METHOD FOR ENCODING AND DECODING MULTI-VIEW VIDEO - Patent application

Abstract

The present invention relates to a method for encoding views simultaneously representing a 3D scene from different positions or different viewing angles, implemented by an encoding device, comprising, for a depth component of at least one view, dividing the depth component into at least one block (C2), obtaining depth information for the at least one block from texture data of at least one texture component of the view (C4), obtaining at least one depth estimation parameter from this information (C5), and, since the depth information of the at least one block is not encoded, encoding the at least one depth estimation parameter (C6).

Description

[Technical Field]

[0001] The present invention relates generally to the field of immersive video, particularly multi-view video. More particularly, the present invention relates to encoding and decoding multiple captured views to generate such immersive video and to synthesizing uncaptured intermediate viewpoints. The present invention may be particularly, but not exclusively, applied to video encoding and corresponding video decoding implemented in current AVC and HEVC video coders and their extensions (MVC, 3D-AVC, MV-HEVC, 3D-HEVC, etc.). [Background technology]

[0002] It is common to use 360° cameras to generate omnidirectional video, e.g., 360° video. Such a 360° camera consists of several 2D (two-dimensional) cameras mounted on a spherical platform. Each 2D camera captures a 3D (three-dimensional) scene at a specific angle from the same viewpoint, and the set of views captured by the cameras makes it possible to generate a video representing the 3D scene according to a 360° x 180° field of view from one viewpoint. It is also possible to use a single 360° camera to capture a 3D scene according to a 360° x 180° field of view. Naturally, such a field of view can be small, e.g., 270° x 135°.

[0003] Such 360° video then allows the user to see the scene as if they were at the center, providing a full 360° view of their surroundings, offering a new way of looking at the scene. Such video is typically played on a virtual reality headset, also known as an HMD, an abbreviation for "head-mounted device." However, the video can also be displayed on a 2D screen equipped with appropriate user interaction means. The number of 2D cameras used to capture the 360° scene varies depending on the platform used.

[0004] However, the above-mentioned 360° approach is limited because the viewer can only view the scene from a single viewpoint.

[0005] To allow the viewer to change the viewpoint, there are multi-view capture systems in which a 3D scene is captured by a set of 2D-type cameras, each capturing a particular angle of the scene. To enhance the sense of immersion in the scene, one or more missing views, i.e., views representing viewpoints not captured by the cameras, are synthesized from the existing views.

[0006] The computation of these missing or intermediate views is performed by so-called "view synthesis" algorithms. For example, the VSRS ("view synthesis reference") software can be used as a view synthesis algorithm.

[0007] Traditionally, such synthesis algorithms are based not only on the texture components of views captured by each camera at different times, but also on the depth components of these views, called "depth maps." The depth map represents the distance between each pixel of a view and the camera that captured this view. Thus, each camera captures a view of the 3D scene from its respective viewpoint in the form of a texture component, which is associated with a depth map of the 3D scene as seen from each of the camera's viewpoints. There are several means for constructing depth maps, including radar, laser, and computational methods that use pixels from the current view and neighboring views.

[0008] There are known techniques for estimating depth from a set of texture components. Thus, the DERS method described in the document "(Non-Patent Document 1)" uses at least two texture components to first estimate the disparity between the two views, i.e., the displacement of one pixel between the two views, measured in number of pixels. To do this, DERS successively applies a "block matching" step to the blocks surrounding the current pixel in the current view, in order to identify the block in another view that minimizes the error relative to the block in the current view. This search is performed horizontally, since the views are considered calibrated. The search is performed within a given disparity interval, i.e., the minimum disparity D min and maximum disparity D max Block matching is performed on all blocks between the pixel and the disparity d. There is a direct link between the disparity d and the depth Z of the scene, and the depth Z of the pixel at disparity d is Z=(b*f) / d where "b" is the distance between the optical centers of the cameras capturing the view, and "f" is the focal length of the cameras. Therefore, the minimum parallax D min is the maximum depth Z expected in the scene max corresponds to the maximum disparity D max is the minimum depth Z expected in the scene min In a specific example, the scene capture corresponds to a Z value of a predetermined value, e.g., 0.3 m. min and Z of, say, 5m max This is done by specifying a value D that determines the number of disparity hypotheses to be evaluated. min and D max is generated directly. For example, D min = 2 pixels and D max= 192 pixels, this means that the block matching technique evaluates the block matching costs of all disparity hypotheses D = 2, 3, 4, ..., 191, 192 pixels, i.e., 191 calculations are stored and compared with each other. Thus, the DERS algorithm determines which disparity to select from among the 191 possible ones as a function of the cost associated with each disparity, and one or more regularization parameters, including the parameter Sc (short for "smoothing coefficient"). This coefficient determines the regularity of the generated depth map. Thus, if this coefficient is low, the depth map will be more accurate but will run the risk of including noise, while if this coefficient is high, the depth map will be very regular with uniform depth zones but will run the risk of only poorly representing small local variations.

[0009] Other depth estimation methods, such as the IVDE (short for "Immersive Video Depth Estimation") method, are presented in "(Non-Patent Document 2)," which jointly determine depth maps for a series of views in parallel.

[0010] In conventional immersive video coders, texture components and their respective depth maps are coded and sent to a decoder, e.g., installed on a user's display device. At the decoder side, each view is decoded, including decoding the view's texture components and decoding the depth map associated with the view's texture components. A synthesis algorithm then constructs an intermediate view corresponding to the viewpoint requested by the user from one or more decoded depth maps and from one or more decoded texture components.

[0011] "(Non-Patent Document 3)" describes an immersive video coder in which the depth map is not coded. Only the texture components of the views are coded and transmitted to the decoder. At the decoder side, the texture components are decoded, and then a depth estimation method, e.g., DERS, is applied to the decoded texture components to generate estimated depth maps. A VSRS (short for "view synthesis reference software") synthesis algorithm then performs view synthesis using these estimated depth maps.

[0012] This technique reduces the encoding bit rate of the omnidirectional video by eliminating the need to encode and transmit a depth map, and also reduces the number of pixels decoded to obtain the synthesized view compared to the number used in conventional immersive video decoders.

[0013] However, the decoding method used in this technique is computationally intensive because it requires the implementation of a depth estimation step at the decoder. Furthermore, the estimated depth itself is not of very good quality because it is based on decoded texture components that are of lower quality than the original texture components. Therefore, view synthesis implemented with this technique results in being suboptimal both in terms of the quality of the image rendered to the user and in terms of consumption of computing resources. [Prior art documents] [Non-patent literature]

[0014] [Non-Patent Document 1] O.Stankiewicz, K.Wegner, M.Tanimoto and M.Domanski, “Enhanced Depth Estimation Reference Software(DERS) for Free-viewpoint Television”, ISO / IEC JTC1 / SC29 / WG11 Doc.MPEG M31518, Geneva, 2013 [Non-patent document 2] Dawid Mieloch,Adrian Dziembowski,Jakub Stankowski,Olgierd Stankiewicz,Marek Domanski,Gwangsoon Lee,Yun Young Jeong[MPEG-I Visual]Immersive video depth estimation,ISO / IEC JTC1 / SC29 / WG11 MPEG2020 m53407 [Non-patent document 3] Patrick Garus,Jung Joel,Thomas Maugey,Christine Guillemot.Bypassing Depth Maps Transmission For Immersive Video Coding.PCS 2019-Picture Coding Symposium,Nov 2019,Ningbo,China.pp.1-5.hal-02397800 Summary of the Invention [Problem to be solved by the invention]

[0015] One of the objects of the present invention is to remedy the above-mentioned drawbacks of the prior art. [Means for solving the problem]

[0016] To this end, one object of the invention is a method for encoding views simultaneously representing a 3D scene from different positions or different viewing angles, implemented by a coding device, comprising the steps of: - dividing the depth component into at least one block; - deriving depth information for the at least one block from texture data of at least one texture component of the view; - deriving at least one depth estimation parameter from the depth information; - encoding the at least one depth estimation parameter, wherein depth information of the at least one block is not encoded. The present invention relates to a method comprising:

[0017] Such an encoding method according to the invention avoids, when encoding a view, encoding the depth blocks of the components or depth maps associated with this view, thereby reducing the computations implemented by the coder while allowing saving on memory resources, eliminating the need to store the encoded data of the depth blocks. Since this depth block is not encoded, the encoded data for this depth block is not transmitted to the decoder, thereby reducing the cost of signaling information transmitted between the coder and the decoder. To compensate for the lack of encoded data for the depth block and enable decoding of the depth block, the encoding method according to the invention implements the encoding of at least one depth estimation parameter associated with the depth block, whereby the depth estimation parameter is used by the decoder to reconstruct the depth block without the need to first decode this depth block.

[0018] According to a particular embodiment, the at least one depth estimation parameter is either a depth value of the at least one block that is greater than the respective depth value of the at least one block or a depth value of the at least one block that is less than the respective depth value of the at least one block.

[0019] According to this particular embodiment, the depth estimator of the decoder does not need to use each pixel of the reconstructed texture block of one or more views to evaluate the likelihood of each possible corresponding depth of this block in order to reconstruct the depth block. Indeed, according to this embodiment, the depth estimator simply estimates the depth of the reconstructed depth block only within the interval between the minimum and maximum depth values ​​of this block. Such depth estimation significantly speeds up the depth estimation step, which is a complex aspect of the prior art.

[0020] According to another particular embodiment, said at least one depth estimation parameter is a parameter used by a depth estimation method.

[0021] According to this particular embodiment, the depth estimation parameters used by known depth estimation methods are advantageously optimized to result in estimated depth blocks that are as close as possible to the original depth blocks. Such parameters are, for example, regularization parameters that allow the depth estimation algorithm to find a depth map that is less noise-free, or confidence parameters that allow the depth estimation algorithm to keep improving the depth values ​​if its confidence would otherwise be too low.

[0022] According to another particular embodiment, the information representative of the depth estimation method is coded.

[0023] According to this particular embodiment, the coder is able to test different available depth estimation methods, each of which may give more or less good results for a given content or block, select the depth estimation method that gives the best depth estimation for the current block, and encode this selection for sending it to the decoder to apply the selected depth estimation to the current block.

[0024] The various above-mentioned embodiments or features may be added individually or in combination with each other to the encoding method defined above.

[0025] The present invention relates to a device for encoding views simultaneously representing a 3D scene from different positions or different viewing angles, the device comprising, for the depth component of at least one view: - dividing the depth component into at least one block; - deriving depth information for the at least one block from texture data of at least one texture component of the view; - deriving at least one depth estimation parameter from the depth information; - encoding the at least one depth estimation parameter, wherein depth information of the at least one block is not encoded. The present invention also relates to a device including a processor configured to implement the method.

[0026] Such a coding device is in particular capable of implementing the coding method described above.

[0027] The present invention relates to a method for decoding views simultaneously representing a 3D scene from different positions or different viewing angles, implemented by a decoding device, comprising the steps of: for a depth component of at least one view, the depth component being divided into at least one block: - reading in a data signal at least one depth estimation parameter associated with said at least one block; - decoding the at least one depth estimation parameter; - deriving depth information for the at least one block from the at least one depth estimation parameter and from texture data of at least one reconstructed texture component of the view. The present invention also relates to a method comprising:

[0028] Such a decoding method according to the invention has low computational complexity and can advantageously be saved on memory resources. Indeed, since the depth information of the blocks is not coded and therefore not transmitted to the decoder, the decoder does not need to decode or store this information. During decoding, the decoder only needs to reconstruct the depth information of the blocks by decoding at least one depth estimation parameter transmitted in the received data signal, said at least one depth estimation parameter being less expensive to transmit than the depth information.

[0029] According to a particular embodiment, the at least one depth estimation parameter is either a depth value of the at least one block that is greater than the respective depth value of the at least one block or a depth value of the at least one block that is less than the respective depth value of the at least one block.

[0030] According to a particular embodiment, said at least one depth estimation parameter is a parameter used by a depth estimation method.

[0031] According to another particular embodiment, the information representative of the depth estimation method is decoded.

[0032] The various above-mentioned embodiments or features may be added individually or in combination with each other to the decoding method defined above.

[0033] The present invention relates to a device for decoding views simultaneously representing a 3D scene from different positions or different viewing angles, the device comprising, for a depth component of at least one view, the depth component being divided into at least one block: - reading in a data signal at least one depth estimation parameter associated with said at least one block; - decoding the at least one depth estimation parameter; - deriving depth information for the at least one block from the at least one depth estimation parameter and from texture data of at least one reconstructed texture component of the view. The present invention also relates to a device including a processor configured to implement the method.

[0034] Such a decoding device is in particular capable of implementing the above-mentioned decoding method.

[0035] The present invention relates to a view synthesis method implemented by a decoding device or a view synthesis device, comprising: - reconstructing the view according to the decoding method according to the invention mentioned above, - synthesizing at least a portion of a view from the reconstructed view and from obtained information representing the depth of said at least one block. The present invention also relates to a view synthesis method including:

[0036] The invention also relates to a computer program comprising instructions for implementing the encoding, decoding or combining method according to the invention, according to any one of the particular embodiments described above, when said program is executed by a processor.

[0037] Such instructions may be persistently stored in a non-transitory memory medium of an encoding device implementing the encoding method described above, a decoding device implementing the decoding method described above, or a compositing device implementing the compositing method described above.

[0038] This program may use any programming language and may be in the form of source code, object code, or an intermediate code between source code and object code, a partially compiled form, or any other desired form.

[0039] The present invention is also directed to a computer readable storage medium or information medium containing instructions for a computer program as mentioned above.

[0040] The storage medium may be any entity or device capable of storing a program. For example, the medium may comprise a storage means such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or even a magnetic storage means, for example a USB key or a hard disk.

[0041] Likewise, the storage medium may be a medium that can be transmitted wirelessly or by other means, such as an electrical or optical signal that can be transmitted via an electrical or optical cable. The program according to the invention can be downloaded, in particular over a network of the Internet type.

[0042] Alternatively, the storage medium may be an integrated circuit having a program embedded therein, the integrated circuit being adapted to perform the above-described encoding method, the above-described decoding method or even the above-described synthesis method or an integrated circuit being used in performing these methods.

[0043] Other features and advantages will become apparent on reading the particular embodiments of the invention provided as illustrative and non-limiting examples and in the accompanying drawings, in which: [Brief explanation of the drawings]

[0044] [Figure 1] 3 illustrates the progression of a method for encoding a view in accordance with a particular embodiment of the invention; [Figure 2A] 2 represents a first embodiment of the step of obtaining depth estimation parameters implemented in the encoding method of FIG. 1; [Figure 2B] 2 illustrates a second embodiment of the step of obtaining depth estimation parameters implemented in the encoding method of FIG. 1; [Figure 3A] 2 represents a first embodiment of the signaling of information coded by the coding method of FIG. 1; [Figure 3B] 2 illustrates a second embodiment of the signaling of information encoded by the encoding method of FIG. 1; [Figure 4] 2 illustrates a video encoding device that implements the encoding method of FIG. 1; [Figure 5] 3 illustrates the progression of a method for decoding a view in accordance with a particular embodiment of the present invention; [Figure 6] 6 illustrates a video decoding device that implements the decoding method of FIG. 5. [Figure 7] 3 illustrates the progression of a missing view synthesis method in a particular embodiment of the present invention. [Figure 8A] 8 illustrates a synthesis device that implements the synthesis method of FIG. 7 in accordance with a specific embodiment of the present invention. [Figure 8B] 8 illustrates a synthesis device for implementing the synthesis method of FIG. 7 in accordance with another specific embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0045] Example implementation of a multi-view video coding method A method for encoding multi-view video is described below, which can use any type of multi-view video coder, for example one that complies with the 3D-HEVC or MV-HEVC standard or other standards.

[0046] Referring to FIG. 1, such an encoding method involves multiple views V1,...,V N is applied to the current view that forms part of the 3D scene, where the multiple views represent the 3D scene from multiple viewing angles or multiple positions / orientations of the camera capturing the scene, respectively.

[0047] The encoding method according to the present invention currently comprises: - View V1, - View V2, - ..., - View V i , - ..., - View V N This encodes the following.

[0048] A view considered out of N may equally be a texture component or a depth component or a map associated with this view. For example, the current view V i (1≦i≦N) is the texture component T of Q (Q≧1) pixels. i and for example texture component T i a depth component P having Q depth values ​​associated with Q pixels of at least one texture component such that i associated with, or among N, view V i As is known, the depth component P i is the texture image T i It can be generated directly from or otherwise by capturing volumetric data of the 3D scene using a device such as LIDAR (an abbreviation for "light detection and ranging").

[0049] In C1, the current view V i is selected at the current time point, and each of the N views is selected in turn in a predetermined order.

[0050] C2, the at least one view V i The depth component P i is multiple blocks B1,B2,...,B j ,...,B M In one possible embodiment, a single depth block is divided into undivided depth components P i The blocks of depth components are of a predefined size (e.g., 64x64 pixels) and can be parameterizable (and the size used is then transmitted in coded form) or otherwise adaptive using signaling of the size used, similar to that implemented in the HEVC standard. In the latter case, the depth component P iis first divided into blocks of maximum size (e.g., 64x64 pixels), and then binary information indicating whether the block should be subdivided into smaller blocks is repeatedly transmitted for each block until a predefined minimum block size (e.g., 4x4 pixels) is reached, for which no information is transmitted. This allows the depth component P i This allows for the definition of a subdivision into blocks.

[0051] In C3, the depth component P i Current Block B j is selected, and the depth component P i are selected one after the other in a predetermined order.

[0052] C4, the at least one current block B j For depth information IP j As is known per se, the texture component T i All or part of the pixels of the block BT and / or N of the view V i For all or part of the texture components of the views other than R and the depth block BP corresponding to the pixel block BT is obtained. j Form.

[0053] In C5, at least one depth estimation parameter PE is added to the depth information IP i is obtained from

[0054] According to the first acquisition embodiment shown in FIG. 2A, in C51a, R depth values ​​Z1 to Z R The maximum depth value Z max and / or minimum depth value Z min is determined.

[0055] In C52a, Z min or Z max or interval [Z min ,Z max] is assigned to the depth estimation parameter PE.

[0056] According to the second acquisition embodiment represented in Fig. 2B, in C51b at least one depth estimation parameter PME of the depth estimation method is selected. This is for example a parameter of the depth estimation algorithm DERS mentioned above. According to another example, it could be a parameter of another depth estimation algorithm, for example an IVDE algorithm. For the DERS algorithm, the selected parameter is for example a regularization parameter or "smoothing factor" Sc that allows the depth estimation algorithm DERS to find a noise-free depth block corresponding to the block of pixels BT. Other examples of parameters of a depth estimation algorithm are, for example: a confidence parameter, such as the parameter λ of the DERS algorithm, which allows the DERS algorithm to continue to refine the depth value if its confidence is too low; smoothing parameters such as the initial smoothing parameter β0 used in the IVDE algorithm and described in the above-mentioned document "Dawid Mieloch, Adrian Dziembowski, Jakub Stankowski, Olgierd Stankiewicz, Marek Domanski, Gwangsoon Lee, Yun Young Jeong [MPEG-I Visual] Immersive video depth estimation, ISO / IEC JTC1 / SC29 / WG11 MPEG2020 m53407" It would also be possible to use etc.

[0057] In C52b, there are X (X≧1) possible values ​​Val1,...,Val of the depth estimation parameter PME of the selected depth estimation method. k ,...,Val X For a finite set of (1≦k≦X), we define X estimated depth blocks BPE1,...,BPE k ,...,BPE XFor the parameter Sc, a finite set of possible values ​​is, for example, {0.01, 0.02, 0.04, 0.08, 0.16}. Of course, other values ​​are also possible depending on the current video context.

[0058] In C53b, X blocks BPE1,...,BPE k ,...,BPE X The estimated depth obtained from the pixel block BT is compared with the original depth block BP j Such selection can be implemented using a distortion measure such as PSNR (short for "Peak Signal to Noise Ratio"), mean square error, sum of absolute differences or any other similar measure. In the example shown, it can be implemented using, for example, the selected estimated depth block BPE k is.

[0059] In C54b, the estimated depth block BPE selected in C53b is k The value of the depth estimation parameter PME of the selected depth estimation method for k is then selected as the value of the depth estimation parameter PE.

[0060] Referring again to Figure 1, in C6 the depth estimation parameter PE is coded using a lossless coding method such as CABAC (an abbreviation for "Context-Adaptive Binary Arithmetic Coding") or even by Huffman coding or Lempel-Ziv coding. In the example of Figure 2A, it is coded by the value Z min or Z max or further interval [Z min ,Z max ], which are coded in C6. In the example of FIG. 2B, it is the value Val of the regularization parameter Sc coded in C6. k or in yet other embodiments, it is also the confidence parameter or smoothing parameter mentioned above.k In addition to , it is possible to encode the value of another parameter, such as the reliability parameter or the smoothing parameter mentioned above, which could be obtained at the end of step C54b. At the end of step C6, the encoded depth estimation parameter PE C is obtained.

[0061] Optionally, and for this reason as represented by the dotted line in Fig. 1, information IMEP, e.g., DERS or IVDE, representing the depth estimation method used in C5, is coded in C7 using a lossless coding method such as CABAC. Such IMEP information is then transmitted to the view V i or a view V located at a different time. i At the end of step C7, the encoded information IMEP C is obtained.

[0062] In C8, the texture component T i is encoded using a conventional video coder such as HEVC. At the end of step C8, the encoded texture component T i C is obtained.

[0063] According to a first embodiment illustrated in FIG. 3A, the coded depth estimation parameters PE C , coded information IMEP C and the encoded texture component T i C The data is contained in the same data signal F that is intended to be transmitted to a decoder as described later in this specification.

[0064] According to a second embodiment illustrated in FIG. 3B, the coded depth estimation parameters PE C and coded information IMEP C are contained in the same data signal F, while the coded texture component T i CThe data is contained in another data signal F', signals F and F' being intended to be transmitted to the decoder mentioned above.

[0065] Advantageously, according to the invention, the coding method generates a coded depth block BP j C As a result, in the example of Figures 3A and 3B, the signal F does not generate the coded depth block BP j C does not contain

[0066] The encoding method just described above then generates the depth component P i Each block B1~B M Then, for each view V1 to V N It can be implemented for each of the following:

[0067] Example of implementation of a video encoding device FIG. 4 presents a simplified structure of a coding device COD suitable for implementing the coding method according to any one of the particular embodiments of the invention.

[0068] According to a particular embodiment of the invention, the actions performed by the encoding method are implemented by computer program instructions, for which purpose the encoding device COD has the architecture of a conventional computer and comprises in particular a memory MEM_C and a processing unit UT_C equipped, for example, with a processor PROC_C, driven by a computer program PG_C stored in the memory MEM_C, which computer program PG_C comprises instructions for implementing the actions of the encoding method as explained above, when the program is executed by the processor PROC_C.

[0069] Upon initialization, the code instructions of the computer program PG_C are loaded, for example, into a RAM memory (not shown), before being executed by the processor PROC_C. The processor PROC_C of the processing unit UT_C implements, in particular, the actions of the coding method described above according to the instructions of the computer program PG_C.

[0070] Example implementation of multi-view video decoding method The following describes a method for decoding multi-view video that can use any type of multi-view video decoder, for example, conforming to the 3D-HEVC or MV-HEVC standard or other standards.

[0071] Referring to FIG. 5, such a decoding method is applied to a data signal representing a current view coded according to the coding method described above, said current view being a set of multiple views V1,...,V N forms part of the

[0072] The decoding method according to the present invention comprises: - a data signal representing the coded view V1, - a data signal representing the encoded view V2, - ..., - Encoded View V i a data signal representing - ..., - Encoded View V N A data signal representing This is to decrypt the

[0073] The decoding method reconstructs the encoded current view V i For data signal F (FIG. 3A) or data signals F and F′ (FIG. 3B) representing:

[0074] In D1, the encoded current view V i is selected at the current time point, and each of the N views is selected one after the other in a predetermined order.

[0075] In D2, the at least one view V i The reconstructed depth component P i is multiple blocks B1,B2,...,B j ,...,B M In one possible embodiment, a single depth block is divided into undivided depth components P i The depth blocks are of a predefined size (for example, 64x64 pixels) and can be parameterizable (and subsequently the size used, transmitted in coded form, is decoded) or otherwise adaptive, with signaling of the size used similar to that implemented in the HEVC standard and read in the signal F. In the latter case, the depth component P i is first divided into blocks of maximum size (for example, 64x64 pixels), and then, for example, in the signal F or another signal, binary information indicating whether a block must be subdivided into smaller blocks is repeatedly read for each block until a predefined minimum block size (for example, 4x4 pixels) is reached, for which no information is read. This results in a depth component P i This allows for the definition of a subdivision into blocks.

[0076] In D3, the depth component P i Current Block B j is selected, and the depth component P i are selected one after the other in a predetermined order.

[0077] In D4, optionally, and for this reason represented by a dotted line in FIG. 5, coded information IMEP representing the depth estimation method used in C5 (FIG. 1) C is read in the data signal F (FIG. 3A or FIG. 3B).

[0078] In D5, the coded information IMEP Cis decoded using a lossless decoding method such as CABAC or even by Huffman decoding or Lampell-Ziv decoding. C is the current view i or a view V located at a different time. i At the end of step D5, the information IMEP is obtained.

[0079] As an alternative to the optional steps D4 and D5, the depth estimation method used in C5 can also be predefined in the decoder, in which case the IMEP information representing the depth estimation method is directly available.

[0080] In D6, the depth component P i of the current block B to be reconstructed j and at least one coded depth estimation parameter PE associated with C is read in the data signal F of FIG. 3A or 3B.

[0081] At D7, the at least one coded depth estimation parameter PE C is decoded using a lossless decoding method such as CABAC, or else by Huffman decoding or Lampell-Ziv decoding. min or Z max If it is or else the interval [Z min ,Z max ], - Depth value Z min is decoded in D7 and assigned to the depth estimation parameter PE, or - or depth value Z max is decoded in D7 and assigned to the depth estimation parameter PE, or - Depth value interval [Z min ,Z max ] is decoded in D7 and assigned to the depth estimation parameter PE Either:

[0082] The value Val of the regularization parameter Sc is coded in C6. k and / or the value of the reliability parameter or smoothing parameter Val according to other embodiments. k , the value Val k This value Val k In addition to this, the value of another parameter can be decoded, such as the reliability parameter or the smoothing parameter mentioned above obtained at the end of step C54b.

[0083] In D8, the at least one view V i Texture component T i is reconstructed using, for example, a conventional video decoder, such as HEVC. The reconstructed texture component T i R is obtained at the end of step D8. Needless to say, step D8 can be implemented before steps D1 to D7 or at any time upon receiving the data signal F (FIG. 3A) or F' (FIG. 3B). For this purpose, the coded texture component T i C The data of is read in the signal F (FIG. 3A) or F′ (FIG. 3B) and then reconstructed to obtain the reconstructed texture component T i R Generate.

[0084] In D9, the current block B j Depth information IP j is the at least one depth estimation parameter P E decoded in D7 and the reconstructed texture component T i R or N out of views V i It is obtained from the texture data (pixels) of the reconstructed texture components of the views other than

[0085] According to a first embodiment of step D9, by using a predefined depth estimation method or one corresponding to the IMEP parameters obtained in D5, the depth search is performed to obtain the reconstructed depth component P i Current Block B j Depth value Z min or depth value Z max Or further, the interval between depth values ​​[Z min ,Z max ] to obtain the reconstructed texture component T i R The advantage of this operation is that the theoretically possible depth is not between the maximum and minimum depths (generally 0 meters to infinity), but at most [Z min ,Z max ] between the current block B j This allows to reduce the complexity of the depth estimation algorithm by significantly reducing the number of depths to be tested. At the end of the first embodiment of step D9, the current block B j The reconstructed depth block BP corresponding to j R is obtained.

[0086] According to a second embodiment of step D9, the predefined depth estimation method or the corresponding IMEP parameters obtained in D5 is used to calculate the value of the regularization parameter Sc (or for example the value of the confidence parameter or smoothing parameter) Val decoded in D7. k By using the current block B j and the reconstructed texture component T i R At the end of step D9 of the second embodiment, the current block B j The reconstructed depth block BP corresponding to j R This second embodiment provides the reconstructed depth block BP j Ris the depth block BP obtained in C4 during the encoding method of Fig. 1 j , and advantageously, according to the invention, this depth block BP j is not encoded or transmitted in signals F or F'.

[0087] In D10, the reconstructed depth block BP j R is then the reconstructed texture component T i R The corresponding, now reconstructed, P i R This is included in the depth component.

[0088] The decoding method just described above then reconstructs each block of pixels B1 to B M Then, for each view V1 to V2, N It is implemented for each.

[0089] Example of implementation of a video decoding device FIG. 6 presents a simplified structure of a decoding device DEC suitable for implementing the decoding method according to any one of the particular embodiments of the invention.

[0090] According to a particular embodiment of the invention, the actions performed by the above-described decoding method are implemented by computer program instructions. To that end, the decoding device DEC has a conventional architecture of a computer and includes in particular a memory MEM_D and a processing unit UT_D, which is for example equipped with a processor PROC_D and driven by a computer program PG_D stored in the memory MEM_D. The computer program PG_D includes instructions for implementing the actions of the decoding method as described above when the program is executed by the processor PROC_D.

[0091] Upon initialization, the code instructions of the computer program PG_D are loaded, for example, into a RAM memory (not shown), before being executed by the processor PROC_D. The processor PROC_D of the processing unit UT_D follows the instructions of the computer program PG_D and in particular implements the actions of the decoding method described above.

[0092] Example implementation of the view synthesis method A view synthesis method using views reconstructed according to the decoding method of FIG. 5 will now be described with reference to FIG.

[0093] As shown in FIG. 7, the synthesis method according to the invention uses the N reconstructed views V1 obtained at the end of the decoding method of FIG. R ,...,V N R Use at least one reconstructed view from

[0094] In S1, there is at least one reconstructed view V q R (1≦q≦N) is selected from the N reconstructed views. The reconstructed view V q R is the reconstructed texture component T q R and the associated reconstructed depth component P q R Includes:

[0095] At S2, at least one synthesized partial PV of a missing or intermediate view, such as a view that the user requested to be displayed on the screen of the user's computer or the user's phone, is displayed. sy is the reconstructed texture component T q R and this reconstructed texture component T q R A reconstructed block of pixels B y R At least one reconstructed depth block BP associated withy R It is calculated from the following equation, where 1≦y≦M. The partial PV of the missing or intermediate view is synthesized. sy is calculated using a conventional synthesis algorithm, such as the VSRS algorithm, the RVS ("Reference View Synthesizer") algorithm, or the VVS ("Generic View Synthesizer") algorithm.

[0096] Example of a synthetic device implementation FIG. 8A or FIG. 8B presents a simplified structure of a synthesis device SYNT suitable for implementing the synthesis method of FIG. 7 according to any one of the particular embodiments of the present invention.

[0097] According to a particular embodiment of the invention, the actions performed by the synthesis method of Fig. 7 are implemented by computer program instructions. To that end, the synthesis device SYNT has a conventional architecture of a computer and includes in particular a memory MEM_S and a processing unit UT_S, which is for example equipped with a processor PROC_S and driven by a computer program PG_S stored in the memory MEM_S. The computer program PG_S includes instructions for implementing the actions of the synthesis method as explained above when the program is executed by the processor PROC_S.

[0098] Upon initialization, the code instructions of the computer program PG_S are loaded, for example, into a RAM memory (not shown), before being executed by the processor PROC_S. The processor PROC_S of the processing unit UT_S follows the instructions of the computer program PG_S and in particular implements the actions of the synthesis method described above.

[0099] According to the embodiment represented in FIG. 8A, the synthesizing device SYNT is arranged at the output of the decoder DEC, as illustrated in FIG. 8A.

[0100] According to the embodiment represented in FIG. 8B, the synthesis device SYNT forms an integral part of the decoder DEC, as illustrated in FIG. 8B. [Explanation of symbols]

[0101] C1 Step C2 Step C3 Step C4 Step C5 Step C6 Step C7 Step C8 Step

Claims

1. 1. A method for encoding views representing a 3D scene simultaneously from different positions or different viewing angles, implemented by an encoding device, comprising: for a depth component of at least one view: - dividing (C2) said depth component into at least one block; deriving (C4) depth information for said at least one block of said depth component from texture data of at least one texture component of said view; - deriving (C5) at least one depth estimation parameter from said depth information of at least one block of said depth components; coding (C6) said at least one depth estimation parameter, wherein said depth information of said at least one block is not coded; A method comprising:

2. The at least one depth estimation parameter is a depth value (Z ｍａｘ ) or a depth value (Z ｍｉｎ 2. The encoding method according to claim 1, wherein the encoding method is either one of the following:

3. The at least one depth estimation parameter is a parameter (Sc; λ; β) used by a depth estimation method. ０ 2. The encoding method according to claim 1, wherein:

4. Coding method according to any one of claims 1 to 3, wherein information representative of the depth estimation method is coded.

5. 1. A device for encoding views simultaneously representing a 3D scene according to different positions or different viewing angles, comprising: for a depth component of at least one view: - dividing said depth component into at least one block; deriving depth information for said at least one block of said depth component from texture data of at least one texture component of said view; - deriving at least one depth estimation parameter from said depth information of at least one block of said depth components; - coding said at least one depth estimation parameter, wherein said depth information of said at least one block is not coded. A device including a processor (UT_C) configured to implement the

6. A computer program comprising program code instructions for implementing the encoding method according to any one of claims 1 to 4, when the computer program is run on a computer.

7. A computer-readable information medium containing instructions for the computer program of claim 6.

8. 1. A method for decoding views simultaneously representing a 3D scene from different positions or different viewing angles, implemented by a decoding device, comprising: for a depth component of at least one view, the depth component being divided into at least one block: - reading (D6) in a data signal at least one depth estimation parameter associated with said at least one block of said depth component; decoding (D7) said at least one depth estimation parameter, deriving (D9) depth information of said at least one block of said depth component from said at least one depth estimation parameter and from texture data of at least one reconstructed texture component of said view; A method comprising:

9. The at least one depth estimation parameter is a depth value (Z ｍａｘ ) or a depth value (Z ｍｉｎ 9. The decoding method according to claim 8, wherein the decoding method is either one of the above.

10. 9. The decoding method of claim 8, wherein said at least one depth estimation parameter is a parameter (Sc) used by a depth estimation method.

11. Decoding method according to any one of claims 8 to 10, wherein information representative of a depth estimation method is decoded.

12. 1. A device for decoding views simultaneously representing a 3D scene from different positions or different viewing angles, the device comprising: for a depth component of at least one view, the depth component being divided into at least one block; - reading in a data signal at least one depth estimation parameter associated with said at least one block of said depth component; decoding said at least one depth estimation parameter; deriving depth information for said at least one block of said depth component from said at least one depth estimation parameter and from texture data of at least one reconstructed texture component of said view; A device including a processor (UT_D) configured to implement the above.

13. 1. A view synthesis method implemented by a decoding device or a view synthesis device, comprising: - reconstructing the view according to a decoding method according to any one of claims 8 to 11, - synthesizing at least a part of a view from said reconstructed view and from said information obtained representing the depth of said at least one block; A view synthesis method comprising:

14. A computer program comprising program code instructions for implementing the decoding method according to any one of claims 8 to 11 or the combining method according to claim 13, when the computer program is run on a computer.

15. A computer-readable information medium containing instructions for the computer program of claim 14.

Images