 7. A method of aspect 6, where the predicted block filter directionality affects the decision to apply boundary smoothing.

[00165] 8. A method of encoding and decoding video data, comprising an interprediction process consisting of a. Configure filter parameters from steps given below according to interprediction parameters b. Adaptively prepare reference samples by applying a filter in the search area used to search for such a block that will be used as a reference after processing with interpolation and prediction filters. w. Apply an interpolation filter to the block that is currently processed d. Apply prediction filters to the block that is currently processed.

[00166] 9. A method of aspect 8, where step d precedes step c.

[00167] 10. A method of filtering a signal comprising a. several iterative filtering steps, where the filter strength depends on additional conditions, b. indication associated with some of the filtering steps, which specifies the filtering intensity in the associated step, and c. control unit that ignores additional conditions in filtering steps depending on the indication associated with the preceding steps.

[00168] 11. A method of aspect 10, wherein indicating a strong filter in filtering step i bypasses a check of additional conditions in steps k>i to select a weaker filter.

[00169] Embodiments of the present invention provide one or more of the following advantages: - Many potential applications in hybrid video coding framework that are compatible with JEM which is the foundation for a next generation video coding standard; - Reduced BD rate and subjective quality improvements compared to JEM1. - Reduced computational complexity for both the encoder and decoder compared to JEM1 with integrated RSAF, which makes the invention potentially attractive for many mobile applications; - Avoid redundant signaling (syntax).

[00170] The above descriptions are only ways of implementing the present invention, the scope of the present invention is not limited to this. Any variations or substitutions can be easily made by a person skilled in the art. Therefore, the scope of protection of the present invention must be subject to the scope of protection of the attached embodiments.

Claims (12)

1. Device (100) for configuring a sequence of filters for video coding, characterized in that it comprises: - a sequence of filters (110; 300; 700; 1000; 1200) that is configurable by one or more primary parameters and one or more secondary parameters, wherein the one or more secondary parameters are not stored in a bit stream; and - a filter controller (120) configured to adjust the one or more secondary parameters based on the one or more primary parameters and based on a strength criterion of the filter sequence, wherein the strength criterion relates to a smoothing global filter sequence, and wherein the filter sequence (110; 300; 700; 1000; 1200) comprises: - a reference sample filter (310; 710) configured to adaptively filter one or more neighboring samples of a block of current video to obtain one or more reference samples, and - an interpolation filter (320; 720; 1010) configured to predict one or more samples of the current video block using an interpolation of the one or more reference samples, wherein the one or more primary parameters comprise a reference sample filter reference parameter and the one or more secondary parameters comprise an interpolation filter selection parameter, wherein the interpolation filter is configured to use an interpolation method in accordance with with the selection parameter, wherein the one or more primary parameters comprise a reference sample filter flag of the reference sample filter and the one or more secondary parameters comprise a filter strength parameter of an intraprediction interpolation filter . 2. Device (100), according to claim 1, characterized by the fact that the intensity criterion comprises: - a criterion for a global for global smoothing, - a criterion for a ratio of an amplification factor to a domain of high frequency and an amplification factor for a low frequency domain, and/or - a criterion for a ratio of a contrast value before filtering and a contrast value after filtering. 3. Device (100) according to claim 1 or 2, characterized in that the one or more primary parameters are predetermined parameters, in particular parameters that are predetermined from an encoded bit stream, a user setting and/or or a parameter lookup loop in an encoding device. 4. Device (100), according to any one of claims 1 to 3, characterized by the fact that the filter sequence (110; 300; 700; 1000; 1200) comprises one or more primary filters that are configurable by one or more primary parameters and one or more secondary filters that are configurable by one or more secondary parameters, wherein the one or more primary filters are located in the filter sequence before the one or more secondary filters. 5. Device (100), according to any one of claims 1 to 4, characterized by the fact that the filter sequence (110; 300; 700; 1000; 1200) comprises a boundary smoothing filter (340; 730) which is configured to perform boundary smoothing for one or more transform units that belong to one or more prediction units that meet a size constraint. 6. Device (100), according to any one of claims 1 to 5, characterized by the fact that the one or more primary parameters include a direction parameter of a predicted block filter and the one or more secondary parameters include a parameter of a boundary smoothing filter, wherein the predicted block filter is configured to compute one or more predicted blocks, and the direction parameter indicates a difference in filtering strength between a vertical direction and a horizontal direction. 7. Device (100), according to any one of claims 1 to 6, characterized by the fact that the filter sequence (110; 300; 700; 1000; 1200) comprises: - a subpixel interpolation filter configured to adaptively filtering samples from a reference block to obtain an interpolated reference block, and - a low-pass and/or a high-pass filter configured to smooth and/or sharpen the interpolated reference block to obtain a filtered reference block, wherein the one or more primary parameters comprise an interpolation filter interpolation parameter and the one or more secondary parameters comprise a sharpening and/or smoothing filter selection parameter, wherein the filter controller (120) is configured to derive the parameters secondaries based on the interpolation parameter of the interpolation filter. 8. Device (100), according to any one of claims 1 to 7, characterized by the fact that the sequence of filters (110; 300; 700; 1000; 1200) further comprises: - an unblocking filter (1210) configured to process vertical edges based on a vertical filter intensity parameter and horizontal edges based on a horizontal filter intensity parameter, - a sample adaptive offset (SAO) filter (1220) configured to sort pixels and add values of offset to pixels according to an SAO class parameter, and - an adaptive loop filter (1230) configured to use two or more codebooks to indicate one or more filter coefficients in a bit stream, wherein the one one or more primary parameters comprise the SAO class parameter and an SAO filter SAO type parameter and the one or more secondary parameters comprise the horizontal filter strength parameter and the vertical filter strength parameter in which the control controller filter (120) is configured to derive the secondary parameters based on the SAO class parameter and/or the SAO type parameter of the SAO filter, and/or the filter controller (120) is configured to select two or more codebooks based on the SAO type parameter. 9. Method (200; 400; 500; 600; 800; 900; 1100; 1300; 1400; 1500; 1600) for configuring a filter sequence (110; 300; 700; 1000; 1200) for video encoding, the method characterized by the fact that it comprises steps of: - adjusting one or more secondary parameters based on one or more primary parameters and based on an intensity criterion of the filter sequence, wherein the intensity criterion relates to a global smoothing of the filter sequence, and the one or more secondary parameters are not stored in a bit stream; and - configuring the filter sequence with the primary and secondary parameters, wherein the filter sequence (110; 300; 700; 1000; 1200) comprises: - a reference sample filter (310; 710) configured to adaptively filter a or more neighboring samples of a current video block to obtain one or more reference samples, and - an interpolation filter (320; 720; 1010) configured to predict one or more samples of the current video block using an interpolation of the one or more more reference samples, wherein the one or more primary parameters comprise a reference sample filter reference parameter and the one or more secondary parameters comprise an interpolation filter selection parameter, wherein the interpolation filter is configured to use an interpolation method in accordance with the selection parameter, wherein the one or more primary parameters include a reference sample filter flag of a reference sample filter and the one or more secondary parameters include a reference intensity parameter filter of an intraprediction interpolation filter. 10. Method (200; 400; 500; 600; 800; 900; 1100; 1300; 1400; 1500; 1600), according to claim 9, characterized by the fact that the intensity criterion comprises: - a criterion for a global smoothing, - a criterion for a ratio of an amplification factor for a high-frequency domain and an amplification factor for a low-frequency domain, and/or - a criterion for a ratio of a contrast value before filtering and a contrast value after filtering. 11. Method (200; 400; 500; 600; 800; 900; 1100; 1300; 1400; 1500; 1600), according to claim 9 or 10, characterized by the fact that it further comprises an initial step of determining the one or more primary parameters of a bit stream. 12. A computer-readable storage medium characterized by the fact that it has instructions stored therein, the instructions, when executed on a computer, cause the computer to execute the method as defined in any one of claims 9 to 11.