Video information notification method and apparatus, and decoding method and apparatus using the same

Abstract

To provide a method for signaling image information in coding / decoding image information, and a decoding method using the method.SOLUTION: The method for signaling image information according to the present invention comprises steps of: performing inter-prediction for a current picture; and signaling information including a result of the inter-prediction and reference picture information indicating reference pictures usable in the inter prediction. The reference picture information contains POC information of the usable reference pictures. The POC information of the usable reference pictures in the reference picture information is configured such that POCs for the pictures existing before the current picture in terms of a POC sequence are located at the front, and POCs for the pictures existing after the current picture in terms of the POC sequence are located following the POCs located at the front.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to video compression technology, and more specifically, to a method and apparatus for efficiently signaling video information, and a decoding method and apparatus using the same.

Background Art

[0002] Recently, the demand for high-resolution and high-quality video has been increasing in various application fields. However, as the video has higher resolution and quality, the amount of information related to the corresponding video also increases.

[0003] Therefore, when transferring video information using a medium such as a conventional wired or wireless broadband line, or storing video information using an existing storage medium, the transfer cost and storage cost of information increase.

[0004] In order to effectively transfer, store, or reproduce high-resolution and high-quality video information, a highly efficient video compression technology can be used.

[0005] To improve the efficiency of video compression, inter prediction and intra prediction can be used. In the inter prediction method, the pixel value of the current picture is predicted by referring to the information of other pictures, and in the intra prediction method, the pixel value is predicted by using the relationship between pixels within the same picture.

[0006] When adopting inter prediction, the encoding device and the decoding device perform prediction based on a reference picture list that indicates the reference pictures available for the current block (current picture).

[0007] The information for creating the reference picture list is transferred from the encoding device to the decoding device. The decoding device can create a reference picture list based on the information received from the encoding device and effectively perform inter prediction.

Summary of the Invention

Problems to be Solved by the Invention

[0008] The present invention aims to provide a method and apparatus for effectively notifying video information in the encoding / decoding of video information.

[0009] The present invention aims to provide a method and apparatus for effectively notifying information for interface prediction in the encoding / decoding of video information.

[0010] The present invention aims to provide a method and apparatus for effectively notifying information for creating a reference picture list for performing interface prediction.

[0011] The present invention aims to provide a method and apparatus for effectively configuring a reference picture list for interface prediction based on received information. [Means for solving the problem]

[0012] One embodiment of the present invention is a video information notification method comprising the steps of performing an interaction prediction for the current picture and notifying information including the interaction prediction result and reference picture information indicating reference pictures available in the interaction prediction, wherein the reference picture information includes picture order count (POC) information of the available reference pictures, and in the reference picture information, the POC information of the available reference pictures is arranged in POC order, with the POCs for pictures prior to the current picture placed at the beginning, followed by the POCs for pictures after the current picture, also in POC order.

[0013] At this time, the POC information is arranged in POC order, with reference pictures prior to the current picture being sorted in descending order of their POC, and with reference pictures after the current picture being sorted in ascending order of their POC.

[0014] The POC information may be the difference in POC between the target reference picture and other reference pictures indicated by the reference picture information, and the POC information of the reference pictures in the reference picture information can be sorted based on the POC of each target reference picture.

[0015] The POC information of a reference picture includes the magnitude and sign of the POC difference between the target reference picture and the reference picture among the reference pictures indicated by the reference picture information, and the POC information of the reference picture can be sorted in the reference picture information based on the POC of the target reference picture.

[0016] In this case, within the reference picture information, the POC information is arranged in POC order, descending order of the POC of the reference picture for reference pictures prior to the current picture, and ascending order of the POC of the reference picture for reference pictures after the current picture. If the target reference picture is one of the reference pictures indicated by the reference picture information that, in POC order, is the picture closest to the current picture among the pictures prior to the current picture, or the picture closest to the current picture among the pictures after the current picture, then the base picture is the current picture. If the target reference picture is not one of the reference pictures indicated by the reference picture information that, in POC order, is either the picture closest to the current picture among the pictures prior to the current picture, or the picture closest to the current picture among the pictures after the current picture, then the base picture may be the reference picture in the reference picture information that corresponds to the POC information immediately preceding the target reference picture's POC information.

[0017] Furthermore, in this case, the sign of the POC difference may be the sign of the difference between the POC of the reference picture and the POC of the current picture.

[0018] In the reference picture information, the POC information of the reference pictures available may be information indicating the magnitude of the POC difference between the target reference picture and the reference picture, the number of cases where the sign of the POC difference is negative, and the number of cases where the sign of the POC difference is positive, among the reference pictures indicated by the reference picture information. The POC information of the reference pictures in the reference picture information can be sorted based on the POC of the target reference picture.

[0019] In this case, the sign of the POC difference may be the sign of the difference between the POC of the reference picture and the POC of the current picture.

[0020] Furthermore, in this case, the POC information within the reference picture information is arranged in POC order, in descending order of POC for reference pictures prior to the current picture, and in POC order, in ascending order of POC for reference pictures after the current picture. If the target reference picture is one of the reference pictures indicated by the reference picture information that, in POC order, is the picture closest to the current picture among the pictures prior to the current picture, or the picture closest to the current picture among the pictures after the current picture, then the base picture is the current picture. If the target reference picture is not one of the reference pictures indicated by the reference picture information that, in POC order, is neither the picture closest to the current picture among the pictures prior to the current picture, or the picture closest to the current picture among the pictures after the current picture, then the base picture may be the reference picture in the reference picture information that corresponds to the POC information immediately preceding the target reference picture's POC information.

[0021] Another embodiment of the present invention is a method for decoding video information, comprising the steps of: entropy decoding the information of a received bitstream to obtain reference picture information, which includes POC information of reference pictures available for predicting the current picture; and making a prediction for the current block using a reference picture list configured based on the POC of each reference picture derived from the reference picture information, wherein in the reference picture information, the POC information of the reference pictures is arranged in POC order, with the POC for pictures prior to the current picture placed at the beginning, followed by the POC for pictures after the current picture, also in POC order.

[0022] The POC information is arranged in POC order. For referenced pictures prior to the current picture, the POC is sorted in descending order of the referenced picture's POC. For referenced pictures after the current picture, the POC is sorted in ascending order of the referenced picture's POC.

[0023] The i-th (i is an integer) POC information, POCi, within the reference picture information is the POC information of the reference picture Pi. POCi includes the magnitude of the POC difference between Pi and the reference picture within the reference picture information. Within the reference picture information, the POC information of the reference picture can be sorted based on the POC of the target reference picture.

[0024] At this time, within the reference picture information, the POC information is arranged in POC order. For reference pictures before the current picture, it is in descending order of the POC of the reference picture; for reference pictures after the current picture, it is in ascending order of the POC of the reference picture. If Pi is one of the reference pictures indicated by the reference picture information, and among the pictures before the current picture in POC order, it is the picture closest to the current picture, and among the pictures after the current picture, it is the picture closest to the current picture, then the reference picture is the current picture. If Pi is not one of the pictures closest to the current picture among the pictures before the current picture in POC order and the pictures after the current picture among the reference pictures indicated by the reference picture information, the reference picture may be the reference picture corresponding to the (i - 1)-th POC information in the reference picture information.

[0025] Also, at this time, the POC information can include information indicating the sign of the difference between the POC of the target reference picture and the POC of the current picture.

[0026] The i-th (i is an integer) POC information POCi in the reference picture information is the POC information of the reference picture Pi. POCi includes the magnitude of the POC difference between Pi and the reference picture in the reference picture information. In the reference picture information, the POC information of the reference picture can be sorted based on the POC of the target reference picture.

[0027] At this time, within the reference picture information, the POC information is arranged in POC order. For reference pictures before the current picture, it is in descending POC order of the reference pictures, and for reference pictures after the current picture, it is in ascending POC order of the reference pictures. Among the reference pictures indicated by the reference picture information, when Pi is either the picture closest to the current picture among the pictures before the current picture in POC order or the picture closest to the current picture among the pictures after the current picture, the reference picture is the current picture. When Pi is not either of these, the reference picture may be the reference picture corresponding to the (i - 1)-th POC information in the reference picture information.

[0028] Also, at this time, the reference picture information can include information indicating the order relationship between the POC of each target reference picture and the POC of the current picture.

[0029] Also, the POC information includes POC difference information and sign information. Among the m reference pictures indicated by the reference picture information, when the number of reference pictures before the current picture in POC order is n, for the POC of the reference picture k corresponding to the k-th (0 ≤ k ≤ n - 1) POC difference information among the POC difference information, it may be the difference between the POC of the first reference picture and POCk, and for the POC of the reference picture j corresponding to the j-th (n ≤ j ≤ m) POC difference information among the POC difference information, it may be the sum of the POC of the second reference picture and POCj.

[0030] At this time, when k is 0, the first reference picture is the current picture. When k is not 0, the first reference picture is the reference picture corresponding to the (k - 1)-th POC difference information. When j is n, the second reference picture is the current picture. When j is not n, the second reference picture may be the reference picture corresponding to the (n - 1)-th POC difference information.

Advantages of the Invention

[0031] According to the present invention, video information can be effectively communicated during the encoding / decoding of video information.

[0032] According to the present invention, information for creating a reference picture list for interaction prediction can be effectively communicated.

[0033] According to the present invention, the transfer overhead when transferring information for creating a reference picture list can be reduced.

[0034] According to the present invention, after receiving information for creating a reference picture list, a reference picture list for interface prediction can be effectively constructed with low complexity based on the received information. [Brief explanation of the drawing]

[0035] [Figure 1] This is a schematic block diagram showing an encoding device (video encoding device) according to one embodiment of the present invention. [Figure 2] This is a schematic block diagram showing an image decoding device according to one embodiment of the present invention. [Figure 3] This diagram provides a schematic explanation of an example of a candidate block that can be used when performing inter-prediction on a block. [Figure 4] This diagram schematically illustrates an example of a set of reference pictures notified from the encoding device to the decoding device. [Figure 5] This figure shows an example of a reference relationship between B-pictures that perform bidirectional prediction. [Figure 6] This diagram schematically illustrates an example of a reference relationship between picture B and picture P. [Figure 7] This flowchart provides a schematic explanation of the encoding method performed by the encoding device according to the present invention. [Figure 8] This flowchart outlines the decoding method performed by the decoding device according to the present invention. [Modes for carrying out the invention]

[0036] Because the present invention can be modified in various ways and has many different embodiments, specific embodiments are illustrated in the drawings and described in detail. However, this does not limit the present invention to any particular embodiment. The terms used herein are used to describe specific embodiments and are not intended to limit the technical idea of ​​the present invention. Singular expressions include plural expressions unless the context makes it clear. In this specification, terms such as “includes” or “has” indicate the existence of features, figures, stages, operations, components, parts or combinations thereof described in the specification, and should be understood not to preclude the existence or possibility of adding one or more other features, figures, stages, operations, components, parts or combinations thereof.

[0037] On the other hand, each configuration shown in the drawings described in this invention is shown independently for the convenience of describing distinct characteristic functions in the video encoding / decoding device, and does not mean that each configuration is embodied by other hardware or other software. For example, two or more of the configurations may be combined to form a single configuration, or one configuration may be divided into multiple configurations. Embodiments in which each configuration is integrated and / or separated are also included within the scope of the invention, as long as they do not depart from the essence of the invention.

[0038] Preferred embodiments of the present invention will be described in further detail below with reference to the attached drawings. Hereafter, the same reference numerals will be used for the same components in the drawings, and redundant descriptions of the same components will be omitted.

[0039] Figure 1 is a schematic block diagram showing an encoding device (video encoding device) according to one embodiment of the present invention. As shown in Figure 1, the encoding device 100 includes a picture division unit 105, a prediction unit 110, a conversion unit 115, a quantization unit 120, a realignment unit 125, an entropy encoding unit 130, an inverse quantization unit 135, an inverse conversion unit 140, a filter unit 145, and a memory 150.

[0040] The picture splitting unit 105 can split the input picture into at least one processing unit block. In this case, the block as a processing unit may be a prediction unit (hereinafter referred to as PU), a conversion unit (hereinafter referred to as TU), or an encoding unit (hereinafter referred to as CU).

[0041] As described later, the prediction unit 110 includes an inter-prediction unit that performs inter-prediction and an intra-prediction unit that performs intra-prediction. The prediction unit 110 performs predictions on the processing units of the picture in the picture division unit 105 and generates prediction blocks. The processing units of the picture in the prediction unit 110 may be CU, TU, or PU. The prediction unit 110 can also determine whether the prediction performed on the processing unit is an inter-prediction or an intra-prediction, and can determine the specific details of each prediction method (e.g., prediction mode). In this case, the processing unit on which the prediction is performed and the processing unit on which the prediction method and the specific details of the prediction method are determined may be different. For example, the prediction method and prediction mode may be determined at the PU level, and the prediction may be performed at the TU level.

[0042] Interpretation can generate prediction blocks by making predictions based on information from at least one picture among the previous and / or subsequent pictures of the current picture. Intrapretation can generate prediction blocks by making predictions based on pixel information within the current picture.

[0043] Interpretation methods include skip mode, merge mode, and motion vector prediction (MVP). In interpretation, a reference picture can be selected for the PU, and a reference block of the same size as the PU can be selected. The reference block can be selected in integer pixel units. Next, a prediction block is generated that minimizes the residual signal with the current PU and also minimizes the magnitude of the motion vector.

[0044] Prediction blocks may be generated in integer sample units, or in units of less than an integer pixel, such as 1 / 2 pixel units or 1 / 4 pixel units. In this case, the motion vector can also be represented in units of less than an integer pixel. For example, luminance samples can be represented in 1 / 4 pixel units, and chrominance samples can be represented in 1 / 8 pixel units.

[0045] Information such as the index of the reference picture selected via interface prediction, motion vector (e.g., MVP), and residual signal is entropi-encoded and transmitted to the decoder. If skip mode is employed, the residual can be converted from the prediction block to the reconstruction block, thus eliminating the need to generate, transform, quantize, and transfer the residual.

[0046] When performing intra-prediction, the prediction mode is determined at the PU (Power Unit) level, and predictions can be performed at the PU level. Alternatively, the prediction mode can be determined at the PU level, and intra-prediction can be performed at the TU (Terminal Unit) level.

[0047] In intra-prediction, the prediction mode can have 33 directional prediction modes and at least 2 non-directional modes. The non-directional modes can include DC prediction modes and planar modes.

[0048] In intra-prediction, a prediction block can be generated after applying a filter to the reference sample. Whether or not to apply a filter to the reference sample can be determined by the intra-prediction mode and / or size of the current block.

[0049] PUs can be blocks of various sizes and shapes. For example, in the case of interpretation, PUs can be 2N×2N blocks, 2N×N blocks, N×2N blocks, or N×N blocks (where N is an integer). In the case of intraprediction, PUs can be 2N×2N blocks or N×N blocks (where N is an integer). In this case, N×N block-sized PUs can be configured to be applied only in specific cases. For example, it can be decided to use N×N block-sized PUs only for the smallest size CU, or to use them only for intraprediction. In addition to the sizes of PUs mentioned above, further PUs such as N×mN blocks, mN×N blocks, 2N×mN blocks, or mN×2N blocks (m<1) can also be defined and used.

[0050] The residual value (residual block or residual signal) between the generated predicted block and the original block is input to the conversion unit 115. In addition, the prediction mode information, motion vector information, etc., used for prediction are encoded together with the residual value in the entropy encoding unit 130 and transmitted to the decoding device.

[0051] The transformation unit 115 performs a transformation on the residual block in transformation units and generates transformation coefficients. The transformation unit in the transformation unit 115 may be a TU and may have a quad tree structure. In this case, the size of the transformation unit can be determined within a predetermined range of maximum and minimum sizes. The transformation unit 115 can transform the residual block using the discrete cosine transform (DCT) and / or discrete sine transform (DST).

[0052] The quantization unit 120 can generate quantization coefficients by quantizing the residual values ​​converted by the conversion unit 115. The values ​​calculated by the quantization unit 120 are provided to the inverse quantization unit 135 and the realignment unit 125.

[0053] The re-arrangement unit 125 re-arranges the quantization coefficients provided by the quantization unit 120. By re-arranging the quantization coefficients, the encoding efficiency in the entropy encoding unit 130 can be increased. The re-arrangement unit 125 can re-arrange the 2D block-shaped quantization coefficients into a 1D vector-shaped form using the coefficient scanning method. The re-arrangement unit 125 can also increase the entropy encoding efficiency in the entropy encoding unit 130 by changing the order of the coefficient scans based on the probabilistic statistics of the coefficients transferred from the quantization unit.

[0054] The entropy coding unit 130 can perform entropy coding on the quantization coefficients realigned by the realignment unit 125. Entropy coding can be performed using coding methods such as Exponential Golomb, Context-Adaptive Variable-Length Coding (CAVLC), and Context-Adaptive Binary Arithmetic Coding (CABAC). The entropy coding unit 130 can encode various types of information transmitted from the realignment unit 125 and the prediction unit 110, including quantization coefficient information and block type information, prediction mode information, division unit information, PU information and transfer unit information, motion vector information, reference picture information, block interpolation information, and filter information.

[0055] Furthermore, the entropy encoding unit 130 can, if necessary, make certain changes to the set of parameters or syntax to be transmitted.

[0056] The inverse quantization unit 135 inversely quantizes the value quantized by the quantization unit 120, and the inverse transformation unit 140 inversely transforms the value inversely quantized by the inverse quantization unit 135. The residual values ​​generated by the inverse quantization unit 135 and the inverse transformation unit 140 can be combined with the predicted block predicted by the prediction unit 110 to generate a restored block.

[0057] Figure 1 illustrates how a reconstruction block is generated by combining the residual block and the predicted block via an adder. In this case, the adder can also be considered as a separate unit (reconstruction block generation unit) that generates the reconstruction block.

[0058] The filter section 145 can apply a deblocking filter, an adaptive loop filter (ALF), and a sample-adaptive offset (SAO) to the restored picture.

[0059] Block distortion removal filters can remove distortion at the boundaries between blocks in the restored picture. After blocks have been removed via the block distortion removal filter, ALF can perform filtering based on a comparison between the restored image and the original image. ALF may only be used when high efficiency is required. SAO restores the offset difference from the original image on a pixel-by-pixel basis for residual blocks treated with the block distortion removal filter, and is employed in forms such as band offset and edge offset.

[0060] On the other hand, the filter unit 145 does not need to apply a filter to the reconstruction block used for interface prediction.

[0061] Memory 150 can store the restored blocks or pictures calculated via the filter unit 145. The restored blocks or pictures stored in memory 150 can be provided to the prediction unit 110, which performs interface prediction.

[0062] Figure 2 is a schematic block diagram showing an image decoding device according to one embodiment of the present invention. As shown in Figure 2, the image decoding device 200 may include an entropy decoding unit 210, a realignment unit 215, an inverse quantization unit 220, an inverse transformation unit 225, a prediction unit 230, a filter unit 235, and a memory 240.

[0063] When a video bitstream is input to a video encoding device, the input bitstream can be decoded according to the procedure by which the video information was processed in the video encoding device.

[0064] For example, if a variable-length coding method such as CAVLC (hereinafter referred to as VLC) is used to perform entropy coding in a video coding device, the entropy decoding unit 210 can also perform entropy decoding by realizing the same VLC table used in the coding device. Furthermore, if CABAC is used to perform entropy coding in the video coding device, the entropy decoding unit 210 can perform entropy decoding using CABAC in response.

[0065] Of the information decoded in the entropy decoding unit 210, the information for generating prediction blocks is provided to the prediction unit 230, and the residual values ​​from the entropy decoding performed in the entropy decoding unit 210 can be input to the re-sorting unit 215.

[0066] The re-alignment unit 215 can re-align the bitstream that has been entropi-decoded in the entropy decoding unit 210 based on the method used for re-alignment in the video encoding device. The re-alignment unit 215 can re-restore coefficients expressed in one-dimensional vector form to two-dimensional block form coefficients and re-align them. The re-alignment unit 215 can receive information related to the coefficient scan performed in the encoding device and perform re-alignment by scanning in reverse order based on the scan order performed in the encoding device.

[0067] The inverse quantization unit 220 can perform inverse quantization based on the quantization parameters provided by the encoding device and the coefficient values ​​of the rearranged blocks.

[0068] The inverse transform unit 225 can perform inverse DCT and / or inverse DST on the DCT and DST performed by the transform unit of the encoding device with respect to the quantization results performed in the video encoding device. The inverse transform can be performed based on the transmission unit or video division unit determined by the encoding device. The DCT and / or DST in the transform unit of the encoding device can be performed selectively according to multiple pieces of information such as the prediction method, the size of the current block, and the prediction direction, and the inverse transform unit 225 of the decoding device can perform the inverse transform based on the transformation information performed in the transform unit of the encoding device.

[0069] The prediction unit 230 can generate predicted blocks based on prediction block generation-related information provided by the entropy decoding unit 210 and previously decoded block and / or picture information provided by the memory 240.

[0070] Currently, when the prediction mode for the PU is set to intra-prediction mode, intra-prediction can be performed to generate prediction blocks based on the pixel information in the current picture.

[0071] Currently, when the prediction mode for the PU is the inter-prediction mode, inter-prediction for the current PU can be performed based on information contained in at least one picture, either a previous picture or a subsequent picture of the current picture. In this case, motion information necessary for inter-prediction of the current PU, such as motion vectors and reference picture indexes, provided by the video encoding device, can be derived from skip flags, merge flags, etc., received from the encoding device.

[0072] The reconstruction block can be generated using the prediction block generated by the prediction unit 230 and the residual block provided by the inverse transformation unit 225. Figure 2 illustrates how the prediction block and the residual block are combined in the adder to generate the reconstruction block. In this case, the adder can be considered as a separate unit (reconstruction block generation unit) that generates the reconstruction block.

[0073] When skip mode is used, the residual is not transferred, and the predicted block can be used as a restored block.

[0074] The restored blocks and / or pictures can be provided to the filter unit 235. The filter unit 235 can employ block distortion removal filtering, SAO and / or ALF, etc., on the restored blocks and / or pictures.

[0075] Memory 240 can store the restored picture or block and make it available as a reference picture or reference block, and can also provide the restored picture to the output unit.

[0076] On the other hand, the encoded or decoded picture is stored in memory, for example, in a decoded picture buffer (DPB). When encoding or decoding the current picture, previous pictures stored in the DPB are referenced to make predictions about the current picture.

[0077] Specifically, the encoding and decoding devices can maintain a reference picture list containing a list of previously encoded / decoded pictures for use in inter prediction.

[0078] When inter prediction is employed, the encoding and decoding devices can make predictions for the target block (current block) of the current picture by referring to other pictures. Inter prediction may also be performed in the prediction unit within the encoding and decoding devices, as shown in Figures 1 and 2.

[0079] When performing interface prediction, as described above, the current block is predicted using information about adjacent and available surrounding blocks. In this case, surrounding blocks include available blocks (hereinafter referred to as "Col blocks" for convenience of explanation) that are co-located in the same position as the current block in the reference picture that the current block can refer to.

[0080] For the sake of explanation, surrounding blocks used to make predictions for the current block in interface prediction are called "candidate blocks."

[0081] Interpretation uses information from candidate blocks to make predictions for the current block. In skip mode or merge mode, motion information (e.g., motion vectors) and reference pictures for the selected block from the candidate blocks are used as motion information and reference pictures for the current block.

[0082] When using MVP, motion information (e.g., motion vector) for a selected block from the candidate blocks is used as the motion vector for the current block, and the reference picture information for the current block is transferred from the encoding device to the decoding device. The difference MVD between the MVP derived from the candidate blocks and the motion vector for the current block is transferred from the encoding device to the decoding device, and the prediction unit of the decoding device can derive motion information for the current block based on the MVP and MVD.

[0083] Figure 3 is a schematic diagram illustrating an example of a candidate block that can be used when performing interpretation prediction on the current block.

[0084] The prediction units of the encoding and decoding devices can utilize blocks at predetermined positions around the current block 400 as candidate blocks. For example, in the example shown in Figure 3, two blocks A0410 and A1420 located at the lower left end of the current block, and three blocks B0430, B1440, and B2450 at the upper right and upper left ends of the current block can be selected as candidate blocks. In addition to spatially adjacent blocks, the aforementioned Col block 460 can also be used as a candidate block in terms of time.

[0085] When performing interaction prediction, the motion information for the current block is either obtained by directly using the motion information of a selected block from the surrounding blocks, as described above, or by deriving it based on the motion information of a selected block from the surrounding blocks.

[0086] On the other hand, regarding the reference picture used for inter prediction, the reference picture for the current block can be derived from the reference picture of a surrounding block or can be specified by the decoder. In skip mode or merge mode, the prediction unit of the decoder can use the reference picture of a surrounding block as the reference picture for the current block. When using MVP, the prediction unit of the decoder can receive information from the encoding device that specifies the reference picture for the current block.

[0087] Pictures encoded / decoded prior to the current picture are stored in memory (e.g., DPB) and can be used to predict the current block (current picture). A list of pictures available for interpretation of the current block is maintained as a reference picture list.

[0088] A P-slice is a slice that is decoded via intra-prediction or inter-prediction using up to one motion vector and one reference picture. A B-slice is a slice that is decoded via intra-prediction or inter-prediction using up to two motion vectors and two reference pictures. In this case, the reference pictures include a short-term reference picture and a long-term reference picture.

[0089] Reference picturelist 0 (hereinafter referred to as "L0" for convenience of explanation) is a reference picturelist used for inter prediction of P-slice or B-slice. Reference picturelist 1 (hereinafter referred to as "L1" for convenience of explanation) is used for inter prediction of B-slice. Therefore, L0 is used for inter prediction of P-slice blocks that perform unidirectional prediction, and L0 and L1 are used for inter prediction of B-slice blocks that perform bidirectional prediction.

[0090] When the decoding device performs decoding of P-slice and B-slice via inter prediction, it constructs a reference picture list. The reference pictures used for inter prediction are specified via the reference picture list. The reference picture index is an index that points to a reference picture on the reference picture list.

[0091] A reference picture list can be created based on a set of reference pictures transferred from the encoding device.

[0092] A referenced picture that creates a referenced picture list via a referenced picture index can be stored in memory (e.g., DPB).

[0093] Pictures stored in memory (pictures previously encoded / decoded from the current picture) are managed by the encoding and decoding devices. The encoding and decoding devices maintain the pictures necessary for predicting the current block and release pictures not used for predicting the current block from memory.

[0094] While using a sliding window method to manage referenced pictures is a simple way to manage them—by storing them in memory and then releasing them after a certain period—it has several drawbacks. For example, even if there are referenced pictures that are no longer needed, they cannot be directly released from memory, reducing efficiency. Also, because they are released from memory after a certain period, it becomes difficult to manage long-term referenced pictures.

[0095] Considering the problems with the sliding window method, a Memory Management Command Operation (MMCO) method may be used, in which the encoding device directly notifies instructions regarding the management of the reference picture. However, even when using the MMCO method, picture loss may occur during the notification process, and if the lost picture contains an MMCO instruction, the lost MMCO information cannot be recovered, making it impossible to maintain the memory (DPB) in an accurate state where the currently required picture is managed. Therefore, inter-prediction may also be inaccurate.

[0096] To solve the aforementioned problem, a method can be used to transfer the list of reference pictures necessary for the slice decoding process from each slice header. A kind of abstract container in the slice header that contains the list of reference pictures can be called "RefPicList". Alternatively, as explained earlier, in order to distinguish between reference picture list 0 and reference picture list 1 configured in the decoding device, the list of reference pictures necessary for the slice decoding process can be called a reference picture set or a collection of reference pictures.

[0097] A reference picture set or RefPicList (hereinafter referred to as "reference picture set" for convenience of explanation and to distinguish it from a reference picture list) contains reference pictures used for referencing current pictures / slices or future pictures / slices. For example, a reference picture set is information transferred from an encoding device to a decoding device, and the pictures included in the reference picture set can be identified by a POC. The POC indicates the display order of the pictures. In this case, the POC of a reference picture included in the reference picture set may be a relative POC with respect to the POC of the current picture.

[0098] Relative POC indicates the POC difference between two pictures in the reference picture set. In POC order, the relative POC of a reference picture prior to the current picture (a reference picture with a smaller POC than the current picture) is the POC difference with the immediately preceding reference picture in the reference picture set. In POC order, the relative POC of a reference picture after the current picture (a reference picture with a larger POC than the current picture) is also the POC difference with the immediately preceding reference picture in the reference picture set. However, in the case of a reference picture whose relative POC has a different sign from (1) the first reference picture in the reference picture set and (2) a previous reference picture in the reference picture set, the magnitude of the relative POC is the POC difference with the current picture.

[0099] The point of reference (POC) difference between two pictures in the reference picture set can be expressed using its absolute value and sign.

[0100] The reference picture set is communicated from the encoding device to the decoding device for each P-slice and B-slice.

[0101] The reference picture lists L0 and L1 may be created based on the reference picture sets received from the encoding device, or they may be explicitly transferred from the encoding device.

[0102] When creating a reference picture list L0, the reference picture index is assigned to the pictures with a smaller POC than the current picture from the received set of reference pictures (pictures in POC order that are before the current picture or have a negative sign for their relative POC) and pictures with a larger POC than the current picture (pictures in POC order that are after the current picture or have a positive sign for their relative POC).

[0103] For example, until all reference picture indices are assigned to create a reference picture list, (i) in the set of reference pictures for the current picture / slice, pictures with a POC smaller than the current picture's POC are assigned lower reference picture in POC order, and then (ii) in the set of reference pictures for the current picture / slice, pictures with a POC larger than the current picture's POC are assigned lower reference picture in POC order, and then pictures with a POC larger than the current picture are assigned lower reference picture in POC order.

[0104] When creating the reference picture list L1, the reference picture index is assigned to the pictures with a larger POC than the current picture from the received set of reference pictures. These include pictures with a smaller POC than the current picture (pictures in POC order that are before the current picture or have a negative sign for their relative POC) and pictures with a larger POC than the current picture (pictures in POC order that are after the current picture or have a positive sign for their relative POC).

[0105] For example, until all reference pictures that make up a reference picture list are assigned, (i) in the set of reference pictures for the current picture / slice, pictures with a POC greater than the current picture's POC are assigned a lower reference picture index in order of POC, and then (ii) in the set of reference pictures for the current picture / slice, pictures with a POC smaller than the current picture's POC are assigned a lower reference picture index in order of POC, and then pictures with a POC smaller than the current picture are assigned a lower reference picture index in order of POC, and so on.

[0106] Here, we have used the case of a short-term reference picture as an example, but in the case of a reference picture list that includes a long-term reference picture, after going through processes (i) and (ii) with respect to L0 and L1, a picture that has been transferred as a long-term reference picture via the reference picture set can be added.

[0107] In this specification, we will describe how to create a set of reference pictures for short-term reference pictures and how to create a list of reference pictures. Hereinafter, "reference picture" means a short-term reference picture.

[0108] In this case, in order to reduce the number of bits in the notified set of reference pictures and to reduce the complexity of the process of creating the list of reference pictures in the decoder, the reference pictures (information about the reference pictures, e.g., POC values) within the set of reference pictures (list of reference pictures) can be transferred in an aligned manner.

[0109] Reference pictures within a set of reference pictures are notified in the following order: (1) Reference pictures with a POC smaller than the current picture's POC are placed at the beginning of the set of reference pictures in descending order of POC, and then (2) Reference pictures with a POC larger than the current picture's POC are placed in ascending order of POC.

[0110] For example, within a set of reference pictures, a reference picture (picture information) with a POC smaller than the current picture's POC is placed first, followed by a reference picture (picture information) with a POC larger than the current picture's POC. In this case, the information of the placed reference picture may be the POC of the reference picture, the relative POC of the reference picture, or the size and sign of the relative POC of the reference picture.

[0111] If the information to be placed is the POC of the reference picture, then in the set of reference pictures, the POCs of reference pictures with smaller POCs than the current picture's POC are placed in order of proximity to the current picture's POC, and then the POCs of reference pictures with larger POCs than the current picture's POC are placed in order of proximity to the current picture's POC.

[0112] When the information to be placed is the relative POC of the reference picture, the relative POCs of reference pictures with a smaller POC than the current picture are placed first, followed by the relative POCs of reference pictures with a larger POC than the current picture. For example, within the set of reference pictures, the relative POCs of reference pictures with a smaller POC than the current picture are placed in descending order of the reference picture's POC, and then the relative POCs of reference pictures with a larger POC than the current picture are placed in ascending order of the reference picture's POC. Here, the relative POC for a reference picture is the difference between the POC of the current picture and the POC of the reference picture.

[0113] Within the reference picture set, a relative POC can be represented by its magnitude (absolute value) and its sign.

[0114] The sign of the relative POC of a reference picture indicates whether the reference picture is before or after the current picture in the POC order. Therefore, instead of transferring information that specifically indicates the sign of the relative POC (whether it is + or -), the magnitude of the relative POC for reference pictures before the current picture in the POC order can be transferred first, and then the magnitude of the relative POC for reference pictures after the current picture can be transferred. A decoder receiving a set of reference pictures can determine that the magnitude of the relative POC received first is for reference pictures before the current picture in the POC order, and the magnitude of the relative POC received later is for reference pictures after the current picture in the POC order. At this time, information representing both the number of reference pictures before the current picture and the number of reference pictures after the current picture in the POC order can be transferred.

[0115] When transferring relative POC sizes, within the set of reference pictures, the relative POC sizes for reference pictures with a smaller POC than the current picture's POC are arranged in descending order of the reference picture's POC, and then the relative POC sizes for reference pictures with a larger POC than the current picture are arranged in ascending order of the reference picture's POC.

[0116] Table 1 shows an example of a method for determining the relative POC size and code in an encoding device. [Table 1]

[0117] The encoding device can determine the relative POC size and code of the reference picture to be notified via the reference picture set using the method shown in Table 1.

[0118] sign_ref_pic[i] identifies the sign of the relative POC for the i-th reference picture in the reference picture set. If the POC of the i-th reference picture (ref_pic[i]) is greater than the POC of the current picture (currentPOC), the value of sign_ref_pic[i] indicates '+'. If the POC of the i-th reference picture (ref_pic[i]) is not greater than the POC of the current picture (currentPOC), the value of sign_ref_pic[i] indicates '-'.

[0119] abs_ref_pic[i] identifies the relative POC size for the i-th reference picture in the reference picture set. When the relative POC sign of the i-th reference picture is the same as the relative POC sign of the (i-1)th reference picture, the difference between the difference between the POC of the i-th reference picture and the reference value (refValue) and the difference between the POC of the (i-1)th reference picture and the reference value (refValue) becomes the relative POC size of the i-th reference picture. That is, when the relative POC sign of the i-th reference picture is the same as the relative POC sign of the (i-1)th reference picture, the relative POC size of the i-th reference picture is the POC difference between adjacent reference pictures in the reference picture set (the difference between the POC of the i-th reference picture and the POC of the (i-1)th reference picture).

[0120] The reference value (refValue) is either transmitted from the encoding device or is a pre-set reference POC value, which becomes the reference POC value for calculating the first relative POC in the reference picture set. For example, the reference value (refValue) may be the POC value of the current picture.

[0121] The relative POC code of the i-th reference picture does not have to be the same as the relative POC code of the (i-1)th reference picture. In this case, the i-th reference picture is either the first picture in the set of reference pictures, or the (i-1)th reference picture is a picture prior to the current picture in POC order, and the i-th reference picture is a picture after the current picture in POC order. Therefore, if the relative POC code of the i-th reference picture is not the same as the relative POC code of the (i-1)th reference picture, the magnitude of the relative POC of the i-th reference picture is the difference between the POC of the i-th reference picture and the reference value (refValue). Next, in the case of the (i+1)th reference picture, the relative POC codes of the i-th reference picture and the i-th reference picture are again the same, so the magnitude of the relative POC of the (i+1)th reference picture is again the difference between the POC of the (i+1)th reference picture and the POC of the i-th reference picture.

[0122] The encoding device can transfer the relative POC size and code of the reference pictures derived as described above as a set of reference pictures. Furthermore, when transferring the relative POC size of a reference picture, the encoding device can first transfer the relative POC size for reference pictures prior to the current picture, in the order in which the codes are transferred, and then transfer the relative POC size for reference pictures after the current picture. In this case, the encoding device can transfer information indicating the number of reference pictures whose relative POC code is '-' (pictures prior to the current picture in POC order) and the number of reference pictures whose relative POC code is '+' (pictures after the current picture in POC order).

[0123] Table 2 shows other examples of methods for determining the relative POC size and code in an encoding device. [Table 2]

[0124] Table 2 illustrates the case where the reference value (refValue) is the current picture's POC value.

[0125] Similar to Table 1, in Table 2, if the POC (ref_pic[i]) of the i-th reference picture is greater than the POC (currentPOC) of the current picture, the value of sign_ref_pic[i], which is the sign of the relative POC for the i-th reference picture, indicates '+'. If the POC (ref_pic[i]) of the i-th reference picture is not greater than the POC (currentPOC) of the current picture, the value of sign_ref_pic[i], which is the sign of the relative POC for the i-th reference picture, indicates '-'.

[0126] When the relative POC code of the i-th reference picture is the same as the relative POC code of the (i-1)th reference picture, the difference between the POC of the i-th reference picture and the POC of the current picture, and the difference between the POC of the (i-1)th reference picture and the POC of the current picture, becomes the magnitude of the relative POC of the first reference picture. In other words, when the relative POC code of the i-th reference picture is the same as the relative POC code of the (i-1)th reference picture, the magnitude of the relative POC of the i-th reference picture is the POC difference between adjacent reference pictures in the reference picture set (the difference between the POC of the i-th reference picture and the POC of the (i-1)th reference picture).

[0127] The relative POC code of the i-th reference picture does not have to be the same as the relative POC code of the (i-1)th reference picture. In this case, the i-th reference picture is either the first picture in the set of reference pictures, or the (i-1)th reference picture is a picture prior to the current picture in POC order, and the i-th reference picture is a picture after the current picture in POC order. Therefore, if the relative POC code of the i-th reference picture is not the same as the relative POC code of the (i-1)th reference picture, the size of the relative POC of the i-th reference picture is the difference between the POC of the i-th reference picture and the POC of the current picture. Next, in the case of the (i+1)th reference picture, the code is again the same as the relative POC of the i-th reference picture, so the size of the relative POC of the (i+1)th reference picture is again the difference between the POC of the (i+1)th reference picture and the POC of the i-th reference picture.

[0128] The encoding device can transfer the relative POC size and code of the reference pictures derived as described above as a set of reference pictures. Furthermore, when transferring the relative POC size of a reference picture, the encoding device can first transfer the relative POC size for reference pictures prior to the current picture, in the order in which the codes are transferred, and then transfer the relative POC size for reference pictures after the current picture. In this case, the encoding device can transfer information indicating the number of reference pictures with a relative POC code of '-' (pictures prior to the current picture in POC order) and the number of reference pictures with a relative POC code of '+' (pictures after the current picture in POC order).

[0129] The decoding device receives information about the reference picture set from the encoding device and can construct or restore the reference picture set based on this information.

[0130] Table 3 shows an example of a method for recovering the reference picture information (POC) in a decoding device that receives a set of reference pictures. [Table 3]

[0131] The decoding device can recover the POC that can be used to predict the current block (picture) based on the reference picture information (size of the relative POC or the size and code of the relative POC) received from the encoding device, using the method shown in Table 3.

[0132] The point of reference (POC) (ref_pic[i]) of the i-th reference picture in the set of reference pictures can be reconstructed based on the relative POC size (abs_ref_pic[i]) and sign (sign_ref_pic[i]) of the i-th reference picture.

[0133] As shown in Table 3, the decoding device can explicitly receive the code of the relative POC from the encoding device and recover the POC of the reference picture.

[0134] If the sign of the i-th reference picture in the reference picture set is the same as the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '-', then the POC of the i-th reference picture is the value obtained by subtracting the relative POC sum from the first reference picture (the 0th reference picture) to the i-th reference picture from the reference value (refValue). If the sign of the i-th reference picture in the reference picture set is the same as the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '+', then the POC of the i-th reference picture is the value obtained by adding the relative POC from the first reference picture (the 0th reference picture) to the i-th reference picture to the reference value (refValue).

[0135] In this case, the reference value (refValue) is either a reference POC value transferred from the encoding device or a pre-set reference POC value, which becomes the reference POC value for calculating the first relative POC in the reference picture set. For example, the reference value (refValue) may be the POC value of the current picture.

[0136] When the sign of the i-th reference picture in the reference picture set is different from the sign of the (i-1)th reference picture, either the i-th reference picture in the i-th reference picture set is the first picture, or the (i-1)th reference picture in the reference picture set is a picture prior to the current picture in POC order, and the i-th reference picture is a picture after the current picture in POC order.

[0137] If the sign of the i-th reference picture in the reference picture set is different from the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '-', then the POC of the i-th reference picture is the reference value (refValue) minus the relative POC of the i-th reference picture. If the sign of the i-th reference picture in the reference picture set is different from the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '+', then the POC of the i-th reference picture is the reference value (refValue) plus the relative POC of the i-th reference picture.

[0138] Furthermore, unlike the example in Table 2, information explicitly indicating the code for the relative POC of the reference picture does not need to be transmitted. In this case, the decoding device can determine that the code of relative POCs located at the beginning of the reference picture set is '-' (minus), and the code of relative POCs located at the end of the reference picture set is '+' (plus). At this time, the encoding device can also transmit information indicating the number of relative POCs with the code '-' and the number of relative POCs with the code '+'. The decoding device can determine that the number of relative POCs with the code '-' indicated by the encoding device are '-', starting from the beginning of the reference picture set, and that the remaining relative POCs are '+', thereby recovering the POC of the i-th reference signal (ref_pic[i]) as described above.

[0139] In other words, the relative POC for the first reference picture in the reference picture set is the POC difference from the reference value (refValue). The relative POC for the reference pictures in the reference picture set that are prior to the current picture, excluding the first reference picture, is the POC difference from the previous reference picture. The relative POC for the first reference picture after the current picture in the reference picture set is the POC difference from the reference value. The relative POC for the remaining reference pictures in the reference picture set (from the second reference picture after the current picture to the last reference picture in the reference picture set) is the POC difference from the previous reference picture. Here, "before" and "after" the current picture are determined by the POC order. Also, the previous reference picture means the picture immediately preceding the current picture in the sorting order within the reference picture set.

[0140] Table 4 shows another example of how a decoding device receiving a set of reference pictures can recover the information of the reference picture (POC). [Table 4]

[0141] The method in Table 4 is explained using the example where the number of reference pictures in the reference picture set is 2, and the reference value (refValue) for calculating the first relative POC value in Table 3 is the POC of the current picture, in order to clearly explain the features of the present invention.

[0142] The decoding device can recover the POC of the reference picture by explicitly receiving the code of the relative POC from the encoding device.

[0143] If the sign of the i-th reference picture in the reference picture set is the same as the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '-', then the POC of the i-th reference picture is the current picture's POC minus the relative POC of the i-th reference picture and the relative POC of the (i-1)th reference picture. If the sign of the i-th reference picture in the reference picture set is the same as the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '+', then the POC of the i-th reference picture is the current picture's POC plus the relative POC of the i-th reference picture and the relative POC of the (i-1)th reference picture.

[0144] When the sign of the i-th reference picture in the reference picture set is different from the sign of the (i-1)th reference picture, either the i-th reference picture in the i-th reference picture set is the first picture, or the (i-1)th reference picture in the reference picture set is a picture prior to the current picture in POC order, and the i-th reference picture is a picture after the current picture in POC order. In this case, the relative POC of the i-th reference picture is derived based on the POC of the current picture, as can be seen from Table 2.

[0145] Therefore, if the sign of the i-th reference picture in the reference picture set is different from the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '-', then the POC of the i-th reference picture will be the current picture's POC minus the relative POC of the i-th reference picture. If the sign of the i-th reference picture in the reference picture set is different from the sign of the (i-1)th reference picture, and the sign of the i-th reference picture is '+', then the POC of the i-th reference picture will be the current picture's POC plus the relative POC of the i-th reference picture.

[0146] Furthermore, as explained in Table 3, information explicitly indicating the code for the relative POC of the reference picture does not need to be transmitted. In this case, the decoding device can determine that the code of relative POCs located at the beginning of the reference picture set is '-' (minus), and the code of relative POCs located at the end of the reference picture set is '+' (plus). At this time, the encoding device can also transmit information indicating the number of relative POCs with the code '-' and the number of relative POCs with the code '+'. The decoding device can determine that the code of relative POCs corresponding to the number of relative POCs with the code '-' indicated by the encoding device is '-', starting from the beginning of the reference picture set, and that the code of the remaining relative POCs is '+', thereby reconstructing the POC of the i-th reference signal (ref_pic[i]) as described above.

[0147] In other words, the relative POC for the first reference picture in the reference picture set is the POC difference with the current picture. The relative POC for the reference pictures in the reference picture set that are prior to the current picture, excluding the first reference picture, is the POC difference with the immediately preceding reference picture. The relative POC for the first reference picture after the current picture in the reference picture set is the POC difference with the current picture. The relative POC for the remaining reference pictures in the reference picture set (from the second reference picture after the current picture to the last reference picture in the reference picture set) is the POC difference with the immediately preceding reference picture. Here, "before" and "after" the current picture are determined by the POC order. Here, the immediately preceding reference picture means the picture immediately preceding in the sorting order within the reference picture set.

[0148] The following describes a specific example in which the present invention is employed, assuming that the reference value (refValue) is the current picture's POC.

[0149] Figure 4 schematically illustrates an example of a set of reference pictures notified from the encoding device to the decoding device. The example in Figure 4 shows the case where nine P-slices (P-pictures) P0 to P9, which perform unidirectional prediction, cross-reference each other.

[0150] Table 5 shows an example where the set of reference pictures notified for the example in Figure 4 consists of the POCs of the reference pictures.

[0151] [Table 5]

[0152] As shown in Figure 4 and Table 5, the set of reference pictures for the current picture includes the points of reference (POCs) of the reference pictures that the current picture can reference. In the set of reference pictures, the reference pictures are ordered by POC, with lower indices assigned to reference pictures closer to the current picture.

[0153] For example, in the examples in Figure 4 and Table 5, if the current picture is P6 (POC=26), the reference pictures that the current picture can access are P5, P4, and P0. Therefore, the set of reference pictures for the current picture P6 with a POC of 26 consists of the POCs P5, P4, and P0, and reference pictures closer to the current picture are assigned lower indices in order of POC.

[0154] In the examples in Figure 4 and Table 5, the reference pictures are sorted in descending order within the reference picture set in order of POC (Point of Access), for reference pictures prior to the current picture, but the reference picture set simply notifies the POC of the reference picture.

[0155] In contrast, as explained earlier, a set of reference pictures can notify the relative POC of the reference picture.

[0156] Table 6 shows an example of the reference picture set notified in the case of Figure 4, and is intended to illustrate an example where the reference picture set is composed of relative POCs of the reference pictures. [Table 6]

[0157] Table 6 shows the reference picture set of the current picture relative to Figure 4, represented by the reference picture's POC, the size of the reference picture's relative POC, and the reference picture's relative POC code.

[0158] Table 5 illustrates the case where the POC of a reference picture is transferred directly within a reference picture set, while the example in Table 6 illustrates the case where the relative POC of a reference picture is transferred within a reference picture set.

[0159] In POC order, the relative POC of a reference picture prior to the current picture (a reference picture with a smaller POC than the current picture) is the POC difference between that reference picture and the immediately preceding reference picture within the reference picture set. Furthermore, the relative POC code included in the reference picture set and transferred indicates whether the reference picture in question is prior to or after the current picture in POC order.

[0160] For example, in the example in Figure 4 and Table 6, if we consider the case where the current picture is P5, the pictures that the current picture can reference are P4, P3, and P0, with POCs of 24, 23, and 20, respectively.

[0161] When a reference picture set transfers relative POCs to P5, the reference picture set transfers the relative POCs of the reference pictures to P5 in a predetermined order, sorting by size and sign. As described above, in the example of Figure 4 representing the reference relationships between P slices, the reference pictures are in POC order, prior to the current picture, and in the reference picture set, the reference pictures are sorted in descending order.

[0162] Therefore, in the reference picture set for P5, the relative POC sizes are sorted in the order of P4, P3, and P0. As shown in Table 6, the relative POC size of P4 transferred to the reference picture set for P5 is 1 and its sign is '-', the relative POC size of P3 is 1 and its sign is '-', and the relative POC size of P0 is 3 and its sign is '-'.

[0163] In this case, it is also possible to transfer a set of reference pictures in POC order, with the reference pictures prior to the current picture (the size of the relative POC of the reference picture) aligned to the front of the reference picture set, and the reference pictures after the current picture (the size of the relative POC of the reference picture) aligned to the back of the reference picture set, without a relative POC sign. In this case, it is also possible to transfer information indicating both the number of reference pictures prior to the current picture (reference pictures with a relative POC sign of '-') and the number of reference pictures after the current picture (reference pictures with a relative POC sign of '+') in POC order.

[0164] Unlike Figure 4, which shows the reference relationships between P-pictures performing unidirectional prediction, Figure 5 shows an example of reference relationships between B-pictures performing bidirectional prediction. Figure 5 schematically shows the reference relationships between nine B-pictures B0 to B8.

[0165] Table 7 shows an example of the reference picture set notified in the case of Figure 5, and is intended to illustrate an example in which the reference picture set is composed of relative POCs of the reference pictures. [Table 7]

[0166] In the examples in Table 7 and Figure 5, the reference picture set can transfer the relative POC of the reference picture instead of transferring the POC of the reference picture directly.

[0167] In POC order, the relative POC of a reference picture prior to the current picture (a reference picture with a smaller POC than the current picture) is the POC difference with the immediately preceding reference picture within the reference picture set. In POC order, the relative POC of a reference picture after the current picture (a reference picture with a larger POC than the current picture) is also the POC difference with the immediately preceding reference picture within the reference picture set. However, (1) in the case of the first reference picture in the reference picture set, and (2) in the case of a reference picture whose relative POC sign is different from that of a previous reference picture in the reference picture set, the magnitude of the relative POC is the POC difference with the current picture. In other words, in the reference picture set in POC order, the relative POC of the reference picture closest to the current picture among the reference pictures prior to the current picture, and the relative POC of the reference picture closest to the current picture among the reference pictures after the current picture, is the POC difference with the current picture.

[0168] Referring to Table 7, and taking the case where the current picture is B5 as an example, the reference picture set consists of B4, B2, B6, and B8. When the reference picture set is constructed from relative POCs, the size of the relative POC to which the lowest index in the reference picture set is assigned is for B4, and is 1, which is the difference between the current picture's POC and B4's POC, and its sign is '-'. Next, the size of the relative POC to which the second index is assigned is for B2, and is 2, which is the difference between B4's POC and B2's POC, and its sign is '-'. The size of the relative POC to which the third index is assigned is for B6. Since the relative POC for B6 has a different sign from the relative POC for the previous reference picture, B2, the size of the relative POC for B6 is 1, which is the POC difference with the current picture, and its sign is '+'. The magnitude of the relative POC to which the last index is assigned is 2, which is the difference between the POC of B6 and B8, relative to B8, and its sign is '+'.

[0169] As described above, instead of transferring the size and sign of all the relative POCs of the reference pictures to the current picture, the reference picture set can transfer only the size of the relative POCs of the reference pictures to the current picture, by transferring the size of relative POCs with a '-' sign before the size of relative POCs with a '+' sign, so that the sign of the corresponding relative POC can be derived without explicitly transferring the sign. In this case, information indicating the number of relative POCs with a '-' sign and the number of relative POCs with a '+' sign can be transferred together.

[0170] For example, considering again the case where the current picture is B5 in Table 7, the encoding device can construct and transfer a reference picture set for B5 using only the relative PC size of the reference picture, as shown in (1 2 1 2). As shown in Table 7, the size of relative POCs with a '-' sign is placed at the beginning of the reference picture set. The sorting order is descending for relative POCs with a '-' sign (POC order, relative to reference pictures before the current picture), as explained earlier, and also maintains ascending order for relative POCs with a '+' sign (POC order, relative to reference pictures after the current picture), etc., as explained earlier. At this time, along with the reference picture set, information indicating the number of relative POCs with '-' and the number of relative POCs with '+' signs can be transferred. For example, if the decoding device receives an instruction that in the reference picture set for B5, there are 2 reference pictures (relative POCs) with the '-' sign and 2 reference pictures (relative POCs) with the '+' sign, the decoding device can determine that the signs for the first two relative POCs in the reference picture set are '-' and the signs for the last two relative POCs are '+'. Therefore, the size of the first two relative POCs in the reference picture set is the size of the relative POC for reference pictures with a smaller POC than the current picture, and the size of the last two relative POCs in the reference picture set is the size of the relative POC for reference pictures with a larger POC than the current picture.

[0171] Figure 6 is a schematic diagram illustrating an example of a reference relationship between picture B and picture P.

[0172] Figure 6 shows the reference relationship between seven P-pictures P0-P6 that perform unidirectional prediction and two B-pictures B0 and B1 that perform bidirectional prediction.

[0173] Table 8 shows an example of the reference picture set notified in the case of Figure 6, and is intended to illustrate an example in which the reference picture set is composed of relative POCs of the reference pictures.

[0174] [Table 8]

[0175] Table 8 and Figure 6 show examples for cases where P-pictures and B-pictures are mixed, but the method for deriving the size and sign of relative POCs, and the method for aligning relative POCs within the reference picture set, are as described above.

[0176] For example, if the current picture is B1, the reference picture set for B1 can consist of relative POCs (Points of Control) of P3, P0, and P6. The reference picture set consists of the size (2 4 2) of the relative POCs for P3, P0, and P6, and their respective codes, and can be transferred to the decoding device.

[0177] In this case as well, instead of transferring information indicating the sign of the relative POCs, it is also possible to transfer information indicating the number of relative POCs with a '-' sign and the number of relative POCs with a '+' sign based on the sorting order, along with the reference picture set containing the size of the relative POCs. For example, if the current picture is B1, it is possible to transfer a reference picture set (2 4 2) containing the size of the relative POCs and information indicating that there are 2 relative POCs with a '-' sign and 1 relative POC with a '+' sign.

[0178] Figure 7 is a flowchart illustrating the encoding method performed by the encoding device according to the present invention. The encoding device that performs the encoding method in Figure 7 corresponds to the encoding device described in Figure 1.

[0179] As shown in Figure 7, the encoding device performs a prediction for the current block (S710). The encoding device can perform either interpretation or intrapretation for the current block. When performing interpretation, the reference picture for the current block can be selected / specified using the reference picture list configured as described above.

[0180] The encoding device transforms / quantizes the prediction result for the current block (S720). The encoding device can transform / quantize the residual block corresponding to the difference between the prediction result and the original block. Furthermore, if intra-prediction is adopted, it can transform / quantize information related to the adopted intra-prediction mode, and if inter-prediction is adopted, it can transform / quantize motion information (information related to motion vectors / reference pictures).

[0181] The encoding device entropy encodes the transformed / quantized information (S730). CABAC can be used as the method for entropy encoding.

[0182] The encoding device notifies the entropy-encoded information (S740). The information notified at this time includes a set of reference pictures for creating a reference picture list for the current picture (current block). The set of reference pictures can be configured per slice and can be included in the header of the slice during transmission.

[0183] A set of reference pictures can be constructed from the points of content (POCs) of the reference pictures relative to the current block. Alternatively, the set of reference pictures can be constructed from the relative POCs of the reference pictures to reduce transfer overhead.

[0184] If the reference picture set consists of relative POCs of the reference pictures, the size and sign of the relative POCs for the pictures that can currently be used as reference pictures of the pictures may be transferred via the reference picture set, or the size and number of relative POCs with a '-' sign and relative POCs with a '+' sign may be transferred. When relative POCs are transferred, relative POCs with a '-' sign are transferred first, followed by relative POCs with a '+' sign. Relative POCs with a '-' sign can be sorted in descending order according to the reference picture's POC, and relative POCs with a '+' sign can be sorted in ascending order according to the reference picture's POC.

[0185] Figure 7 provides a schematic explanation of the operation of the encoding device, taking into account the contents of the reference picture set, in order to facilitate understanding of the invention. However, this is for the sake of explanation only, and in the present invention, the operation of the encoding device includes the operations described in Figure 1.

[0186] Figure 8 is a flowchart illustrating the decoding method performed by the decoding device according to the present invention.

[0187] As shown in Figure 8, the decoder receives the bitstream from the encoder and performs entropy decoding (S810). The bitstream received from the encoder includes a set of reference pictures. The set of reference pictures can be received in the slide header.

[0188] The set of reference pictures may consist of the points of content (POCs) of the reference pictures relative to the current block, or it may consist of the relative points of content (POCs) of the reference pictures.

[0189] The decoding device can receive information indicating which pictures can be used as reference pictures for the current picture via the reference picture set. For example, it can receive the POC of a picture that can be used as a reference picture for the current picture via the reference picture set. If the reference picture set includes (1) the size and sign of the relative POC for the reference picture, or (2) the size of the relative POC for the reference picture and the number of relative POCs with '-' and '+' signs, the POC of the relevant reference picture can be derived using the method in Table 4 or the like based on the received information.

[0190] When receiving relative POCs via a set of reference pictures, relative POCs with a '-' sign are received first, followed by relative POCs with a '+' sign. Relative POCs with a '-' sign can be sorted in descending order according to the POCs in the reference pictures, and relative POCs with a '+' sign can be sorted in ascending order according to the POCs in the reference pictures.

[0191] The decoding device makes a prediction for the current block based on the entropy-decoded information (S920). The prediction method for the current block can be transmitted from the encoding device. If the prediction method for the current block is an inter prediction, the decoding device can make the prediction using a reference picture list constructed based on the received set of reference pictures.

[0192] The method for creating a reference picture list from a set of reference pictures is as described above. The constructed reference picture list can be stored in the memory of the decoder.

[0193] The decoding device restores the video (S930). Based on the prediction for the current block, the current block is restored, and the video can be restored through the restored block. If skip mode is adopted, the residual is not transferred, so the predicted block can be used as the restored block. If merge mode is applied or MVP is used, the current block can be restored by combining the predicted block and the residual block.

[0194] In this specification, the expressions "picture included in the reference picture set" and "the x-th picture in the reference picture set" are used, but it should be noted that these are for the sake of clarity. A picture included in the reference picture set means a picture that contains the POC information corresponding to the reference picture set, and the x-th picture in the reference picture set means a picture in which the corresponding POC information is aligned in the x-th position within the reference picture set.

[0195] On the other hand, the reference relationships between pictures shown in Figures 4 to 6 were explained as an example where the temporal level is not adopted, but this is for the purpose of understanding the invention, and the present invention is not limited to this. The present invention can also be applied in the case where the temporal level is considered and only pictures of a lower level than itself are referenced, and in this case the reference relationships in Tables 5 to 8 can be modified to reflect this.

[0196] In the exemplary systems described above, the methods of the present invention are illustrated in a flowchart in a series of steps or blocks; however, the present invention is not limited to the order of the steps, and some steps may occur in a different order or simultaneously with other steps different from those described above. Furthermore, the embodiments described above include examples of various modes. Accordingly, the present invention should be understood to include all other substitutions, modifications, and changes that fall within the scope of the following claims.

[0197] In the above description of the present invention, when it is stated that one component is "connected" or "linked" to another component, it should be understood that the other component is directly connected to or linked to the other component, but that other components may exist between the two components. Conversely, when it is stated that one component is "directly connected" or "directly linked" to another component, it should be understood that no other components exist between the two components.

Claims

1. Interpretation method performed by a decoding device, Steps include obtaining the POC (picture order count) information within the slice header, The steps include: deriving a POC value for a reference picture based on the POC information; The steps include: configuring a list of reference pictures based on the POC value of the reference picture; The steps include: performing interpretation on the current block based on the aforementioned reference picture list to derive a prediction sample for the current block; The POC information in the slice header includes POC difference information, The POC value of the i-th reference picture is derived based on the POC difference derived from the POC difference information. In POC order, the aforementioned reference pictures after the current picture are sorted in ascending order. For the reference picture following the current picture in the POC order, If i is equal to 0, the POC difference represents the difference between the POC value of the current picture and the POC value of the i-th reference picture, and, If i is greater than 0, the POC difference represents the difference between the POC value of the i-th reference picture and the POC value of the (i-1)-th reference picture. A method in which reference pictures having a POC value greater than the POC value of the current picture are sorted in ascending order of their POC values ​​relative to the reference pictures.

2. A video encoding method performed by an encoding device, The steps include: deriving the POC (picture order count) value of the reference picture, A step of deriving POC information for the reference picture based on the derived POC value of the reference picture, wherein the POC information represents the POC difference for the reference picture, The step includes encoding video information including the POC information, The aforementioned POC information includes POC difference information, The aforementioned POC information is included in the slice header, The POC difference information identifies the POC difference with respect to the POC value of the i-th reference picture. In POC order, the aforementioned reference pictures after the current picture are sorted in ascending order. For the reference picture following the current picture in the POC order, If i is equal to 0, the POC difference represents the difference between the POC value of the current picture and the POC value of the i-th reference picture, and, If i is greater than 0, the POC difference represents the difference between the POC value of the i-th reference picture and the POC value of the (i-1)-th reference picture. A method in which reference pictures having a POC value greater than the POC value of the current picture are sorted in ascending order of their POC values ​​relative to the reference pictures.

3. A method for transmitting data related to video signals, A step of acquiring a bitstream relating to the video signal, wherein the bitstream is The steps include: deriving the POC (picture order count) value of the reference picture, A step of deriving POC information for the reference picture based on the derived POC value of the reference picture, wherein the POC information represents the POC difference for the reference picture, A step of encoding video information including the POC information, and a step of generating based on the above, The step of transmitting the data, which includes the bitstream, The aforementioned POC information includes POC difference information, The aforementioned POC information is included in the slice header, The POC difference information identifies the POC difference with respect to the POC value of the i-th reference picture. In POC order, the aforementioned reference pictures after the current picture are sorted in ascending order. For the reference picture following the current picture in the POC order, If i is equal to 0, the POC difference represents the difference between the POC value of the current picture and the POC value of the i-th reference picture, and, If i is greater than 0, the POC difference represents the difference between the POC value of the i-th reference picture and the POC value of the (i-1)-th reference picture. A method in which reference pictures having a POC value greater than the POC value of the current picture are sorted in ascending order of their POC values ​​relative to the reference pictures.

Images