Encoder, decoder, and method for improved lossless compression

Abstract

A decoder for decoding a picture from a data stream is configured to check (106) whether a plurality of coding parameters (98) contained in the data stream (14) indicates a coding parameter setting corresponding to lossless prediction residual coding, the plurality of coding parameters (98) relating to a predetermined portion (100) of a picture (12) and controlling a prediction residual transform mode and a quantization precision with respect to the predetermined portion. The decoder is configured to, in response to the plurality of coding parameters indicating the coding parameter setting corresponding to lossless prediction residual coding, set one or more predetermined coding options associated with one or more tools of the decoder for processing prediction reconstruction modified by prediction residual with respect to the predetermined portion such that the one or more tools are disabled with respect to the predetermined portion.

Description

Technical Field

[0001] Embodiments according to the invention relate to encoders, decoders, methods and computer programs for improved lossless compression.The present invention relates to extensions for lossless compression in modern hybrid image and video coding solutions.

[0002] Industrial achievements

[0003] In the following, various inventive embodiments and aspects will be described. Furthermore, other embodiments will be defined by the accompanying claims.

[0004] It should be noted that any embodiment defined by the claims may be supplemented by any of the details (features and functions) described below in the different inventive embodiments and aspects.

[0005] In addition, it should be noted that the various aspects described herein can be used alone or in combination. Therefore, details can be added to each of the individual aspects without adding details to another aspect of the aspect.

[0006] It should also be noted that the present disclosure explicitly or implicitly describes features that can be used in both an encoder (a device for providing an encoded representation of an input signal) and a decoder (a device for providing a decoded representation of a signal based on the encoded representation). Thus, any feature described herein can be used in the context of an encoder as well as in the context of a decoder.

[0007] Furthermore, the features and functions disclosed herein in relation to the methods may also be used in a device (configured to perform such functions). Furthermore, any features and functions disclosed herein with respect to the devices may also be used in the corresponding methods. In other words, the methods disclosed herein may be supplemented by any features and functions described with respect to the devices.

[0008] Furthermore, as will be described in the "Alternative Implementations" section, any of the features and functions described herein may be implemented in hardware or software, or using a combination of hardware and software. Background Art

[0009] In existing lossy video compression, the encoder uses a specific quantization step size Δ to quantize the prediction residual or the transformed prediction residual. The smaller the step size, the finer the quantization and the smaller the error between the original signal and the reconstructed signal. Recent video coding standards (such as H.264 and H.265) use an exponential function of the so-called quantization parameter (QP) to derive the quantization step size Δ, for example:

[0010]

[0011] The exponential relationship between the quantization step size and the quantization parameter allows for finer adjustments to the resulting bitrate. The decoder needs to know the quantization step size to correctly scale the quantized signal. This stage is sometimes called "inverse quantization," even though quantization is irreversible. This is why the decoder parses the scaling factor, or QP, from the bitstream. QP signaling is typically performed hierarchically, meaning a base QP is signaled at a higher level in the bitstream (e.g., the picture level). At the sub-picture level (where a picture can consist of multiple slices, tiles, or bricks), only increments to the base QP are signaled. To adjust the bitrate with finer granularity, incremental QPs can even be signaled per block or region of blocks, for example, in HEVC, per transform unit within an NxN region of a coding block. Encoders typically use incremental QP techniques for subjective optimization or rate control algorithms. Without loss of generality, the following assumes that the base unit in this disclosure is a picture; thus, the encoder signals a base QP for each picture consisting of a single slice. In addition to this base QP (also called slice QP), a delta QP may be signaled for each transform block (or any union of transform blocks, also called quantization group).

[0012] Existing video coding schemes such as the High Efficiency Video Coding (HEVC) or the upcoming Versatile Video Coding (VVC) standards optimize energy compression for various residual signal types by allowing additional transforms beyond the widely used integer approximation of the discrete cosine transform type II (DCT-II). The HEVC standard further specifies integer approximations of the discrete sine transform type VII (DST-VII) for 4×4 transform blocks using specific intra-frame directional modes. Due to this fixed mapping, there is no need to signal whether DCT-II or DST-VII is used. In addition, the identity transform can be selected for the 4×4 transform block. Here, the encoder needs to signal whether to apply DCT-II / DST-VII or the identity transform. Since the identity transform is equivalent to multiplying a matrix by 1, it is also called transform skipping. In addition, current VVC developments allow the encoder to select more transforms in the DCT / DST family for the residual as well as additional inseparable transforms, which are applied after the DCT / DST transform at the encoder and before the inverse DCT / DST at the decoder. Both the extended set of DCT / DST transforms and the additional non-separable transforms require additional signaling for each transform block.

[0013] Figure 1bA hybrid video coding method is illustrated with a forward transform and subsequent quantization of the residual signal 24 at the encoder 10 and scaling of the quantized transform coefficients followed by an inverse transform for the decoder 36. The transform and quantization related blocks 28 / 32 and 52 / 54 are highlighted.

[0014] Modern image and video coding solutions, such as High Efficiency Video Coding (HEVC, H.265, ISO / IEC 23008-3) and the currently developing Versatile Video Coding (VVC, H.266), allow for efficient compression of still or moving picture content even at very low bit rates. A typical use case for these codec (encoder-decoder) solutions is the lossy compression of high-resolution video material for broadcast (e.g., television) and streaming (e.g., IP video) applications. Nevertheless, the codecs also support lossless compression, allowing a mathematically perfect reconstruction of the encoded input signal upon decoding. More specifically, HEVC provides several pulse code modulation (PCM) related coding tools, as well as the so-called transform quantization bypass coding mode, which facilitates lossless coding by simplifying the entropy coding process and by disabling the quantization, transform (DCT or DST) and deblocking steps. Detailed information can be found in the publicly available HEVC syntax and decoding specifications [1].

[0015] In the current version of VVC (the successor to HEVC under development), the lossless coding capabilities of HEVC have been largely inherited (at least in the reference codec software [2]). This means that PCM-related coding tools as well as the transformed quantization bypass coding mode are available for activation by the HEVC encoder and the current VVC reference encoder. In addition, the transform quantization bypass coding flag (a 1-bit syntax element) is specified not only for the entire bitstream or pictures (frames) of the bitstream, but also for individual sub-blocks (coding units CU or transform units TU) of the bitstream or frame. In other words, in HEVC and VVC, transformed quantization bypass coding can be enabled on a sub-block basis, allowing each sub-block to disable quantization, transform and deblocking coding tools individually.

[0016] Recently, the above-mentioned contribution to the VVC standardization activity within JVET has been introduced, which corrects a specific lossless coding related flaw in the current VVC draft affected by the transform skip coding functionality (which is also present in HEVC) and stipulates that for a given coded sub-block (CU or TU), the inverse residual coefficient transform operation (inverse DCT or DST) is bypassed. More specifically, the contribution proposes to limit the quantization step size controlled by the quantization parameter (QP) to values ​​greater than or equal to 1 (represented by a QP of 4) in case transform skip is activated in a sub-block. Therefore, with a QP of 4 and disabled loop filtering in the spatial area covered by the sub-block, lossless coding can be achieved when transform skip is activated. However, this behavior is the same as the use of the transform quantization bypass coding mode, since quantization with QP=4 (i.e. unity step size) effectively represents a bypass of the quantization algorithm.

[0017] Furthermore, in order to achieve lossless coding at acceptably low bitrates using VVC or any other video codec with similar lossy coding capabilities and toolset, the behavior of some other newly introduced coding tools (which are not available in HEVC and previous video coding standards) may have to be changed. Specifically,

[0018] It may be necessary to disable the newly introduced adaptive loop filter (ALF) and shaping tools;

[0019] • The matrix of the joint inter-chroma transform [3] needs to be changed to allow lossless coding.

[0020] The object of the present invention is to provide a solution to the two above mentioned drawbacks of redundant lossless coding capabilities (regarding transformed quantization bypass and transform skip functionality) and the necessity to modify the behavioral details of some coding tools when lossless coding is required.

[0021] Therefore, it would be desirable to provide concepts for lossless encoding of pictures or videos at acceptably low bit rates and / or improved lossless compression.

[0022] This is achieved by the subject-matter of the independent claims of the present application.

[0023] Further embodiments according to the invention are defined by the subject matter of the dependent claims of the present application. Summary of the Invention

[0024] According to a first aspect of the present invention, the inventors of the present application recognized that one problem encountered when attempting to improve lossless coding of pictures or videos stems from the fact that some tools in decoders or encoders, such as adaptive loop filters (ALFs) and shaping tools, result in lossy coding. According to the first aspect of the present application, this difficulty is overcome by disabling one or more tools for processing prediction reconstruction modified by prediction residuals to avoid information loss after lossless reconstruction of predetermined portions of a picture. In other words, post-reconstruction modifications to samples of the predetermined portion of the picture are disabled. Furthermore, the first aspect is based on the idea that one or more tools can be disabled for lossless coding without requiring syntax elements to indicate lossless prediction residual coding for the predetermined portion of the picture. The decoder and encoder can be configured to infer the use of lossless prediction residual coding for the predetermined portion of the picture based on multiple coding parameters included in the data stream, thereby eliminating the need for explicit signaling of lossless coding. The decoder is configured to advantageously switch between lossless and lossy coding for each portion of the picture. Additional signaling indicating lossless coding for each portion of the picture is not required. This results in a reduced amount of data to be encoded and a reduced amount of data used to decode a predetermined portion of a picture. Improved lossless coding compression can be achieved. The data stream may include a lossless coding syntax element that indicates whether a predetermined area of ​​a picture or the entire picture is encoded using lossless coding, but the data stream no longer needs to include a lossless coding syntax element for each portion of the picture. This is due to the ability of the decoder and / or encoder to check whether the coding parameters indicate lossless prediction residual coding. It has also been found that higher coding efficiency and reduced bit rate can be achieved by combining disabling one or more tools of the decoder or encoder and checking whether multiple coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding.

[0025] Thus, according to a first aspect of the present application, a decoder for decoding a picture from a data stream is configured to check whether a plurality of coding parameters, such as a quantization parameter (QP) and / or a transform mode (TM), indicate a coding parameter setting corresponding to lossless prediction residual coding, wherein the plurality of coding parameters are included in the data stream, are associated with a predetermined portion of the picture, and control the prediction residual transform mode and quantization precision for the predetermined portion. Such a coding parameter setting corresponding to lossless prediction residual coding is represented, for example, by (QP, TM) = (4, transform skipped) or (QP, TM) = (1…4, transform skipped). The coding parameter setting corresponds to lossless prediction residual coding because such coding is immediately signaled by the coding parameter setting or is caused by the decoder being configured to interpret or change the coding parameter setting as a coding parameter setting that results in lossless residual coding, for example, by mapping a quantization parameter less than four (QP<4) to a quantization parameter equal to four (QP=4) when the transform mode is transform skipped. This interpretation or change of the coding parameter setting may be based on the plurality of coding parameters. The decoder may be configured to derive from the plurality of coding parameters whether to use lossy prediction residual coding or lossless prediction residual coding for a predetermined portion of the picture. In response to the plurality of coding parameters indicating coding parameter settings corresponding to lossless prediction residual coding, the decoder is configured to set one or more predetermined coding options related to one or more tools (e.g., deblocking, sample adaptive offset filtering (SAO), and / or adaptive loop filtering (ALF)) used by the decoder to process prediction reconstruction corrected by the prediction residual for the predetermined portion, such that the one or more tools are disabled for the predetermined portion.

[0026] In parallel with the decoder, an encoder for encoding a picture into a data stream is configured to signal a plurality of coding parameters in the data stream and to check whether the plurality of coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, wherein the plurality of coding parameters are associated with a predetermined portion of the picture and control a prediction residual transform mode and a quantization precision with respect to the predetermined portion. In response to the plurality of coding parameters indicating the coding parameter settings corresponding to lossless prediction residual coding, the encoder is configured to set one or more predetermined coding options associated with one or more tools of the encoder for processing prediction reconstruction corrected by the prediction residual with respect to the predetermined portion, such that the one or more tools are disabled in a prediction loop of the encoder for the predetermined portion.

[0027] According to the second aspect of the present invention, the inventors of the present application realized that one problem encountered when trying to improve the lossless encoding of pictures or videos stems from the fact that lossless encoding generally produces a significantly higher bit rate than lossy encoding. According to the second aspect of the present application, this difficulty is overcome by using different codes in the binarization for lossless encoding and lossy encoding. It is possible to switch between different binarizations for lossless encoding and lossy encoding. It was found that binarization with reduced computational complexity relative to binarization in the lossy case significantly reduces the bit rate of lossless encoding. For example, compared to lossy encoding, in the case of lossless coding, computational complexity is reduced by encoding more binary values ​​using an equal probability bypass mode in CABAC using context modeling and binary arithmetic coding rather than a conventional coding mode. Therefore, processing bottlenecks can be prevented.

[0028] Therefore, a second aspect of the present application relates to a decoder for decoding a picture from a data stream and an encoder for encoding a picture into a data stream. The decoder (encoder) is configured to: for a predetermined portion of the picture, determine whether to (is to) be encoded into the data stream using lossless coding or lossy coding. The decoder (encoder) is configured to: if the predetermined portion of the picture is (to) be encoded into the data stream using lossless coding, then use the prediction residual data to binarize the binary value string in a first manner and perform context-adaptive binary entropy decoding (encoding) on ​​the binary value string, and decode the prediction residual for the predetermined portion from the data stream (encode the prediction residual for the predetermined portion into the data stream); and if the predetermined portion of the picture is (to) be encoded into the data stream using lossy coding, then use the prediction residual data to binarize the binary value string in a second manner and perform context-adaptive binary entropy decoding (encoding) on ​​the binary value string, and decode the prediction residual for the predetermined portion from the data stream (encode the prediction residual for the predetermined portion into the data stream). The first manner is different from the second manner, so that the computational complexity is reduced in the first manner compared to the second manner.

[0029] According to a third aspect of the present invention, the inventors of the present application have recognized that one problem encountered when attempting to improve lossless coding of pictures or videos stems from the fact that the transform of the encoder or the inverse transform of the decoder may require a rounding step to an integer, thereby prohibiting a mathematically perfect reconstruction of the picture or part of the picture. Therefore, currently for lossless coding, the transform process is bypassed or disabled. According to the third aspect of the present application, this difficulty is overcome by using a fully reversible transform for lossless coding. The decoder / encoder is configured to switch between different transform modes for lossless coding and lossy coding. Using a fully reversible transform, lossless coding can be achieved because no rounding errors occur. Fully reversible transforms are, for example, inter-component transforms or transforms within spectral decomposition color components. Additional syntax elements may be required in the data stream to indicate and / or control the fully reversible transform, but it has been found that the bit rate can be significantly reduced, thereby improving coding efficiency.

[0030] Therefore, a third aspect of the present application relates to a decoder for decoding a picture from a data stream and an encoder for encoding a picture into a data stream. The decoder (encoder) is configured to: for a predetermined portion of the picture, determine whether to use lossless coding or lossy coding to encode it into the data stream. The decoder (encoder) is configured to: if the predetermined portion of the picture is (to be) encoded into the data stream using lossless coding, perform a fully reversible transform on the prediction residual or a reconstruction corrected by the prediction residual (for example, in the prediction loop of the encoder) of the predetermined portion; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossy coding, perform a non-fully reversible transform on the prediction residual or a reconstruction corrected by the prediction residual (for example, in the prediction loop of the encoder) of the predetermined portion.

[0031] According to a fourth aspect of the present invention, the inventors of the present application have recognized that one problem encountered when attempting to improve lossless encoding of pictures or videos stems from the fact that smoothing filters result in information loss. According to the fourth aspect of the present application, this difficulty is overcome by disabling filtering for parts of a picture to be losslessly decoded or losslessly encoded, or by reducing the filtering used to obtain the prediction signal compared to lossy encoding. Disabling filtering for this part results in the absence of filtering-related syntax elements in the data stream, thereby achieving improved bitstream compression. Alternatively, the prediction signal for this part of the picture is filtered less frequently in the lossless encoding case than in the lossy encoding case. For example, in the lossless encoding case, an interpolation filter can be applied, where the higher-frequency half of the entire spatial spectrum of the prediction signal has higher energy in the lossless encoding case than in the lossy encoding case. This is based on the idea that in the lossless encoding case, it is advantageous to modify the prediction signal as little as possible when deriving it. It has been found that special filtering options for lossless encoding (resulting in higher information content in the prediction signal) result in a lower bitrate, thereby leading to higher efficiency.

[0032] Therefore, a fourth aspect of the present application relates to a decoder for decoding a picture from a data stream and an encoder for encoding a picture into a data stream. The decoder (encoder) is configured to: for a predetermined portion of the picture, determine whether to use lossless coding or lossy coding to encode it into the data stream. The decoder (encoder) is configured to: if the predetermined portion of the picture is (to be) encoded into the data stream using lossless coding, derive a prediction signal for the predetermined portion in a first manner; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossy coding, derive a prediction signal for the predetermined portion in a second manner. The first manner is different from the second manner, so that the prediction signal is less filtered in the first manner than in the second manner, for example, by using an interpolation filter, or the prediction signal is not filtered in the first manner but is filtered in the second manner.

[0033] According to a fifth aspect of the present invention, the inventors of the present application have realized that one problem encountered when trying to improve lossless coding of pictures or videos stems from the fact that loop filters and / or quantization cause loss of information. According to a fifth aspect of the present application, this difficulty is overcome by disabling one or more tools of the decoder / encoder for processing prediction reconstruction corrected by prediction residuals for a predetermined portion and / or by disabling one or more tools of the decoder / encoder for requantization / quantization of prediction residuals. Disabling the tools for this portion results in the absence of post-processing-related syntax elements and / or (re)quantization-related syntax elements in the data stream, or a reduction in the amount of post-processing-related syntax elements and / or (re)quantization-related syntax elements, thereby achieving improved bitstream compression.

[0034] Therefore, a fifth aspect of the present application relates to a decoder for decoding a picture from a data stream and an encoder for encoding a picture into a data stream. The decoder (encoder) is configured to: for a predetermined portion of the picture, determine whether to use lossless coding or lossy coding to encode it into the data stream. The decoder (encoder) is configured to: if the predetermined portion of the picture is (to be) encoded into the data stream using lossless coding, infer that one or more predetermined coding options related to one or more tools of the decoder (encoder) for processing (for example, in the prediction loop of the encoder) the prediction reconstruction corrected by the prediction residual or for requantization (quantization) of the prediction residual with respect to the predetermined portion are to be set, so that the one or more tools are disabled with respect to the predetermined portion; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossy coding, infer that the one or more predetermined coding options are to be set by deriving the one or more predetermined coding options from a plurality of coding parameters.

[0035] According to a sixth aspect of the present invention, the inventors of the present application have realized that one problem encountered when trying to improve lossless encoding of pictures or videos stems from the fact that the difference between the bit depth (i.e., the input bit depth associated with the video to be encoded or associated with the video to be output by the decoder) and the internal bit depth causes information loss. According to the sixth aspect of the present application, this difficulty is overcome by performing bit depth conversion. Therefore, the input bit depth or the difference between the input bit depth and the internal bit depth is encoded into the data stream by the encoder and read from the data stream by the decoder. Alternatively, the input bit depth or the aforementioned difference can be indicated by one or more encoding parameters (e.g., a quantization parameter (QP)). Using information about the input bit depth or the difference, the decoder is configured to output the video at the input bit depth, whereby the bit depth of the video output is equal to the bit depth of the video input encoded by the encoder. Thus, lossless encoding can be achieved. No additional syntax elements are required when the bit depth is indicated by one or more parameters, thereby achieving improved bitstream compression. Note that the input bit depth (which gets its name from the fact that the input bit depth is the bit depth of the video the encoder receives for encoding, and it is the bit depth the decoder outputs) need not be equal to the internal bit depth. Rather, the encoder and decoder allow different settings in the data stream, which may include one setting if the two bit depths are the same, and one or more settings if the two bit depths are different.

[0036] Therefore, a sixth aspect of the present application relates to a video decoder and a video encoder. The decoder is configured to: perform video decoding from a data stream at an internal bit depth, and receive a video input with an input bit depth or an internal bit depth. In parallel with the decoder, the encoder is configured to: perform video encoding into the data stream at an internal bit depth, and receive a video input at an input bit depth or an internal bit depth. For example, in the case where the bit depth of the input video and / or the output video (i.e., the input bit depth) is equal to the internal bit depth, the output video and / or the input video is associated with the internal bit depth. In addition, the decoder (encoder) is configured to: read a syntax element indicating the difference between the input bit depth and / or the internal bit depth and the input bit depth from the data stream (encode the syntax element indicating the difference between the input bit depth and / or the internal bit depth and the input bit depth into the data stream).

[0037] Therefore, a sixth aspect of the present application relates to a video decoder configured to perform video decoding from a data stream at an internal bit depth and output video at an input bit depth or an internal bit depth. In parallel with the video decoder, a video encoder is configured to perform video encoding into the data stream at the internal bit depth and receive video input having the input bit depth or the internal bit depth. For example, when the bit depth of the input video and / or output video (i.e., the input bit depth) is equal to the internal bit depth, the output video and / or the input video is associated with the internal bit depth. In addition, the decoder (encoder) is configured to read a syntax element indicating a minimum value of a quantization step size scaling parameter (QP minimum, i.e., an absolute or relative quantization step size, requantization step size, or scaling parameter) from the data stream (encode the syntax element indicating the minimum value of the quantization step size scaling parameter (QP minimum, i.e., an absolute or relative quantization step size, requantization step size, or scaling parameter) into the data stream). The minimum quantization step size scaling parameter can represent a unit step size, i.e., a quantization parameter equal to 4. For example, this syntax element indicates the minimum value of the quantization step size scaling parameter (due to a non-zero difference between the internal bit depth and the input bit depth). For example, in addition to the QP minimum value, the decoder can also be informed of the difference between the internal bit depth and the input bit depth by transmitting the difference as described above, or by transmitting both values ​​(i.e., the internal bit depth and the input bit depth). Another QP different from 4 can result in or almost result in lossless coding. If the QP minimum value is set to 4, the decoder can adapt a QP less than 4 to a QP equal to 4, resulting in lossless coding.

[0038] Therefore, a sixth aspect of the present application relates to a video decoder (encoder) configured to: decode an indication of an internal bit depth and an input bit depth or a difference between them from a data stream (encode an indication of the internal bit depth and the input bit depth or the difference between them into the data stream), and perform video decoding from the data stream (video encoding into the data stream) at the internal bit depth. In addition, the decoder is configured to: output video at the input bit depth, and the encoder is configured to: receive video with an input bit depth (i.e., video input). The decoder (encoder) is configured to: check whether the internal bit depth drops below the input bit depth, and change the internal bit depth to correspond to the input bit depth. Therefore, the difference between the input bit depth and the internal bit depth is corrected, and lossless coding is achieved.

[0039] The embodiments relate to methods based on the same considerations as the above-mentioned encoder and / or decoder. Incidentally, these methods can be implemented with all the features and functions also described with respect to the encoder and / or decoder.

[0040] An exemplary embodiment relates to a computer program having a program code for performing the method described herein when the program code runs on a computer.

[0041] Embodiments relate to a data stream obtained by a method for block-based encoding of a picture signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings are not necessarily drawn to scale, emphasis instead generally being placed upon illustrating the principles of the invention. In the following description, various embodiments of the invention are described with reference to the following drawings, in which:

[0043] Figure 1a shows a schematic diagram of an encoder;

[0044] Figure 1b A schematic diagram of an alternative encoder is shown;

[0045] Figure 2 shows a schematic diagram of a decoder;

[0046] Figure 3 A schematic diagram of block-based encoding is shown;

[0047] Figure 4 A schematic diagram showing decoder-side scaling and inverse transform in recent video coding standards;

[0048] Figure 5 A schematic diagram illustrating decoder-side scaling and inverse transformation according to an example is shown;

[0049] Figure 6A schematic diagram illustrating lossless coding verification performed by a decoder and / or encoder according to an embodiment is shown;

[0050] Figure 7 A block diagram showing different binarization methods used depending on lossless or lossy coding;

[0051] Figure 8 A block diagram illustrating a method for lossless encoding using a fully reversible transform;

[0052] Figure 9 A block diagram showing a method for deriving a prediction signal based on lossless or lossy coding;

[0053] Figure 10 A block diagram illustrating a method for setting one or more predetermined encoding options according to lossless or lossy encoding;

[0054] Figure 11a and Figure 11b A block diagram showing a method for considering input bit depth and internal bit depth according to a first embodiment;

[0055] Figure 12a and Figure 12b A block diagram showing a method of considering input bit depth and internal bit depth according to a second embodiment; and

[0056] Figure 13a and Figure 13b A block diagram illustrating a method considering input bit depth and internal bit depth according to a third embodiment is shown. DETAILED DESCRIPTION

[0057] The same or equivalent elements or elements having the same or equivalent functions are denoted by the same or equivalent reference numerals in the following description even if they appear in different drawings.

[0058] In the following description, a number of details are set forth to provide a more comprehensive explanation of the embodiments of the present invention. However, it will be apparent to those skilled in the art that embodiments of the present invention can be practiced without these specific details. In other cases, well-known structures and devices are shown as block diagrams rather than in detail to avoid obscuring the embodiments of the present invention. Furthermore, unless otherwise specifically stated, features of the different embodiments described below may be combined with each other.

[0059] The following description of the drawings begins by presenting a description of an encoder and decoder of a block-based prediction codec for encoding video pictures, forming an example of a coding framework within which embodiments of the present invention may be built. Figures 1a to 3 The embodiments of the inventive concept described herein can be constructed into Figure 1a 、 Figure 1band Figure 2 In the encoder and decoder, although Figures 4 to 6 The described embodiments can also be used to form Figure 1a 、 Figure 1b and Figure 2 The encoder and decoder are based on the coding framework that operates the encoder and decoder.

[0060] Figure 1a An apparatus (eg, a video encoder and / or a picture encoder) is shown for predictively encoding a picture 12 into a data stream 14, exemplarily using transform-based residual coding. The apparatus or encoder is denoted using reference numeral 10. Figure 1b Also shown is an arrangement for predictive coding of a picture 12 into a data stream 14, wherein a possible prediction module 44 is shown in more detail. Figure 2 A corresponding decoder 20 is shown, i.e. an apparatus 20 configured to predictively decode a picture 12′ from the data stream 14, also using transform-based residual decoding, wherein a prime sign is used to indicate that the picture 12′ reconstructed by the decoder 20 deviates from the picture 12 originally encoded by the apparatus 10 with respect to coding losses introduced by quantization of the prediction residual signal. Figure 1a 、 Figure 1b and Figure 2 Transform-based prediction residual coding is used as an example, but the embodiments of the present application are not limited to such prediction residual coding. Figure 1a 、 Figure 1b and Figure 2 The same is true for the other details described.

[0061] The encoder 10 is configured to perform a spatial-to-spectral transform on the prediction residual signal and encode the prediction residual signal thus obtained into the data stream 14. Likewise, the decoder 20 is configured to decode the prediction residual signal from the data stream 14 and perform a spectral-to-spatial transform on the prediction residual signal thus obtained.

[0062] Internally, the encoder 10 may include a prediction residual signal former 22 that generates a prediction residual 24 to measure the deviation of a prediction signal 26 from the original signal (i.e., from the picture 12), wherein, according to an embodiment of the present invention, the prediction signal 26 may be interpreted as a linear combination of a set of one or more predictor blocks. The prediction residual signal former 22 may, for example, be a subtractor that subtracts the prediction signal from the original signal (i.e., from the picture 12). The encoder 10 then further includes a transformer 28 that performs a spatial-to-spectral transform on the prediction residual signal 24 to obtain a spectral-domain prediction residual signal 24', which is then quantized by a quantizer 32, which is also included in the encoder 10. The thus quantized prediction residual signal 24" is encoded into the bitstream 14. To this end, the encoder 10 may optionally include an entropy encoder 34 that entropy encodes the transformed and quantized prediction residual signal into the data stream 14.

[0063] The prediction signal 26 is generated by the prediction stage 36 of the encoder 10 based on the prediction residual signal 24 ″ which is encoded into the data stream 14 and can be decoded from the data stream 14. For this purpose, Figure 1a As shown, the prediction stage 36 may internally comprise a dequantizer 38, which dequantizes the prediction residual signal 24" to obtain a spectral domain prediction residual signal 24'", which corresponds to the signal 24" except for quantization losses; the dequantizer 38 is followed by an inverse transformer 40, which inversely transforms the prediction residual signal 24'", i.e., spectrally to spatially transforms the prediction residual signal 24'", to obtain a prediction residual signal 24", which corresponds to the original prediction residual signal 24 except for quantization losses. A combiner 42 of the prediction stage 36 then recombines the prediction signal 26 and the prediction residual signal 24"", e.g., by addition, to obtain a reconstructed signal 46, i.e., a reconstruction of the original signal 12. The reconstructed signal 46 may correspond to the signal 12'. The prediction module 44 of the prediction stage 36 then generates the prediction signal 26 based on the signal 46 by using, for example, spatial prediction (i.e., intra-picture prediction) and / or temporal prediction (i.e., inter-picture prediction), as shown. Figure 1b Shown in more detail.

[0064] Likewise, if Figure 2 As shown, the decoder 20 may be internally composed of components corresponding to the prediction stage 36 and interconnected in a manner corresponding to the prediction stage 36. In particular, the entropy decoder 50 in the decoder 20 may entropy decode the quantized spectral domain prediction residual signal 24" from the data stream, and then the dequantizer 52, inverse transformer 54, combiner 56 and prediction module 58, which are interconnected and cooperate in the manner of the modules described above with respect to the prediction stage 36, may recover the reconstructed signal based on the prediction residual signal 24", such that Figure 2 As shown, the output of combiner 56 results in a reconstructed signal, ie picture 12'.

[0065] Although not specifically described above, it is readily apparent that the encoder 10 may set some encoding parameters, including, for example, prediction modes, motion parameters, etc., according to some optimization schemes (e.g., in a manner that optimizes certain rate and distortion related criteria, i.e., encoding cost). For example, the encoder 10 and decoder 20, and corresponding modules 44 and 58, may each support different prediction modes, e.g., intra-frame coding mode and inter-frame coding mode. The granularity at which the encoder and decoder switch between these prediction mode types may correspond to the subdivision of the pictures 12 and 12' into coding segments or coding blocks, respectively. For example, a picture may be subdivided into intra-frame coded blocks and inter-frame coded blocks, based on these coding segments.

[0066] An intra-coded block is predicted based on a spatially encoded / decoded neighborhood (e.g., a current template) of a corresponding block (e.g., a current block), as outlined in more detail below. There may be multiple intra-coding modes, and multiple intra-coding modes may be selected for a corresponding intra-coded segment, including directional or angular intra-coding modes, according to which the corresponding segment is filled by extrapolating sample values ​​of the neighborhood to the corresponding intra-coded segment along a certain direction specified for the corresponding directional intra-coding mode. For example, the intra-coding modes may also include one or more other modes, such as a DC coding mode, according to which a DC value is assigned to all samples within the corresponding intra-coded segment for the prediction of the corresponding intra-coded block; and / or a planar intra-coding mode, according to which the prediction of the corresponding block is approximated or determined as a spatial distribution of sample values ​​at sample positions of the corresponding intra-coded block described by a two-dimensional linear function, wherein the tilt and offset of the plane defined by the two-dimensional linear function are driven based on neighboring samples.

[0067] In contrast, inter-coded blocks can be temporally predicted, for example. For an inter-coded block, a motion vector can be signaled within the data stream 14, which indicates the spatial displacement of a portion of a previously coded picture (e.g., a reference picture) of the video to which the picture 12 belongs, at which portion the previously coded / decoded picture was sampled to obtain a prediction signal for the corresponding inter-coded block. This means that, in addition to the residual signal encoding included in the data stream 14, for example, an entropy-coded transform coefficient level representing a quantized spectral domain prediction residual signal 24", the data stream 14 may also have encoded therein coding mode parameters for assigning coding modes to various blocks, prediction parameters for some of these blocks (e.g., motion parameters for inter-coded segments), and optionally other parameters (e.g., parameters for controlling and signaling the subdivision of the pictures 12 and 12', respectively, into segments). The decoder 20 uses these parameters to subdivide the pictures in the same way as the encoder, assign the same prediction modes to the segments, and perform the same prediction to produce the same prediction signal.

[0068] Figure 3 The relationship between two signals is illustrated, one of which is the reconstruction signal, (ie the reconstructed picture 12') and the other is the combination of the prediction residual signal 24"" signaled in the data stream 14 and the prediction signal 26. As mentioned above, the combination can be additive. Figure 3 The prediction signal 26 is shown as a subdivision of the picture region into intra-frame coding blocks schematically shown with hatching and inter-frame coding blocks schematically shown without hatching. The subdivision can be any subdivision, such as a regular subdivision of the picture region into square blocks or non-square blocks with multiple rows and columns, or a multi-tree subdivision of the picture 12 from a tree root block into multiple leaf blocks of varying sizes, such as a quadtree subdivision, etc. Figure 3 A hybrid thereof is shown, where the picture region is first subdivided into multiple rows and columns of tree root blocks, and then the tree root blocks are further subdivided into one or more leaf blocks according to a recursive multi-tree subdivision.

[0069] In addition, the data stream 14 may have an intra-coding mode encoded therein for an intra-coded block 80, which assigns one of several supported intra-coding modes to the corresponding intra-coded block 80. The data stream 14 may also have one or more motion parameters encoded therein for an inter-coded block 82. In general, the inter-coded block 82 is not limited to being temporally encoded. Alternatively, the inter-coded block 82 may be any block predicted from a previously coded portion other than the current picture 12 itself (e.g., a previously coded picture of the video to which the picture 12 belongs, or from a picture of another view or lower layer in the hierarchy if the encoder and decoder are scalable encoders and decoders, respectively).

[0070] Figure 3The prediction residual signal 24"" in is also shown to subdivide the picture area into blocks 84. These blocks may be referred to as transform blocks in order to distinguish them from the coding blocks 80 and 82. In practice, Figure 3 It is shown that the encoder 10 and the decoder 20 can use two different subdivisions of the picture 12 and the picture 12', respectively, into blocks, namely one into coding blocks 80 and 82, respectively, and another into transform blocks 84. The two subdivisions may be identical, i.e. each coding block 80 and 82 may simultaneously form a transform block 84, but Figure 3 The situation is shown as follows: for example, the subdivision into transform block 84 forms an extension of the subdivision into coding blocks 80 and 82, such that any boundary between two blocks 80 and 82 overlaps a boundary between two blocks 84, or in other words, each block 80 and 82 coincides with either one of transform blocks 84 or a group of transform blocks 84. However, the subdivision can also be determined or selected independently of each other, such that transform block 84 can alternatively straddle a block boundary between blocks 80 and 82. With respect to the subdivision into transform block 84, the description given above regarding the subdivision into blocks 80 and 82 is similar, i.e., block 84 can be the result of a regular subdivision of the picture region into blocks (arranged or not arranged into rows and columns), the result of a recursive multi-tree subdivision of the picture region, or a combination thereof or any other type of blocking. Incidentally, it should be noted that blocks 80, 82, and 84 are not limited to square, rectangular, or any other shape.

[0071] Figure 3 It is further illustrated that the combination of the prediction signal 26 and the prediction residual signal 24"" directly produces the reconstructed signal 12'. However, it should be noted that according to alternative embodiments more than one prediction signal 26 may be combined with the prediction residual signal 24"" to produce the picture 12'.

[0072] exist Figure 3 In the embodiment described below, the transform blocks 84 should have the following meanings. The transformer 28 and the inverse transformer 54 perform their transforms in units of these transform blocks 84. For example, many codecs use some kind of DST (discrete sine transform) or DCT (discrete cosine transform) for all transform blocks 84. Some codecs allow skipping of transforms so that for some of the transform blocks 84, the prediction residual signal is encoded directly in the spatial domain. However, according to the embodiment described below, the encoder 10 and the decoder 20 are configured in such a way that they support several transforms. For example, the transforms supported by the encoder 10 and the decoder 20 may include:

[0073] DCT-II (or DCT-III), where DCT stands for discrete cosine transform

[0074] DST-IV, where DST stands for discrete sine transform

[0075] DCT-IV

[0076] DST-VII

[0077] Identity Transformation (IT)

[0078] Naturally, while the transformer 28 will support all forward transformed versions of these transforms, the decoder 20 or inverse transformer 54 will support their corresponding backward or inverse versions:

[0079] Inverse DCT-II (or inverse DCT-III)

[0080] ·Inverse DST-IV

[0081] Inverse DCT-IV

[0082] ·Inverse DST-VII

[0083] Identity Transformation (IT)

[0084] The subsequent description provides more details about the transforms that may be supported by the encoder 10 and decoder 20. In any case, it should be noted that the set of supported transforms may include only one transform, such as a spectral-to-spatial or spatial-to-spectral transform, but it is also possible that the encoder or decoder does not use any transform at all, or does not use any transform for a single block 80, 82, 84.

[0085] As already outlined above, Figures 1a to 3 have been presented as examples in which the inventive concepts described herein may be implemented to form specific examples of encoders and decoders according to the present application. In this regard, Figure 1a 、 Figure 1b and Figure 2 The encoder and decoder of may represent possible implementations of the encoder and decoder described previously herein, respectively. However, Figure 1a 、 Figure 1b and Figure 2 However, an encoder according to an embodiment of the present application may perform block-based encoding of picture 12 using the concepts outlined in more detail above or below, which is different from the Figure 1a or Figure 1b The difference between the encoder and Figure 3 The subdivision into blocks 80 is performed in the illustrated manner and / or no transform is used at all or for individual blocks (e.g., transform skipping / identity transform). Likewise, a decoder according to an embodiment of the present application may perform block-based decoding of a picture 12' from the data stream 14 using the coding concept outlined further below, but the decoder may be different, for example, from Figure 2 The decoder 20 shown is different in that it is different from the Figure 3The described approach subdivides the picture 12 ′ into blocks and / or derives the prediction residual from the data stream 14 , for example, not in the transform domain (but in the spatial domain) and / or does not use any transform at all or for individual blocks.

[0086] According to an embodiment, the inventive concept described below may relate to the transformer 28 / inverse transformer 40 and the quantizer 32 / dequantizer 38 of an encoder, or the inverse transformer 54 and dequantizer 52 of a decoder. According to an embodiment, the transformer 28, inverse transformer 40, 54, quantizer 32 and / or dequantizer 38, 52 may be disabled for lossless coding of blocks of a picture, where lossless coding is indicated by coding parameters (e.g., quantization parameter and transform mode). The inventive concept may also relate to further processing of the prediction residual 24" and / or the prediction signal 26 and / or the prediction residual-corrected prediction reconstruction 46, where the focus is on the functionality of the decoder / encoder during lossless coding.

[0087] The quantization step size (i.e., quantization accuracy) can vary depending on the selected transform and transform block size, as described below. This description is from the decoder's perspective, and the decoder-side scaling 52 (multiplication) by the quantization step size can be viewed as the inverse (irreversible) operation of the encoder-side division by the step size.

[0088] On the decoder side, the scaling 52 (i.e., dequantization) of the (quantized) transform coefficient levels in current video coding standards (e.g., H.265 / HEVC) is designed for transform coefficients generated by higher-precision DCT / DST integer transforms, e.g. Figure 4 As shown in Figure 2. The variable bitDepth specifies the bit depth of the image samples, for example, 8 bits or 10 bits. The variables log2TbW and log2TbH specify the binary logarithms of the transform block width and height, respectively. Figure 4 Decoder-side scaling 52 and inverse transform 54 in recent video coding standards such as H.265 / HEVC are shown.

[0089] It should be noted that at the decoder, the two 1D DCT / DST based integer transforms 1281 introduce an additional factor This needs to be compensated by scaling with its inverse. For non-square blocks with odd log2TbH+log2TbW, the scaling consists of This can be taken into account by increasing the scaling factor 181 / 256 or using a different set of levelScale values ​​that contain this factor for this case, for example, levelScale[] = {29, 32, 36, 40, 45, 51}. This does not apply for the identity transform or transform skip case 1282.

[0090] It can be seen that for QPs less than 4, the step size or scaling factor becomes less than 1, because the levelScale of these QPs is less than 64=2 6 For the transform coefficients, this is not a problem because the integer forward transform 1281 increases the precision of the residual signal and therefore increases the dynamic range. However, for the residual signal in the case of identity transform or transform skip 1282, there is no increase in dynamic range. In this case, a scaling factor less than 1 may introduce distortion for QP < 4, which is not present for QP 4 with a scaling factor of 1. This contradicts the quantizer design intention that reducing the QP should reduce distortion.

[0091] Changing the quantization step size depending on the selected transform (e.g., whether the transform is skipped) can be used to derive different quantization step sizes for transform skipping 1282. In particular for the lowest QPs (0, 1, 2, and 3), this will solve the problem of quantization step sizes / scaling factors being less than 1 for the lowest QPs. Figure 5 In one example shown, a solution could be to clip 53 the quantization parameter to the minimum allowed value of 4 (QP'), resulting in a quantization step size that cannot be less than 1. Furthermore, the size-dependent normalization 541 with bdShift1 and the final rounding 542 to the bit depth with bdShift2 required for the transform could be moved to the transform path 54. This would simplify transform skip scaling to a 10-bit downshift with rounding. In another example, instead of clipping the QP value to 4, a bitstream restriction could be defined that does not allow the encoder to use QP values ​​that result in a scaling factor less than 1 for transform skipping. Figure 5 Improved decoder-side scaling 52 and inverse transform 54 according to the present invention are shown.

[0092] At the other end of the bitrate range, i.e., for lower bitrates, the quantization step size for the identity transform 1282 can be reduced by an offset, resulting in higher fidelity for blocks with no transform applied or with the identity transform 1282 applied. This enables the encoder to select an appropriate QP value for transform-skipped blocks to achieve higher compression efficiency. This aspect is not limited to the identity transform / transform-skipped 1282; it can also be used to modify the QP for other transform types 1281 by an offset. The encoder will determine this offset, for example, in a manner that improves coding efficiency, such as by maximizing perceived visual quality or minimizing objective distortion (e.g., squared error for a given bitrate), or by reducing the bitrate for a given quality / distortion. The optimal derivation of this slice QP (in terms of the applied criteria) depends, for example, on the content, bitrate, or complexity operating point, as well as other factors such as the selected transform and transform block size. This disclosure describes methods for signaling QP offsets for the case of multiple transforms. Without loss of generality, given two alternative transforms, the encoder can transmit a fixed QP offset for each of the two alternative transforms in a high-level syntax structure (e.g., sequence parameter set, picture parameter set, tile group header, slice header, etc.). Alternatively, when the encoder has selected an alternative transform, the QP offset is transmitted by the encoder, for example, for each transform block. A combination of these two approaches is to signal the base QP offset in the high-level syntax structure and signal an additional offset for each transform block using the alternative transform. The offset can be a value that is added to or subtracted from the base QP, or an index into a set of offset values. The set can be predefined or signaled in the high-level syntax structure.

[0093] • The QP offset relative to the base QP for the identity transform may be signaled in a high-level syntax structure (eg, at sequence level, picture level, tile group level, tile level, or slice level).

[0094] • Alternatively, a QP offset relative to the base QP for the identity transform can be signaled for each coding unit or a predefined set of coding units.

[0095] • Alternatively, for each transform unit to which the identity transform is applied, a QP offset relative to the base QP for the identity transform is signaled.

[0096] Figure 6 A decoder (eg, Figures 1a to 2 Furthermore, the following description is primarily from the perspective of the decoder, although it is apparent that the encoder may include parallel features.

[0097] The picture 12 is encoded by the encoder into a data stream 14, and the decoder is configured to provide a reconstructed picture based on the data stream 14, wherein in the case of lossless coding, the reconstructed picture is equal to the picture 12 or has no identifiable minimum difference from the picture 12. The picture 12 can be divided into parts 100 (i.e., blocks) and regions 104. The region 104 includes a plurality of parts 100. The predetermined part 100 is within the predetermined region 104. The data stream 14 can include part individual information (such as a plurality of encoding parameters 98) and region individual information (such as an optional lossless encoding syntax element 102). The plurality of encoding parameters 98 are associated with the predetermined part 100 of the picture 12, and for example, in the case of Figures 1a to 2 The transformer 28 and inverse transformers 40, 54 shown control the prediction residual transform mode with respect to the predetermined portion 100 and (for example, in Figures 1a to 2 The quantizer 28 and dequantizers 40, 54 shown internally control the quantization precision with respect to the predetermined portion 100. The prediction residual transform mode may be controlled by the transform mode indication syntax element 981. The quantization precision may be controlled by the quantization parameter (QP) 982.

[0098] A decoder for decoding a picture 12 from a data stream 14 is configured to check 106 whether a plurality of coding parameters 98 indicates a coding parameter setting corresponding to lossless prediction residual coding 1061, wherein the plurality of coding parameters 98 corresponds to a coding parameter setting for lossless prediction residual coding 1061, either due to immediate signaling of such coding or due to a decoder configured to interpret or change such a setting as a setting that results in lossless residual coding (e.g., mapping QP < 4 to QP = 4 in the case of TM = transform skipped). An encoder for encoding the picture 12 into the data stream 14 may also be configured to check whether the plurality of coding parameters 98 indicates a coding parameter setting corresponding to lossless prediction residual coding 1061. (QP, TM) = (4, transform skipped) or (QP, TM) = (1 ... 4, transform skipped) may indicate a coding parameter setting corresponding to lossless prediction residual coding 1061. The encoding parameter setting of the quantization parameter 982 that achieves a quantization precision equal to or finer than a predetermined quantization precision and / or the encoding parameter setting of the transform mode indication syntax element 981 that uses transform skipping as the transform mode can correspond to lossless prediction residual encoding 1061.

[0099] According to an embodiment, the decoder / encoder is configured to read / signal a plurality of coding parameters 98 from / to the data stream 14, and check whether the prediction residual transform mode indicated by the plurality of coding parameters 98 (e.g., indicated by the transform mode indication syntax element 981) corresponds to a transform skip mode, and whether the quantization precision indicated by the plurality of coding parameters 98 (e.g., indicated by the quantization parameter 982) corresponds to a quantization step size finer than a predetermined quantization step size, for example, whether it corresponds to QP < 4 (corresponding to no quantization), and if so, change the quantization step size to a predetermined quantization step size, for example, QP = 4. If the transform mode is set to transform skip, the decoder / encoder is configured to map the quantization parameter 982 finer than the predetermined quantization precision (QP < 4) to a quantization parameter 982 equal to the predetermined quantization precision to achieve lossless encoding. The predetermined quantization precision may represent no quantization 32, 52 or bypassed or disabled quantization 32, 52.

[0100] In response to the plurality of encoding parameters 98 indicating an encoding parameter setting corresponding to lossless prediction residual encoding 1061, the decoder / encoder is configured to: set 110 one or more predetermined encoding options associated with one or more tools of the decoder / encoder for processing (e.g. in a prediction loop of the encoder) a prediction reconstruction of the prediction residual correction with respect to the predetermined portion 100, such that the one or more tools are disabled with respect to the predetermined portion 100. The prediction reconstruction of the prediction residual correction may represent Figures 1a to 2 The output of the combiner 56 or 42 is shown, ie the reconstructed signal 46. The one or more tools relate to, for example, deblocking, SAO, ALF, and may be located downstream of the output 46 of the combiner 56 or 42, respectively.

[0101] According to one embodiment, the decoder / encoder is configured to: if the multiple encoding parameters indicate an encoding parameter setting corresponding to lossless prediction residual encoding 1061, set one or more predetermined encoding options with respect to the predetermined part 100 so that the one or more tools are disabled 110 with respect to the predetermined part 100; and if the multiple encoding parameters do not indicate an encoding parameter setting corresponding to lossless prediction residual encoding (i.e., in the lossy encoding case 1062), set one or more predetermined encoding options to a predetermined tool state.

[0102] The decoder is optionally configured to read from the data stream 14 a lossless coding syntax element 102 (e.g. determined and encoded by the encoder) indicating whether a predetermined region 104 of a picture covering or containing the predetermined portion 100 (not exclusively, but for the portion where the check 106 is performed) is coded into the data stream 14 using lossless coding or using lossy coding. The decoder is configured to, if the lossless coding syntax element 102 indicates that the predetermined region 104 of the picture 12 is coded into the data stream 14 using lossless coding 1061 and if the plurality of coding parameters 98 indicates a coding parameter setting corresponding to lossless prediction residual coding 1061, set 110 one or more predetermined coding options such that the one or more tools are disabled with respect to the predetermined portion 100. In addition, the decoder is configured to set 120 one or more predetermined coding options to a predetermined tool state if the plurality of coding parameters 98 do not indicate (i.e., are not equal to) a coding parameter setting corresponding to lossless prediction residual coding 1061 or if the lossless coding syntax element 102 indicates that a predetermined region 104 of the picture 12 is encoded into the data stream 14 using lossy coding 1062. The lossless coding syntax element 102 is signaled for the region 104 of the picture 12 and for each portion 100 within the region 104 (wherein the decoder / encoder checks 106 for each portion 100 whether the plurality of coding parameters 98 indicates lossless coding 1061 or lossy coding 1062 for the individual portion 100). As a result, some of the portions may be decoded / encoded differently than indicated by the lossless coding syntax element 102 for the entire region 104.

[0103] According to an embodiment, the decoder is configured to determine a predetermined tool state based on one or more syntax elements 108 (e.g., syntax elements related to SAO, ALF, etc.) in the data stream 14. In the case of lossless coding 1061, the decoder is configured to skip 122 reading the one or more tool syntax elements 108 because the one or more tools are disabled 110. Optionally, at least one of the one or more syntax elements 108 is not present in the data stream 14 if one or more predetermined coding options are set such that the one or more tools are disabled 110 for the predetermined portion 100, cf. aspect 3.4.

[0104] According to an embodiment, the decoder is configured to: in response to the plurality of encoding parameters 98 indicating an encoding parameter setting corresponding to the lossless prediction residual encoding 1061, set 130 one or more other encoding options (for example, those described in the following description with respect to aspects 1 to 3.4 below) with respect to the predetermined portion 100 to a default state. The default state may mean: in the process of deriving the prediction signal for the predetermined portion 100 (for example, Figures 1a to 2reduction of filtering or disabling of filtering in terms of low-pass filtering of the prediction signal 26 shown in , see aspect 3.3; and / or to filter the prediction residual signal (eg, Figures 1a to 2 3.2). Other coding options may involve: binarization of the prediction residual data into a string of binary values ​​(see aspect 3.1) and context-adaptive binary entropy decoding of said string of binary values, and / or the use of a fully reversible transform of the prediction residual 24'", 24"" or the prediction residual modified reconstruction 46 of the predetermined portion 100 (see aspect 3.2), and / or disabling or reducing filtering for deriving the prediction signal 26 for the predetermined portion (see aspect 3.3), and / or disabling requantization of the prediction residual or processing of the prediction residual modified prediction reconstruction 46 with respect to the predetermined portion 100 (see aspect 3.4). See aspect 3.4). Similarly, for the encoder, other encoding options may involve: binarization of the prediction residual data into a binary value string and context-adaptive binary entropy coding of the binary value string, and / or the use of a fully reversible transform of the prediction residual 24'", 24"" or the prediction residual modified reconstruction 46 of the predetermined portion in the prediction loop of the encoder, and / or the disabling or reduction of filtering for deriving the prediction signal 26 for the predetermined portion, and / or the disabling of prediction residual requantization or processing of the prediction residual modified prediction reconstruction 46 with respect to the predetermined portion 100.

[0105] The encoder can be configured to determine and encode a plurality of coding parameters 98, lossless coding syntax elements 102, and / or tool syntax elements 108 into the data stream 14. The encoder can include parallel features and / or functionality described with respect to the decoder. This applies at least to the prediction loop 36 of the encoder, which has the same features and / or functionality as the decoder. However, it will be apparent that features related to the inverse transformer 54 or dequantizer 52 of the decoder can also be similarly applied to the transformer 28 and / or quantizer 32 of the encoder.

[0106] One or more of the following aspects may be integrated in the Figure 6 Alternatively, the following aspects may be implemented separately in a decoder or encoder.

[0107] Aspect 1: Reducing the number of syntax elements for lossless image or video coding

[0108] The basic approach in this regard is to make the per-subblock transform quantization bypass coding flag obsolete, because with the above modifications, when using a coding quantization parameter (QP) of 4 (unit step size) and disabling post-reconstruction modifications of block samples by coding tools (such as loop filters), its functionality is completely provided by the per-subblock transform skip function. The deactivation of such post-filter can be achieved by adjusting the transform skip mode and the QP lower than or equal to 4.

[0109] In the following paragraphs, different possible implementations are specified.

[0110] One embodiment of aspect 1 is to use a 1-bit transform skip indicator per coding block (here, per transform block and per color component), called transform_skip_flag. This indicator is transmitted as part of the transform unit coding syntax, for example, as shown in Table 1. The plurality of coding parameters 98 may include transform_skip_flag, i.e., a transform mode indication syntax element 981, which controls the prediction residual transform mode for a predetermined portion 100 of a picture 12. The transform_skip_flag defines transform skipping as a transform mode, i.e., a prediction residual transform mode, for the predetermined portion 100.

[0111]

[0112]

[0113]

[0114] Table 1 Transformation unit coding syntax

[0115] Based on this transform_skip_flag (which has a value of 1 if transform skipping is used and 0 otherwise), the actual quantization parameter qP 982 can be determined for palette coding as described in Item 1.1, for example, as specified in equations (436)-(438); and the actual quantization parameter qP 982 can also be determined for transform coefficient scaling as described in Item 1.2, for example, as specified in equation (1153). The plurality of coding parameters 98 may include the quantization parameter qP 982, which controls the quantization accuracy for a predetermined portion 100 of the picture 12.

[0116] 1.1 Decoding process for palette mode

[0117] The inputs to this process are:

[0118] - position (xCbComp, yCbComp) which specifies the top left sample of the current coding block relative to the top left sample of the current picture,

[0119] - a variable treeType which specifies whether a single tree or a dual tree is used, and if a dual tree is used, whether the current tree corresponds to the luma component or the chroma component,

[0120] - a variable cIdx, which specifies the color component of the current block,

[0121] - Two variables nCbW and nCbH, which specify the width and height of the current coding block respectively.

[0122] The output of this process is an array recSamples[x][y], which specifies the reconstructed sample values ​​for the block, where x = 0..nCbW-1, y = 0..nCbH-1.

[0123] Depending on the value of treeType, the variables startComp, numComps, and maxNumPalettePredictorSize are derived as follows:

[0124] - If treeType is equal to SINGLE_TREE:

[0125] startComp=0 (422)

[0126] numComps=sps_chroma_format_idc==0?1:3 (423)

[0127] maxNumPalettePredictorSize=63 (424)

[0128] Otherwise, treeType is equal to DUAL_TREE_LUMA:

[0129] startComp=0 (425)

[0130] numComps=1 (426)

[0131] maxNumPalettePredictorSize=31 (427)

[0132] Otherwise, treeType is equal to DUAL_TREE_CHROMA:

[0133] startComp=1 (428)

[0134] numComps=2 (429)

[0135] maxNumPalettePredictorSize=31 (430)

[0136] Based on the value of cIdx, the variables nSubWidth and nSubHeight are derived as follows:

[0137] -If cIdx is greater than 0 and startComp is equal to 0, nSubWidth is set equal to SubWidthC and nSubHeight is set equal to SubHeightC.

[0138] Otherwise, nSubWidth is set equal to 1 and nSubHeight is set equal to 1.

[0139] The (nCbW x nCbH) block at position (xCbComp, yCbComp) of the reconstructed sample array recSamples is denoted by recSamples[x][y], where x = 0..nCbW-1 and y = 0..nCbH-1, and for each x in the range of 0 to nCbW-1 (inclusive) and each y in the range of 0 to nCbH-1 (inclusive), the value of recSamples[x][y] is derived as follows:

[0140] - The variables xL, yL, xCbL and yCbL are derived as follows:

[0141] xL=x*nSubWidth (431)

[0142] yL=y*nSubHeight (432)

[0143] xCbL=xCbComp*nSubWidth (433)

[0144] yCbL=yCbComp*nSubHeight (434)

[0145] -The variable bIsEscapeSample is exported as follows:

[0146] - If PaletteIndexMap[xCbL+xL][yCbL+yL] is equal to MaxPaletteIndex and Palette_escape_val_present_flag is equal to 1, then bIsEscapeSample is set equal to 1.

[0147] Otherwise, bIsEscapeSample is set equal to 0.

[0148] If bIsEscapeSample is equal to 0, the following applies:

[0149] recSamples[x][y]=CurrentPaletteEntries[cIdx]

[0150] [PaletteIndexMap[xCbL+xL][yCbL+yL]] (435)

[0152] Otherwise (bIsEscapeSample is equal to 1), the following sequence of steps applies:

[0153] 1. The quantization parameter qP is derived as follows:

[0154] - If cIdx is equal to 0,

[0155] qP=Max(QpPrimeTsMin, Qp′Y) (436)

[0156] - Otherwise, if cIdx is equal to 1,

[0157] qP=Max(QpPrimeTsMin, Qp′Cb) (437)

[0158] - Otherwise (cIdx is equal to 2),

[0159] qP=Max(QpPrimeTsMin, Qp′Cr) (438)

[0160] 1.2 Scaling process of transform coefficients

[0161] The inputs to this process are:

[0162] - luma position (xTbY, yTbY) which specifies the top left sample of the current luma transform block relative to the top left luma sample of the current picture,

[0163] - the variable nTbW, which specifies the transform block width,

[0164] - the variable nTbH, which specifies the transform block height,

[0165] - a variable predMode which specifies the prediction mode of the coding unit,

[0166] - A variable cIdx which specifies the color component of the current block.

[0167] The output of this process is an (nTbW)x(nTbH) array d of scaled transform coefficients, whose elements are d[x][y].

[0168] The quantization parameter qP and the variable QpActOffset are derived as follows:

[0169] - If cIdx is equal to 0, the following applies:

[0170] qP=Qp' Y (1142)

[0171] QpActOffset=cu_act_enabled_flag[xTbY][yTbY]? -5:0 (1143)

[0172] Otherwise, if TuCResMode[xTbY][yTbY] is equal to 2, the following applies:

[0173] qP=Qp' CbCr (1144)

[0174] QpActOffset=cu_act_enabled_flag[xTbY][yTbY]? 1:0 (1145)

[0175] - Otherwise, if cIdx is equal to 1, the following applies:

[0176] qP=Qp' Cb (1146)

[0177] QpActOffset=cu_act_enabled_flag[xTbY][yTbY]? 1:0 (1147)

[0178] Otherwise (cIdx equals 2), the following applies:

[0179] qP=Qp' Cr (1148)

[0180] QpActOffset=cu_act_enabled_flag[xTbY][yTbY]? 3:0 (1149)

[0181] Modify the quantization parameter qP and export the variables rectNonTsFlag and bdShift as follows:

[0182] - If transform_skip_flag[xTbY][yTbY][cIdx] is equal to 0, the following applies:

[0183] qP=Clip3(0, 63+QpBdOffset, qP+QpActOffset) (1150)

[0184] rectNonTsFlag=(((Log2(nTbW)+Log2(nTbH))&1)

[0185] ==1)? 1:0 (1151)

[0186] bdShift=BitDepth+rectNonTsFlag+

[0187] ((Log2(nTbW)+Log2(nTbH)) / 2)-5+

[0188] sh_dep_quant_used_flag (1152)

[0189] Otherwise (transform_skip_flag[xTbY][yTbY][cIdx] is equal to 1), the following applies:

[0190] qP=Clip3(QpPrimeTsMin, 63+QpBdOffset, qP+QpActOffset) (1153)

[0192] rectNonTsFlag=0 (1154)

[0193] bdShift=10 (1155)

[0194] According to one embodiment, the value of qP is constrained to be greater than or equal to a minimum value QpPrimeTsMin, which is a constant (across a given video sequence) specified in equation (69):

[0195] sps_internal_bitdepth_minus_input_bitdepth specifies the minimum allowed quantization parameter for transform skip mode as follows:

[0196] QpPrimeTsMin=4+6*sps_internal_bitdepth_minus_input_bitdepth (69)

[0198] The value of sps_internal_bitdepth_minus_input_bitdepth should be in the range of 0 to 8, inclusive.

[0199] In other words, the decoder / encoder is configured to read / signal a plurality of coding parameters 98 from / in the data stream 14, and to check whether the prediction residual transform mode indicated by the plurality of coding parameters corresponds to a transform skip mode and to a quantization step size finer than a predetermined quantization step size (i.e., QpPrimeTsMin), which corresponds to no quantization; and if so, to change the quantization step size to the predetermined quantization step size. Finally, QpPrimeTsMin is controlled by the parameter sps_internal_bitdepth_minus_input_bitdepth, which is transmitted sequence by sequence (in the sequence header). The predetermined quantization step size may depend on the internal bit depth and the input bit depth, for example, it may depend on the difference between the internal bit depth and the input bit depth. The decoder / encoder may be configured to derive a minimum value of the quantization step size scaling parameter, i.e., the predetermined quantization step size and / or QpPrimeTsMin, based on this difference, and to adhere to the minimum quantization step size scaling parameter for the portion 100 in the prediction residual transform skip mode. The decoder / encoder may be configured to, when observing a minimum quantization step size scaling parameter for a portion 100 encoded in prediction residual transform skip mode, change the signaled quantization step size scaling parameter 982 signaled in the data stream 14 for the portion 100 to be equal to the minimum quantization step size scaling parameter if the signaled quantization step size scaling parameter 982 falls below the minimum quantization step size scaling parameter. Note that when this parameter is equal to zero (which is a common case), QpPrimeTsMin yields a value of 4.

[0200] Therefore, by using the element sps_internal_bitdepth_minus_input_bitdepth, the element transform_ski_pflag and the variable QpPrimeTsMin, lossless coding of prediction residual corrected prediction reconstructed picture samples in a specific picture region (eg, coding unit or transform unit) and a specific color component can be achieved.

[0201] In order to achieve lossless coding even in the presence of post-reconstruction deblocking filtering, the deblocking filter parameters are also conditioned on the actual quantization parameter qP, so that the filter is effectively bypassed when the value of qP is less than or equal to QpPrimeTsMin. This bypassing is achieved by specifying the deblocking parameters β' and tc' to be equal to 0, as shown in Table 2 and equations (1280) and (1282) in item 1.3. Note that in the case where QpPrimeTsMin of equation (69) is greater than 4 (caused by sps_internal_bitdepth_minus_input_bitdepth>0), it may be necessary to transmit specific suitable values ​​of sh_luma_beta_offset_div2 for equation (1280) and sh_luma_tc_offset_div2 for equation (1282). The decoder / encoder is configured to: if a predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossless coding 1061, infer that one or more predetermined coding options related to one or more tools of the decoder / encoder for processing prediction reconstruction corrected by prediction residuals or for requantization / quantization of prediction residuals with respect to the predetermined portion 100 are to be set 110, such that the one or more tools are disabled with respect to the predetermined portion 100; and if the predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossy coding 1062, infer that the one or more predetermined coding options are to be set by deriving 120 the one or more predetermined coding options from a plurality of coding parameters 98.

[0202] 1.3 Decision process of luminance block edge

[0203] The inputs to this process are:

[0204] -Picture sample array recPicture,

[0205] - position (xCb, yCb) which specifies the top left sample of the current coding block relative to the top left sample of the current picture,

[0206] - position (xBl, yBl) which specifies the top left sample of the current block relative to the top left sample of the current coding block,

[0207] - the variable edgeType, which specifies whether to filter vertical (EDGE_VER) or horizontal (EDGE_HOR) edges,

[0208] - variable bS, which specifies the boundary filter strength,

[0209] - the variable maxFilterLengthP, which specifies the maximum filter length,

[0210] - The variable maxFilterLengthQ, which specifies the maximum filter length.

[0211] The output of this process is:

[0212] - variables dE, dEp and dEq containing the decision,

[0213] - modified filter length variables maxFilterLengthP and maxFilterLengthQ,

[0214] -Variable t C .

[0215] Sample value P i，k and q j，k Derived as follows, where i = 0..Max(2, maxFilterLengthP), j = 0..Max(2, maxFilterLengthQ), and k = 0 and 3:

[0216] - If edgeType is equal to EDGE_VER, then the following applies:

[0217] q j，k =recPicture[xCb+xBl+j][yCb+yBl+k] (1273)

[0218] P i，k =recPicture[xCb+xBl-i-1][yCb+yBl+k] (1274)

[0219] Otherwise (edgeType equals EDGE_HOR), the following applies:

[0220] q j，k =recPicture[xCb+xBl+k][yCb+yBl+j] (1275)

[0221] P i，k =recPicture[xCb+xBl+k][yCb+yBl-i-1] (1276)

[0222] The variable qpOffset is derived as follows:

[0223] - If sps_ladf_enabled_flag is equal to 1, the following applies:

[0224] -The variable lumaLevel of the reconstructed brightness level is exported as follows:

[0225] lumaLevel=((P 0，0 +P0，3 +q 0，0 +q 0，3 )>>2), (1277)

[0226] -The variable qpOffset is set equal to sps_ladf_lowest_interval_qp_offset and modified as follows:

[0227] for(i=0;i<sps_num_ladf_intervals_minus2+1;i++){

[0228] if(lumaLevel>SpsLadfIntervalLowerBound[i+1])

[0229] qpOffset=sps_ladf_qp_offset[i] (1278)

[0230] else

[0231] break

[0232] }

[0233] Otherwise, qpOffset is set equal to 0.

[0234] variable Qp Q and Qp P is set equal to the Qp of the coding unit Y Value, these coding units include sample q 0，0 and p 0，0 The coding block.

[0235] The variable qP is derived as follows:

[0236] qP=((Qp Q +Qp P +1)>>1)+qpOffset (1279)

[0237] The value of the variable β' is determined according to the quantization parameter Q derived as follows, as specified in Table 2:

[0238] Q=Clip3(0,63,qP+(sh_luma_beta_offset_div2<<1)) (1280)

[0239] Where sh_luma_beta_offset_div2 is for samples containing q 0,0 The value of the syntax element sh_luma_beta_offset_div2 of the slice.

[0240] The variable β is derived as follows:

[0241] β=β'*(1<<(BitDepth-8)) (1281)

[0242] variable t C The value of ' is determined based on the quantization parameter Q derived as follows, as specified in Table 2:

[0243] Q=Clip3(0, 65, qP+2*(bS-1)+(sh_luma_tc_offset_div2<<1)) (1282)

[0245] Where sh_luma_tc_offset_div2 is for samples containing q 0，0 The value of the syntax element sh_luma_tc_offset_div2 of the slice.

[0246] variable t C Export as follows:

[0247] roundOffset=1<<(9-BitDepth) (1283)

[0248] tC=BitDepth<10?(t C ′+roundOffset)>>(10-BitDepth):t C '*

[0249] (1<<(BitDepth-10)) (1284)

[0250] Q 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 β′ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 <![CDATA[t c ′]]> 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Q 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 β′ 7 8 9 10 11 12 13 14 15 16 17 18 20 22 24 26 28 <![CDATA[t C ′]]> 0 3 4 4 4 4 5 5 5 5 7 7 8 9 10 10 11 Q 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 β′ 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 <![CDATA[t C ′]]> 13 14 15 17 19 21 24 25 29 33 36 41 45 51 57 64 71 Q 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 β′ 64 66 68 70 72 74 76 78 80 82 84 86 88 - - <![CDATA[t C ′]]> 80 89 100 112 125 141 157 177 198 222 250 280 314 352 395

[0251] Table 2: Threshold variables β' and t derived from input Q C '

[0252] Aspect 2: Introducing a new “global” lossless_coding flag

[0253] In order to streamline the VVC configuration for lossless coding capabilities, it is proposed to introduce a new one-bit flag in the sequence, frame, picture, sub-picture, tile group or slice header syntax of the image or video coding bitstream, here called lossless_coding. In other words, it is proposed to introduce a new "global" (relative to the sub-block level) flag, whose value can be 0 (non-lossless coding operation, i.e., normal mode) or 1 (lossless mode), and which controls the operation (activation or deactivation or algorithm details) of at least two coding tools provided (i.e., specified) by the affected image or video codec. More specifically, the operation of at least 2 tools in the following tool list depends on the lossless_coding flag:

[0254] Entropy encoder for any residual (spatial or transformed) signal coefficients

[0255] Joint Inter-Component Transform (ICT) in terms of transformation matrix operations

[0256] Smoothing filtering of some predictive coding tools (will be disabled)

[0257] Loop filter: Deblocking, Shape Adaptive Offset, Adaptive Loop Filter (ALF), Shaper (disabled)

[0258] Quantization of residual space or transform coefficients (disabled)

[0259] Aspect 3: The expected effect of the "global" lossless_coding flag on individual encoding tools

[0260] This third aspect specifies in detail how the value of the lossless_coding flag or the inferred meaning of the transform mode (indicating transform skip functionality) and QP (indicating a specific step size) as described in aspect 1 above changes the entropy coding, inter-component transform, prediction and loop filtering tools for a given picture area or portion associated with the lossless_coding flag or the inferred meaning.

[0261] 3.0 Quantification

[0262] The quantization parameter (QP) is equal to 4, and the use of transform skip mode results in a lossless quantization mode for the residual signal. When the input bit depth and the internal bit depth are different, a bit depth correction is required. The difference between the input bit depth and the internal bit depth can be implemented by signaling the input bit depth in the higher-level syntax. Alternatively, the difference between the input bit depth and the internal bit depth can be transmitted in the higher-level syntax. Another alternative is to signal the minimum allowed QP for transform skip in the higher-level syntax, which is either an absolute QP or a value relative to the base QP.

[0263] Embodiments relate to a video decoder configured to decode video from a data stream 14 at an internal bit depth and output video at either an input bit depth or an internal bit depth. The video decoder is configured to read syntax elements from the data stream indicating the difference between the internal bit depth and the input bit depth. In parallel, a video encoder is configured to encode video into the data stream 14 at the internal bit depth and receive video input at either the input bit depth or the internal bit depth. The video encoder is configured to encode syntax elements indicating the difference between the internal bit depth and the input bit depth into the data stream 14.

[0264] A bit depth conversion may be performed from the internally reconstructed video version 46 to the input bit depth. The decoder may be configured to perform this conversion before video output, and the encoder may be configured to perform this conversion in the prediction loop 36.

[0265] According to one embodiment, the decoder / encoder is configured to: derive a minimum value of a quantization step size scaling parameter (e.g., quantization parameter 982) based on the difference (e.g., due to a non-zero difference), a QP different from 4 that can result in lossless or nearly lossless coding, and to adhere to the minimum quantization step size scaling parameter for video portions encoded in prediction residual transform skip mode. The decoder / encoder is configured to: map a quantization step size scaling parameter 982 finer than the minimum value (e.g., QP < 4) to a quantization step size scaling parameter 982 equal to the minimum value to achieve lossless coding. For lossless coding, the minimum value can be associated with no quantization 32, 52 or bypassing or disabling quantization 32, 52. The decoder / encoder may be configured to, when observing a minimum quantization step size scaling parameter for a video portion 100 encoded in a prediction residual transform skip mode, change the signaled quantization step size scaling parameter to be equal to the minimum quantization step size scaling parameter if the signaled quantization step size scaling parameter signaled in the data stream 14 for the video portion 100 falls below the minimum quantization step size scaling parameter. The decoder / encoder may only observe the minimum quantization step size scaling parameter within the video region 104 for which the data stream 14 signals a lossless coding mode 1061.

[0266] Another embodiment relates to a video decoder configured to decode video from a data stream 14 at the internal bit depth and output video at either the input bit depth or the internal bit depth, and to read a syntax element from the data stream 14 indicating a minimum value for a quantization step size scaling parameter (QP minimum) (e.g., due to a non-zero difference between the internal and input bit depths). For example, the decoder may be informed of the difference by transmitting the difference or by transmitting two values ​​(i.e., the input bit depth and the internal bit depth) in addition to the minimum QP. A QP different from 4 may result in or nearly result in lossless encoding. In parallel, a video encoder may be configured to perform video encoding into the data stream 14 at the internal bit depth, receive video input having either the input bit depth or the internal bit depth, and encode a syntax element into the data stream 14 indicating the minimum value for the quantization step size scaling parameter. The decoder / encoder may be configured to respect a minimum quantization step size scaling parameter for a video portion 100 encoded in a prediction residual transform skip mode and, optionally, partially change the signaled quantization step size scaling parameter to be equal to the minimum quantization step size scaling parameter if the signaled quantization step size scaling parameter signaled in the data stream 14 for that video portion 100 falls below the minimum quantization step size scaling parameter. The decoder / encoder may respect the minimum quantization step size scaling parameter only within a video region 104 for which the data stream 14 signals a lossless coding mode 1061.

[0267] Another embodiment relates to a video decoder / encoder configured to derive from / encode into a data stream 14 an indication of an internal bit depth and an input bit depth, or a difference therebetween, and to decode / encode video from / into the data stream 14 at the internal bit depth. The decoder is configured to output video at the input bit depth, and the encoder is configured to receive video input at the input bit depth. Furthermore, the decoder / encoder is configured to detect if the internal bit depth has dropped below the input bit depth and to change the internal bit depth to correspond to the input bit depth. Optionally, the decoder / encoder is configured to derive from / encode into the data stream 14 an indication of a lossless coded video portion, perform the detection and change within the lossless coded video portion, and use the internal bit depth (e.g., as derived from the data stream 14) for the lossy coded video portion. The internal bit depth may be signaled by the encoder in the data stream.

[0268] 3.1 Entropy Coding

[0269] Since lossless coding (e.g., QP of 4) typically results in a higher bit rate than lossy coding, the entropy coding engine can be switched to a high-throughput mode to prevent processing bottlenecks. This can be done, for example, by using a different code in binarization or by processing more binary values ​​in a bypass coding mode in CABAC using context modeling and binary arithmetic coding after binarization, rather than in a regular coding mode.

[0270] The embodiment relates to a decoder (encoder) configured to: determine, for a predetermined portion 100 of a picture 12, whether to be (or to be) encoded into a data stream 14 using lossless coding 1061 or lossy coding 1062, and if the predetermined portion 100 of the picture 12 is (or to be) encoded into the data 14 using lossless coding 1061, decode the prediction residual for the predetermined portion 100 from the data stream 14 using a first manner (referred to as residual_ts_coding() in Table 1) using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy decoding (encoding) of the binary value string. and if the predetermined portion 100 of the picture 12 is (to be) encoded into the data stream 14 using lossy coding 1062, decoding the prediction residual for the predetermined portion 100 from the data stream 14 (encoding the prediction residual for the predetermined portion 100 into the data stream) in a second manner (referred to as residual_coding() in Table 1) using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy decoding (encoding) of the binary value string, wherein the first manner is different from the second manner, so that the computational complexity is reduced in the first manner compared to the second manner.

[0271] The determination of whether the predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossless encoding 1061 or lossy encoding 1062 may be performed (e.g., Figure 6 The decoder / encoder described in the foregoing may be based on the data stream 14, or by reading a portion-by-portion transform quantization bypass coding flag, or by a different means. According to one embodiment, the decoder is configured to perform the determination by reading a lossless coding syntax element (e.g., 102) from the data stream 14, the syntax element indicating whether the predetermined portion 100 of the picture 12 or the predetermined region 104 containing the predetermined portion 100 is encoded in the data stream 14 using lossless coding 1061 or lossy coding 1062, and performing the determination based on the lossless coding syntax element. The encoder may encode the lossless coding syntax element.

[0272] Compared with the second approach, the computational complexity of the first approach can be reduced by at least one of the following:

[0273] - the number of binary values ​​of the binary value string encoded using the equally probable bypass mode in the first manner is greater than in the second manner, and

[0274] - Use different binarization codes in the first and second modes, for example, Exponential Golomb code, truncated unary code, etc.

[0275] According to one embodiment, the number of binary values ​​of the binary value string belonging to the unary code or the truncated unary code in the first manner is less than that in the second manner.

[0276] 3.2 Inter-Component Transform (ICT)

[0277] The coefficients of the ICT inverse transform matrix do not allow lossless coding because there is no corresponding forward transform matrix that, in the absence of quantization of the residual signal, leads to perfect lossless reconstruction. It is therefore proposed that when lossless_coding is equal to 1, a different ICT inverse transform matrix (specified as a list of coefficients or a mathematical operation such as a shift or multiplication) be used in the codec than when lossless_coding is equal to 0 (normal operation). Specifically, when lossless_coding is equal to 0, a traditional ICT upmix operation can be used, for example:

[0278] cb=c1+c2,

[0279] cr=sign*(c1-c2),

[0280] where c1 and c2 are the two transmitted residual block signals, and cb and cr are the associated output residual signals produced by the inverse ICT. However, when lossless_coding is equal to 1, a lifting transform [4, 5] or a modulo transform [6] operation can be applied to c1 and c2 instead of the above operations, which allows perfect reconstruction of the cb and cr prediction residuals in the absence of quantization and loop filtering. A different but equivalent way to describe this perfect reconstruction property is to point out that the transform is also mathematically completely reversible, even in practical applications running on, for example, computers (where the transform operation typically requires a rounding-to-integer step that prohibits mathematically perfect reconstruction). For example, forward lossless ICT:

[0281] c1=cb+sign*cr,

[0282] c2=cb-sign*INT(c1 / 2)

[0283] and the corresponding inverse lossless ICT

[0284] cb'=c2+sign*INT(c1 / 2),

[0285] cr'=sign*(c1-cb')

[0286] where INT() represents the floor (rounding towards negative infinity), ceiling (rounding towards positive infinity), or rounding (rounding to the nearest integer) operator, and sign is equal to 1 or -1; perfect reconstruction of both cb and cr is achieved (i.e., cb'=cb, cr'=cr). Therefore, when lossless encoding is required, the above-mentioned inverse lossless ICT operation resulting in cb' and cr' is preferably applied in the decoder. Note that perfect reconstruction can also be achieved by switching between the above-mentioned forward and inverse lossless operations, so that the forward lossless ICT is given by:

[0287] c1=cr+sign*INT(cb / 2),

[0288] c2=sign*(cb-c1)

[0289] And the corresponding inverse lossless ICT is given by:

[0290] cb'=c1+sign*c2,

[0291] cr'=c1-sign*INT(cb' / 2).

[0292] Also note that the + and - signs in the above equations may vary in a particular implementation, but will lead to equivalent results (i.e., cb' = cb, cr' = cr). Finally, it is worth noting that slightly different formulas (e.g., the equivalent formula for integer mid-side (M / S) processing in HD-AAC described in [5]) can be used as the inverse lossless transform during decoding.

[0293] According to one embodiment, the decoder (encoder) is configured to: determine, for a predetermined portion 100 of the picture 12, whether to (or is to) use lossless coding 1061 or lossy coding 1062 to encode it into the data stream 14, and if the predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossless coding 1061, then (for example, in the prediction loop of the encoder) perform a fully reversible transform on the prediction residual 24"', 24"" or the reconstruction 46 corrected by the prediction residual of the predetermined portion 100; and if the predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossy coding 1062, then (for example, in the prediction loop of the encoder) perform a non-fully reversible transform on the prediction residual 24"', 24"" or the reconstruction 46 corrected by the prediction residual of the predetermined portion 100.

[0294] The determination of whether the predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossless encoding 1061 or lossy encoding 1062 may be performed (e.g., Figure 6 The decoder / encoder described in the foregoing may be based on the data stream 14, or by reading a portion-by-portion transform quantization bypass coding flag, or by a different means. According to one embodiment, the decoder is configured to perform the determination by reading a lossless coding syntax element (e.g., 102) from the data stream 14, the syntax element indicating whether the predetermined portion 100 of the picture 12 or the predetermined region 104 containing the predetermined portion 100 is encoded in the data stream 14 using lossless coding 1061 or lossy coding 1062, and performing the determination based on the lossless coding syntax element. The encoder may encode the lossless coding syntax element.

[0295] The fully reversible transform and the non-fully reversible transform can be a transform between color components or a transform within a spectrally decomposed color component.

[0296] 3.3 Smoothing Filters in Some Predictors

[0297] For lossless coding with high bitrates, filters applied to the prediction signal to attenuate compression artifacts (e.g., smoothing filters to reduce quantization artifacts) may not be advantageous. Therefore, in the case of transform skip with quantization skip, these filters can be disabled for lossless coding.

[0298] According to one embodiment, the decoder (encoder) is configured to: determine, for a predetermined portion 100 of the image 12, whether to (is to) be encoded into the data stream 14 using lossless coding 1061 or lossy coding 1062; and if the predetermined portion 100 of the image 12 is (is to) be encoded into the data stream 14 using lossless coding 1061, derive a prediction signal 26 for the predetermined portion 100 in a first manner; and if the predetermined portion 100 of the image 12 is (is to) be encoded into the data stream 14 using lossless coding 1062, derive a prediction signal 26 for the predetermined portion 100 in a second manner, wherein the first manner is different from the second manner, so that the prediction signal 26 is filtered less in the first manner, for example, by an interpolation filter, than in the second manner, or is not filtered in the first manner but is filtered in the second manner.

[0299] The determination of whether the predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossless encoding 1061 or lossy encoding 1062 may be performed (e.g., Figure 6The decoder / encoder described in the foregoing may be based on the data stream 14, or by reading a portion-by-portion transform quantization bypass coding flag, or by a different means. According to one embodiment, the decoder is configured to perform the determination by reading a lossless coding syntax element (e.g., 102) from the data stream 14, the syntax element indicating whether the predetermined portion 100 of the picture 12 or the predetermined region 104 containing the predetermined portion 100 is encoded in the data stream 14 using lossless coding 1061 or lossy coding 1062, and performing the determination based on the lossless coding syntax element. The encoder may encode the lossless coding syntax element.

[0300] According to one embodiment, in terms of low-pass filtering, the prediction signal 26 is filtered less in the first manner than in the second manner, or is not filtered in the first manner but is filtered in the second manner, for example, when derived based on the first manner, the prediction signal 26 has higher energy in the higher frequency half of the entire spatial spectrum of the prediction signal 26 compared to the second manner.

[0301] 3.4 Loop Filter and Quantization

[0302] When lossless_coding is equal to 1, all loop filters (e.g., deblocking, shape adaptive offset (SAO), reshaper) and quantization are disabled, and their related syntax elements (including their respective per-sequence or per-frame activation flags and any QP or QP increment) shall not appear in the coded bitstream.

[0303] According to one embodiment, the decoder (encoder) is configured to: determine, for a predetermined portion 100 of the picture 12, whether to (is to) be encoded into the data stream 14 using lossless coding 1061 or lossy coding 1062, and if the predetermined portion 100 of the picture 12 is (to) be encoded into the data stream 14 using lossless coding 1061, infer that one or more predetermined coding options related to one or more tools of the decoder (encoder) for processing the prediction reconstruction 46 corrected by the prediction residual and / or for the prediction residual requantization 52, 38 (quantization 52) regarding the predetermined portion 100 are to be set, so that the one or more tools are disabled for the predetermined portion 100; and if the predetermined portion 100 of the picture 12 is (to) be encoded into the data stream 14 using lossy coding 1062, infer that the one or more predetermined coding options are to be set by deriving the one or more predetermined coding options from a plurality of coding parameters 98.

[0304] The determination of whether the predetermined portion 100 of the picture 12 is encoded into the data stream 14 using lossless encoding 1061 or lossy encoding 1062 may be performed (e.g., Figure 6The decoder / encoder described in the foregoing may be based on the data stream 14, or by reading a portion-by-portion transform quantization bypass coding flag, or by a different means. According to one embodiment, the decoder is configured to perform the determination by reading a lossless coding syntax element (e.g., 102) from the data stream 14, the syntax element indicating whether the predetermined portion 100 of the picture 12 or the predetermined region 104 containing the predetermined portion 100 is encoded in the data stream 14 using lossless coding 1061 or lossy coding 1062, and performing the determination based on the lossless coding syntax element. The encoder may encode the lossless coding syntax element.

[0305] In the following, different embodiments of the method are described.

[0306] Figure 7 A method 200 for decoding a picture from a data stream (encoding a picture into a data stream) is shown, comprising: determining 106, for a predetermined portion of a picture, whether it is (to be) encoded into the data stream using lossless coding or lossy coding; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossless coding, decoding the prediction residual for the predetermined portion from the data stream (encoding the prediction residual for the predetermined portion into a binary value string) in a first manner 210 using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy coding of the binary value string. and if the predetermined part of the picture is (to be) encoded into the data stream using lossy coding, then in a second manner 220, using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy coding of the binary value string, decoding the prediction residual for the predetermined part from the data stream (encoding the prediction residual for the predetermined part into the data stream), wherein the first manner is different from the second manner, so that the computational complexity of the first manner 210 is reduced compared to the second manner 220.

[0307] Figure 8 A method 300 for decoding a picture from a data stream (encoding a picture into a data stream) is shown, comprising: determining 106, for a predetermined portion of a picture, whether it is (to be) encoded into the data stream using lossless coding or lossy coding; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossless coding, performing a fully reversible transform 310 on a prediction residual or a reconstruction corrected by the prediction residual of the predetermined portion; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossy coding, performing a non-fully reversible transform 320 on a prediction residual or a reconstruction corrected by the prediction residual of the predetermined portion.

[0308] Figure 9A method 400 for decoding a picture from a data stream (encoding a picture into a data stream) is shown, comprising: determining 106 whether a predetermined portion of a picture is to be encoded into the data stream using lossless coding or lossy coding; and deriving a prediction signal for the predetermined portion in a first manner 410 if the predetermined portion of the picture is to be encoded into the data stream using lossless coding; and deriving a prediction signal for the predetermined portion in a second manner 420 if the predetermined portion of the picture is to be encoded into the data stream using lossy coding, wherein the first manner is different from the second manner, such that the prediction signal is less filtered in the first manner 410 than in the second manner 420, for example by an interpolation filter, or is not filtered in the first manner 410 but is filtered in the second manner 420.

[0309] Figure 10 A method 500 for decoding a picture from a data stream (encoding a picture into a data stream) is shown, comprising: determining 106 whether a predetermined portion of a picture is (to be) encoded into the data stream using lossless coding or lossy coding; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossless coding, inferring that one or more predetermined coding options related to one or more tools of a decoder / encoder for processing prediction reconstruction corrected by prediction residuals or for quantizing prediction residuals with respect to the predetermined portion are to be set, such that the one or more tools are disabled for the predetermined portion 510; and if the predetermined portion of the picture is (to be) encoded into the data stream using lossy coding, inferring that the one or more predetermined coding options are to be set by deriving 520 the one or more predetermined coding options 520 from a plurality of coding parameters.

[0310] Figure 11a A method 600 is shown comprising performing 610 video decoding from a data stream at an internal bit depth and performing 620 video output at an input bit depth or an internal bit depth, and reading 630 a syntax element from the data stream indicating a difference between the internal bit depth and the input bit depth. Figure 11b A parallel method 600 is shown, comprising performing 612 video encoding into a data stream at an internal bit depth, receiving 622 a video input having an input bit depth or an internal bit depth, and encoding 632 a syntax element into the data stream indicating a difference between the internal bit depth and the input bit depth.

[0311] Figure 12a A method 700 is shown comprising performing 610 video decoding from a data stream at an internal bit depth and performing 620 video output at an input bit depth or an internal bit depth, and reading 730 a syntax element from the data stream indicating a minimum value for a quantization step size scaling parameter. Figure 12bA parallel method 700 is shown comprising performing 612 video encoding into a data stream at an internal bit depth and receiving 622 a video input having an input bit depth or an internal bit depth, and encoding 732 a syntax element indicating a minimum value of a quantization step size scaling parameter into the data stream.

[0312] Figure 13a A method 800 is shown comprising deriving 810 an indication of an internal bit depth and an input bit depth or a difference therebetween from a data stream, performing 610 video decoding from the data stream at the internal bit depth and performing 620 video output at the input bit depth, checking 820 whether the internal bit depth has dropped below the input bit depth, and changing the internal bit depth to correspond to the input bit depth. Figure 13b A parallel method 800 is shown comprising encoding 812 into a data stream video an indication of an internal bit depth and an input bit depth or a difference therebetween, performing 612 video encoding into said data stream at the internal bit depth and receiving 622 a video input having the input bit depth, checking 820 whether the internal bit depth has dropped below the input bit depth, and changing 830 the internal bit depth to correspond to the input bit depth.

[0313] References

[0314] 【1】ITU-T, Recommendation H.265and ISO / IEC, Int.Standard 23008-2, "Highefficiency video coding," Geneva, Switzerland, Feb. 2018. Online: http: / / www.itu.int / rec / T-REC-H.265.

[0315] 【2】JVET / Fraunhofer HHI, "VVCSoftware VTM: VVC VTM reference software," Jul.2019.Online: https: / / vcgit.hhi.fraunhofer.de / jvet / VVCSoftware_VTM.

[0316] 【3】C.Helmrich, H.Schwarz, T.Nguyen, C.Rudat, D.Marpe, and T.Wiegand, "CE7: Joint Chroma Resi-dual Coding with Multiple Modes (Tests CE7-2.1, CE7-2.2),"Joint Video Experts Group, document JVET-O0105, Gothenburg, June 2019.

[0317] 【4】I.Daubechies and WSweldens, "Factoring Wavelet Transforms into Lifting Steps," J.Four.Anal.Appl., Sep.1996.

[0318] 【5】R.Geiger, R.Yu, J.Herre, S.Rahardja, S.-W.Kim, X.Lin, and M.Schmidt, "ISO / IEC MPEG-4High-Definition Scalable Advanced Audio Coding," J.Aud.Eng.Soc., vol.55, no.1 / 2, pp.27-43, Jan.2007.

[0319] 【6】S.Srinivasan, "Modulo transforms-An alternative to lifting," IEEETrans.on Signal Processing, vol.54, no.5, pp.1864-1874, June 2006.

[0320] Implementation plan options :

[0321] Although some aspects have been described in the context of devices, it is clear that these aspects also represent descriptions of corresponding methods, where blocks or devices correspond to method steps or features of method steps. Similarly, aspects described in the context of method steps also represent descriptions of corresponding blocks or items or features of corresponding devices. Some or all of the method steps can be performed by (or using) hardware devices (e.g., microprocessors, programmable computers, or electronic circuits). In some embodiments, one or more of the most important method steps can be performed by such devices.

[0322] Depending on certain implementation requirements, embodiments of the present invention may be implemented in hardware or software. The implementation may be performed using a digital storage medium, such as a floppy disk, DVD, Blu-ray, CD, ROM, PROM, EPROM, EEPROM, or flash memory, having electronically readable control signals stored thereon that cooperate (or are capable of cooperating) with a programmable computer system to perform the corresponding method. Thus, the digital storage medium may be computer-readable.

[0323] Some embodiments according to the invention comprise a data carrier having electronically readable control signals, which are capable of cooperating with a programmable computer system, such that one of the methods described herein is performed.

[0324] Generally, embodiments of the present invention can be implemented as a computer program product with a program code, the program code being operative for performing one of the methods when the computer program product runs on a computer.The program code may, for example, be stored on a machine-readable carrier.

[0325] Other embodiments comprise the computer program for performing one of the methods described herein, stored on a machine readable carrier.

[0326] In other words, an embodiment of the inventive method is, therefore, a computer program having a program code for performing one of the methods described herein, when the computer program runs on a computer.

[0327] A further embodiment of the inventive method is, therefore, a data carrier (or a digital storage medium, or a computer-readable medium) comprising, recorded thereon, the computer program for performing one of the methods described herein. The data carrier, digital storage medium or recorded medium is typically tangible and / or non-transitory.

[0328] Therefore, a further embodiment of the inventive method is a data stream or a sequence of signals representing the computer program for performing one of the methods described herein.The data stream or the sequence of signals may, for example, be configured to be transmitted via a data communication connection, for example via the Internet.

[0329] A further embodiment comprises a processing means, for example a computer or a programmable logic device, configured to or adapted to perform one of the methods described herein.

[0330] A further embodiment comprises a computer having installed thereon the computer program for performing one of the methods described herein.

[0331] Another embodiment according to the present invention comprises an apparatus or system configured to transfer (e.g., electronically or optically) a computer program for performing one of the methods described herein to a receiver. For example, the receiver may be a computer, a mobile device, a storage device, etc. For example, the apparatus or system may comprise a file server for transferring the computer program to the receiver.

[0332] In some embodiments, a programmable logic device (e.g., a field programmable gate array) can be used to perform some or all of the functions of the methods described herein. In some embodiments, the field programmable gate array can collaborate with a microprocessor to perform one of the methods described herein. Typically, these methods are preferably performed by any hardware device.

[0333] The apparatus described herein may be implemented by hardware devices, by computers, or by a combination of hardware devices and computers.

[0334] The apparatus described herein or any component of an apparatus described herein may be implemented at least partially in hardware and / or software.

[0335] The methods described herein may be performed using a hardware device, or using a computer, or using a combination of a hardware device and a computer.

[0336] Any component of a method described herein or an apparatus described herein may be performed at least in part by hardware and / or software.

[0337] The above embodiments are intended to illustrate the principles of the present invention only. It should be understood that modifications and variations of the arrangements and details described herein will be apparent to those skilled in the art. Accordingly, it is intended that the present invention be limited solely by the scope of the forthcoming patent claims, and not by the specific details presented through the description and explanation of the embodiments herein.

Claims

1. A decoder for decoding a picture from a data stream, comprising: processor, and a memory storing instructions that, when executed by the processor, cause the decoder to: checking (106) whether a plurality of coding parameters (98) contained in a data stream (14) indicates a coding parameter setting corresponding to lossless prediction residual coding, said plurality of coding parameters (98) relating to a predetermined portion (100) of a picture (12) and controlling a prediction residual transform mode and a quantization precision with respect to said predetermined portion; reading a lossless coding syntax element (102) from the data stream, the lossless coding syntax element (102) indicating whether a predetermined area (104) of a picture is encoded into the data stream using lossless coding or lossy coding, the predetermined area (104) covering or including the predetermined portion; If the lossless coding syntax element indicates that the predetermined region of the picture is encoded into the data stream using lossless coding, and if the plurality of coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, setting one or more predetermined coding options associated with one or more tools of the decoder for processing prediction reconstruction corrected by prediction with respect to the predetermined portion such that the one or more tools are disabled with respect to the predetermined portion; as well as If the multiple coding parameters are not equal to the coding parameter settings corresponding to lossless prediction residual coding, or the lossless coding syntax element indicates that a predetermined area of ​​the picture is encoded into the data stream using lossy coding, the one or more predetermined coding options are set to a predetermined tool state.

2. The decoder of claim 1, wherein the instructions, when executed by the processor, further cause the decoder to: determine the predetermined tool state based on one or more syntax elements (108) in the data stream.

3. The decoder of claim 1 , wherein the instructions, when executed by the processor, further cause the decoder to: in response to the plurality of encoding parameter indications corresponding to encoding parameter settings for lossless prediction residual encoding, set one or more other encoding options with respect to the predetermined portion to a default state.

4. The decoder of claim 3 , wherein the default state represents the following: a reduction in filtering or a disabling of filtering in terms of low-pass filtering for deriving the prediction signal (26) for the predetermined portion; and / or A fully reversible transform is to be performed on the prediction residual signal.

5. The decoder of claim 3 , wherein the other encoding options relate to: Binarization of the prediction residual data into a binary value string and context-adaptive binary entropy decoding of the binary value string; Use of a fully reversible transformation of the prediction residual or the prediction residual-corrected reconstruction (46) of said predetermined portion; disabling or reducing filtering for deriving a prediction signal (26) for said predetermined portion; and / or Disabling of a process of requantizing the prediction residual or reconstructing the prediction with respect to the predetermined portion of the prediction residual.

6. The decoder of claim 1 , wherein the instructions, when executed by the processor, further cause the decoder to: If the multiple coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, then the prediction residual for the predetermined portion is decoded from the data stream using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy decoding of the binary value string in a first manner; and if the multiple coding parameters do not indicate coding parameter settings corresponding to lossless prediction residual coding, then the prediction residual for the predetermined portion is decoded from the data stream using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy decoding of the binary value string in a second manner, wherein the first manner is different from the second manner, so that the computational complexity is reduced in the first manner compared to the second manner.

7. The decoder of claim 1 , wherein the instructions, when executed by the processor, further cause the decoder to: If the plurality of encoding parameters indicate encoding parameter settings corresponding to lossless prediction residual encoding, a fully reversible transform is performed on the prediction residual of the predetermined portion or the reconstruction (46) corrected by the prediction residual; and if the plurality of encoding parameters do not indicate encoding parameter settings corresponding to lossless prediction residual encoding, a non-fully reversible transform is performed on the prediction residual of the predetermined portion or the reconstruction (46) corrected by the prediction residual.

8. The decoder of claim 1 , wherein the instructions, when executed by the processor, further cause the decoder to: If the plurality of encoding parameters indicate an encoding parameter setting corresponding to lossless prediction residual encoding, a prediction signal (26) for the predetermined portion is derived in a first manner, and if the plurality of encoding parameters do not indicate an encoding parameter setting corresponding to lossless prediction residual encoding, a prediction signal (26) for the predetermined portion is derived in a second manner, wherein the first manner is different from the second manner, such that the prediction signal is less filtered in the first manner than in the second manner, or is not filtered in the first manner but is filtered in the second manner.

9. The decoder of claim 1 , wherein the instructions, when executed by the processor, further cause the decoder to: If the multiple coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, it is inferred that one or more predetermined coding options related to one or more tools of the decoder for processing prediction reconstruction corrected by prediction residual or for requantization of prediction residual with respect to the predetermined part are to be set, so that the one or more tools are disabled with respect to the predetermined part; and if the multiple coding parameters do not indicate coding parameter settings corresponding to lossless prediction residual coding, it is inferred that the one or more predetermined coding options related to one or more tools of the decoder for processing prediction reconstruction corrected by prediction residual or for requantization of prediction residual with respect to the predetermined part are to be set by deriving the one or more predetermined coding options related to the one or more tools of the decoder for processing prediction reconstruction corrected by prediction residual or for requantization of prediction residual with respect to the predetermined part from the multiple coding parameters.

10. The decoder of claim 1 , wherein the instructions, when executed by the processor, further cause the decoder to: If the plurality of encoding parameters indicate encoding parameter settings corresponding to lossless prediction residual encoding, setting one or more predetermined encoding options with respect to the predetermined portion such that the one or more tools are disabled with respect to the predetermined portion, and If the plurality of encoding parameters do not indicate an encoding parameter setting corresponding to lossless prediction residual encoding, one or more predetermined encoding options with respect to the predetermined portion are set to a predetermined tool state.

11. The decoder of claim 10, wherein the instructions, when executed by the processor, further cause the decoder to: determine the predetermined tool state based on one or more syntax elements in the data stream.

12. The decoder of claim 2, wherein: If the one or more predetermined coding options with respect to the predetermined portion are set such that the one or more tools are disabled with respect to the predetermined portion, at least one syntax element of the one or more syntax elements is not present in the data stream.

13. The decoder of claim 1 , wherein the instructions, when executed by the processor, further cause the decoder to: Reading the plurality of encoding parameters from the data stream, Checking whether a prediction residual transform mode indicated by the plurality of encoding parameters corresponds to a transform skip mode and to a quantization step size finer than a predetermined quantization step size, wherein the predetermined quantization step size corresponds to no quantization, and if so, changing the quantization step size to the predetermined quantization step size.

14. An encoder for encoding a picture into a data stream, comprising: processor, and a memory storing instructions that, when executed by the processor, cause the encoder to: Signaling a plurality of encoding parameters (98) in a data stream (14); encoding a lossless coding syntax element (102) into the data stream, the lossless coding syntax element (102) indicating whether a predetermined area (104) of the picture is to be encoded into the data stream using lossless coding or lossy coding, the predetermined area covering or containing a predetermined portion of the picture; checking (106) whether the plurality of coding parameters (98) indicate coding parameter settings corresponding to lossless prediction residual coding, the plurality of coding parameters (98) relating to a predetermined portion (100) of a picture (12) and controlling a prediction residual transform mode and a quantization precision with respect to the predetermined portion; If the lossless coding syntax element indicates that a predetermined region of the picture is to be encoded into the data stream using lossless coding, and if the plurality of coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, setting one or more predetermined coding options associated with one or more tools of the encoder for processing prediction reconstruction corrected by prediction residuals with respect to the predetermined region such that the one or more tools are disabled with respect to the predetermined region in a prediction loop of the encoder; as well as If the multiple coding parameters are not equal to the coding parameter settings corresponding to lossless prediction residual coding, or the lossless coding syntax element indicates that a predetermined area of ​​the picture is to be encoded into the data stream using lossy coding, setting the one or more predetermined coding options to a predetermined tool state.

15. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: determining one or more syntax elements (108) and encoding them into the data stream; and The predetermined tool state is determined based on the one or more syntax elements (108).

16. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: in response to the plurality of encoding parameter indications corresponding to encoding parameter settings for lossless prediction residual encoding, set one or more other encoding options with respect to the predetermined portion to a default state.

17. The encoder of claim 16, wherein: The default state indicates the following: a reduction in filtering or a disabling of filtering in terms of low-pass filtering for deriving the prediction signal (26) for the predetermined portion; and / or A fully reversible transform is to be performed on the prediction residual signal.

18. The encoder of claim 16, wherein: The other encoding options mentioned relate to: Binarization of the prediction residual data into a binary value string and context-adaptive binary entropy coding of the binary value string; use of a fully reversible transform of the prediction residual or a reconstruction (46) corrected by the prediction residual of the predetermined portion in a prediction loop of the encoder; disabling or reducing filtering for deriving a prediction signal (26) for said predetermined portion; and / or Disabling of a process of quantizing the prediction residual or correcting the prediction reconstruction with respect to the predetermined portion.

19. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: If the multiple coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, the prediction residual for the predetermined part is encoded into the data stream using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy coding of the binary value string in a first manner; and if the multiple coding parameters do not indicate coding parameter settings corresponding to lossless prediction residual coding, the prediction residual for the predetermined part is encoded into the data stream using binarization of the prediction residual data into a binary value string and context-adaptive binary entropy coding of the binary value string in a second manner, wherein the first manner is different from the second manner, so that the computational complexity is reduced in the first manner compared to the second manner.

20. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: If the plurality of encoding parameters indicate encoding parameter settings corresponding to lossless prediction residual encoding, a fully reversible transform is performed on the prediction residual of the predetermined portion or the reconstruction (46) corrected by the prediction residual; and if the plurality of encoding parameters do not indicate encoding parameter settings corresponding to lossless prediction residual encoding, a non-fully reversible transform is performed on the prediction residual of the predetermined portion or the reconstruction (46) corrected by the prediction residual.

21. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: If the plurality of encoding parameters indicate an encoding parameter setting corresponding to lossless prediction residual encoding, the prediction signal (26) for the predetermined portion is encoded in a first manner, and if the plurality of encoding parameters do not indicate an encoding parameter setting corresponding to lossless prediction residual encoding, the prediction signal (26) for the predetermined portion is encoded in a second manner, wherein the first manner is different from the second manner, such that the prediction signal is less filtered in the first manner than in the second manner, or is not filtered in the first manner but is filtered in the second manner.

22. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: If the multiple coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, it is inferred that one or more predetermined coding options related to one or more tools of the encoder for processing prediction reconstruction corrected by prediction residual or for quantizing prediction residual with respect to the predetermined part are to be set, so that the one or more tools are disabled with respect to the predetermined part; and if the multiple coding parameters do not indicate coding parameter settings corresponding to lossless prediction residual coding, it is inferred that the one or more predetermined coding options related to one or more tools of the encoder for processing prediction reconstruction corrected by prediction residual or for quantizing prediction residual with respect to the predetermined part are to be set by deriving the one or more predetermined coding options related to the one or more tools of the encoder for processing prediction reconstruction corrected by prediction residual or for quantizing prediction residual with respect to the predetermined part from the multiple coding parameters.

23. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: If the plurality of encoding parameters indicate encoding parameter settings corresponding to lossless prediction residual encoding, setting one or more predetermined encoding options with respect to the predetermined portion such that the one or more tools are disabled with respect to the predetermined portion, and If the plurality of encoding parameters do not indicate an encoding parameter setting corresponding to lossless prediction residual encoding, one or more predetermined encoding options with respect to the predetermined portion are set to a predetermined tool state.

24. The encoder of claim 23, wherein the instructions, when executed by the processor, further cause the encoder to: determining one or more syntax elements (108) and encoding them into the data stream; and The predetermined tool state is determined based on the one or more syntax elements.

25. The encoder of claim 15, wherein at least one of the one or more syntax elements is not encoded into the data stream if the one or more predetermined encoding options with respect to the predetermined portion are set such that the one or more tools are disabled with respect to the predetermined portion.

26. The encoder of claim 14, wherein the instructions, when executed by the processor, further cause the encoder to: signaling said plurality of encoding parameters in said data stream, Checking whether a prediction residual transform mode indicated by the plurality of encoding parameters corresponds to a transform skip mode and a quantization step size finer than a predetermined quantization step size, wherein the predetermined quantization step size corresponds to no quantization, and if so, changing the quantization step size to the predetermined quantization step size.

27. A method for decoding a picture from a data stream, comprising: checking (106) whether a plurality of coding parameters (98) contained in a data stream (14) indicates a coding parameter setting corresponding to lossless prediction residual coding, said plurality of coding parameters (98) relating to a predetermined portion (100) of a picture (12) and controlling a prediction residual transform mode and a quantization precision with respect to said predetermined portion; reading a lossless coding syntax element (102) from the data stream, the lossless coding syntax element (102) indicating whether a predetermined area (104) of a picture is encoded into the data stream using lossless coding or lossy coding, the predetermined area (104) covering or including the predetermined portion; If the lossless coding syntax element indicates that a predetermined region of the picture is encoded into the data stream using lossless coding, and if the plurality of coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, setting one or more predetermined coding options associated with one or more tools of a decoder for processing prediction reconstruction corrected by prediction with respect to the predetermined portion such that the one or more tools are disabled with respect to the predetermined portion; as well as If the multiple coding parameters are not equal to the coding parameter settings corresponding to lossless prediction residual coding, or the lossless coding syntax element indicates that a predetermined area of ​​the picture is encoded into the data stream using lossy coding, the one or more predetermined coding options are set to a predetermined tool state.

28. A method for encoding a picture into a data stream, comprising: Signaling a plurality of encoding parameters (98) in a data stream (14); encoding a lossless coding syntax element (102) into the data stream, the lossless coding syntax element (102) indicating whether a predetermined area (104) of the picture is to be encoded into the data stream using lossless coding or lossy coding, the predetermined area covering or containing a predetermined portion of the picture; checking (106) whether the plurality of coding parameters (98) indicate coding parameter settings corresponding to lossless prediction residual coding, the plurality of coding parameters (98) relating to a predetermined portion (100) of a picture (12) and controlling a prediction residual transform mode and a quantization precision with respect to the predetermined portion; If the lossless coding syntax element indicates that a predetermined region of the picture is to be encoded into the data stream using lossless coding, and if the plurality of coding parameters indicate coding parameter settings corresponding to lossless prediction residual coding, setting one or more predetermined coding options associated with one or more tools of an encoder for processing prediction reconstruction corrected by a prediction residual with respect to the predetermined portion, such that the one or more tools are disabled with respect to the predetermined portion in a prediction loop of the encoder; as well as If the multiple coding parameters are not equal to the coding parameter settings corresponding to lossless prediction residual coding, or the lossless coding syntax element indicates that a predetermined area of ​​the picture is to be encoded into the data stream using lossy coding, setting the one or more predetermined coding options to a predetermined tool state.

Images