• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The present invention relates to methods, devices or streams for encoding, transmitting and decoding two-dimensional point clouds. When encoding point clouds as frames, a large number of pixels are not used. A dense mapping operator optimizes the use of pixels, but requires a lot of data to be encoded in the stream and its inverse operator is difficult to calculate. According to the present principles, a simplified mapping operator is generated according to a dense mapping operator and is stored as two-dimensional coordinate arrays representative of an exposed grid that requires little space in the stream. The inverse operator is easy to generate according to the exposed grid.
    公开号:BR102018009143A2
    申请号:R102018009143-3
    申请日:2018-05-04
    公开日:2019-03-19
    发明作者:Julien Fleureau；Bertrand Chupeau；Renaud Dore
    申请人:Thomson Licensing；
    IPC主号:
    专利说明:

    METHOD AND APPARATUS TO CODE AND DECODE CLOUDS FROM TWO DIMENSION POINTS
    1. Field of technique [001] The present disclosure refers to the domain of encoding and decoding data representing two-dimensional point clouds as pixel frames, especially when the point cloud comprises dense and empty parts.
    2. Precedents [002] A frame is a series of pixels containing a value, for example, a color value or a depth value. A frame can be used to represent an image, in which case, each pixel is set to a significant value. A frame can also be used to encode other types of data, for example, a two-dimensional point cloud. A two-dimensional point cloud is a set of two-dimensional coordinates associated with a value, for example, color or depth information. The coordinates of the points are expressed in the frame's reference frame. A two-dimensional point cloud can be the result of projecting a n-dimensional point cloud onto a surface, with n greater than two.
    [003] In this case, some pixels contain dots and are associated with a significant value (for example, color or depth) and some others do not contain dots and have no value at all. To encode such an image, a default value is assigned to the unused pixels. The default value must be excluded from the range of significant values. In addition, since the frame is a swept surface, two nearby points may have to be encoded on the same pixel. This is a problem on the decoding side, as it is no longer possible to distinguish these two points from a single piece of information. This problem can be solved by using an unused pixel to encode one of the colon and thus not
    Petition 870180036897, of 05/04/2018, p. 10/44
    2/25 mix your value in a single pixel. The problem then is to retrieve the original coordinates of the points.
    [004] Several techniques exist to optimize the use of pixels in a frame containing unused pixels. Grid generation algorithms, for example, generate a mapping operator, in order to better expand the points of the two-dimensional point cloud on the board. The principles include encoding distinct points into distinct pixels as much as possible by enlarging dense parts of the projected point cloud and narrowing the empty parts of the projected point cloud. The mapping operator is often not a function (but a more complex algorithm) and is sometimes not invertible. Such a mapping operator requires a lot of space to be coded in a stream, since it is not a parameterized function, but a complex algorithm. In addition, on the decoding side, the inverse of the mapping operator needs to be calculated in order to recover the original point cloud. The computation of such a reverse operator is long and resource-consuming, even if possible. In reality, some of the mapping optimization solutions found do not have inverse operators. There is no technical solution to optimize the expansion of the point cloud in a frame with an efficient inclusion of the reverse mapping operator within the flow.
    3. Summary [005] One aspect of the present disclosure involves overcoming the lack of an encoding and decoding method for storing, compressing and transmitting two-dimensional point clouds as frames.
    [006] Another aspect of the present disclosure relates to a method of decoding a two-dimensional point cloud of a bit stream. The method comprises:
    - obtain, from the bit stream, a frame of pixels and representative data of
    Petition 870180036897, of 05/04/2018, p. 11/44
    3/25 an exposed grid, the data being associated with the frame;
    - determine an operator to undo the mapping according to said exposed grid and
    - decode the point cloud by applying the operator to undo the mapping on the board.
    [007] In one mode, the operator to undo the mapping can be a bilinear interpolator in parts parameterized with the grid exposed.
    [008] According to another aspect, the exposed grid can be associated with a group of frames in the bit stream and the method further comprises decoding the frames of the group of frames by applying the operator to undo the mapping determined according to said grid exposed in the group's boards.
    [009] The present disclosure also refers to a method of encoding a two-dimensional point cloud into a bit stream. The method comprises:
    - generate an exposed grid by mapping a regular grid according to a dense mapping operator;
    - generate a pixel frame by applying a mapping operator to the point cloud, said mapping operator being determined according to the exposed grid, and
    - generating the bit stream by encoding said frame associated with the representative data of said grid exposed in the bit stream.
    [010] The method generates an operator by optimizing the distribution of points on the board, in order to minimize the number of 'unused' pixels and maximize the number of pixels of information. An advantage of the method can be linked to the small size of the data required to encode the mapping operator in the bit stream and the few processing resources needed to calculate the operator and its inverse in decoding.
    Petition 870180036897, of 05/04/2018, p. 12/44
    4/25 [011] In one mode, the mapping operator is a bilinear interpolator in parts parameterized with the grid exposed.
    [012] According to one aspect, the dense mapping operator is determined for a group of point clouds, and the method further comprises generating a frame for each point cloud in the point cloud group and associating the exposed grid with the generated group of frames in the bit stream.
    [013] In one embodiment, the two-dimensional point cloud can be a projection of a n-dimensional point cloud onto a surface, n being greater than 2.
    [014] The present disclosure also relates to a device comprising a memory associated with at least one processor configured for:
    - obtaining, from a bit stream, a frame of pixels and data representative of an exposed grid, said data being associated with said frame;
    - determine an operator to undo the mapping according to the exposed grid and
    - decode the point cloud by applying an operator to undo the mapping on the board.
    [015] In one mode, the operator to undo the mapping comprises a bilinear interpolator in parts parameterized with the exposed grid.
    [016] According to another aspect, the exposed grid can be associated with a group of frames in the bit stream and at least one processor can be further configured to decode frames in the group of frames by applying the operator to undo the determined mapping of according to the grid shown in the tables.
    [017] The present disclosure also refers to a device that comprises a memory associated with at least one configured processor
    Petition 870180036897, of 05/04/2018, p. 13/44
    5/25 to:
    - generate an exposed grid by mapping a regular grid according to a dense mapping operator;
    - generate a pixel frame by applying a mapping operator to a two-dimensional point cloud, the mapping operator being determined according to the exposed grid, and
    - generate a bit stream by encoding the frame associated with data representative of the grid exposed in the bit stream.
    [018] In one mode, the mapping operator comprises a bilinear interpolator in parts parameterized with the exposed grid.
    [019] According to another aspect, the dense mapping operator can be determined for a group of point clouds, and at least one processor can be further configured to generate a frame for each point cloud of said group of point clouds. points and associate said exposed grid with the generated group of frames in the bit stream.
    [020] In one embodiment, the two-dimensional point cloud can be a projection of a n-dimensional point cloud onto a surface, n being greater than 2.
    [021] The present disclosure also refers to a stream carrying data representative of a two-dimensional point cloud, where the data comprises:
    - a first element of syntax related to an exposed grid generated by mapping a regular grid according to a dense mapping operator, and
    - a second element of syntax relative to at least one pixel frame generated by the application of a mapping operator in the two-dimensional point cloud, the mapping operator being determined according to
    Petition 870180036897, of 05/04/2018, p. 14/44
    6/25 the exposed grid;
    wherein the exposed grid can be associated with said at least one frame.
    4. List of figures [022] The present disclosure will be better understood, and other specific characteristics and advantages will appear with the reading of the following description, the description referring to the attached drawings in which:
    - Figure 1 illustrates a mapping operator to encode a cloud of points in a scanned frame, according to a modality of the present principles,
    - Figure 2 illustrates a dense mapping operator (DMO) that maps the points of the point cloud of figure 1 on a table in a non-orthogonal manner, according to a modality of the present principles,
    - Figure 3 shows diagrammatically the calculation of a simplified mapping operator (SMO) of a dense mapping operator in figure 2, according to the present principles,
    - Figure 4 shows diagrammatically the coding of a point cloud of figure 1 in a rectangular frame using an SMO as described in figure 3, according to one modality,
    - Figure 5 illustrates diagrammatically the decoding of a point cloud of a frame associated with the data respectively corresponding to the frame in Figure 4 and the data in Figure 3, according to one modality,
    - Figure 6 illustrates how interpolation in bilinear parts can be used in accordance with these principles,
    - Figure 7 illustrates a method for encoding two-dimensional point clouds according to a non-restrictive modality of the present principles,
    Petition 870180036897, of 05/04/2018, p. 15/44
    7/25
    - Figure 8 illustrates a method of decoding a cloud of two-dimensional points in a stream comprising a frame and data representative of an exposed two-dimensional grid, according to a modality of the present principles,
    - Figure 9 shows an exemplary architecture of a device that can be configured to carry out a method described in relation to Figures 7 and / or 8, according to a modality of the present principles,
    - Figure 10 illustrates an example of a transmission of the frame and data of figures 3, 4 and 5 between two remote devices of figure 9 through a communication network, according to a modality of the present principles,
    - Figure 11 shows an example of a syntax modality of such a flow when data is transmitted through a packet-based transmission protocol, according to a modality of the present principles.
    5. Detailed description of the modalities [023] The exposed material is now described with reference to the drawings, in which equal reference numerals are used to refer to equal elements by all of them. In the following description, for the purpose of explanation, numerous specific details are presented in order to provide a complete understanding of the exposed matter. It is understood that the modalities of the exposed matter can be practiced without these specific details.
    [024] In accordance with a non-limiting embodiment of the present disclosure, a method and device for encoding, transmitting and decoding two-dimensional point clouds are disclosed in the present document.
    [025] A two-dimensional point cloud is a set of two-dimensional coordinates expressed in a frame of reference and associated with a value (for example, a color or a depth). A two-dimensional point cloud can be the result of a projection onto a
    Petition 870180036897, of 05/04/2018, p. 16/44
    8/25 surface of a point cloud of n dimensions with n greater than two. A two-dimensional point cloud can be encoded in a scanned frame (that is, a series of pixels) mapped to the reference frame. Each point in the point cloud is mapped to a pixel. Multiple points can be mapped to a single pixel. When two or more points are mapped to a single pixel, their associated information is combined, for example, mediated. When, for a given pixel, no points are mapped, that pixel is set to an 'unused' value. Since color or depth pixels are not intended to include an 'unused' value, it is common in the art to create an additional frame in which the pixels include a Boolean ('not used' / 'used') value and a predefined value is given for unused pixels. Such a frame is called a "key frame" (or key image). Keyframes are used as a mask to distinguish between significant pixels and unused pixels. Keyframes are generally not compressed or compressed without loss because a small error in a key image can generate a large error in the decoded image. The same key board can be used for several information boards (for example, a color board and a depth board). Both the information and key frames are encoded in a stream and transmitted together. In a variation, information boards and key boards are transmitted separately.
    [026] Figure 1 illustrates the usual mapping operator for encoding a point cloud in a swept frame. Each point in the point cloud 10 has coordinates in a frame of reference associated with the space and includes at least one information, for example, a depth or a color. A mapping operator 11 is used to map each point to a pixel in a frame 12. Frame 12 has a width (that is, a number of columns of pixels) and a height (that is, a number of rows of pixels) . The identity mapping operator 11 (IMO) expresses the point coordinates of the point cloud 10 accordingly
    Petition 870180036897, of 05/04/2018, p. 17/44
    9/25 with the width and height of the frame 12. The identity mapping operator 11 associates at a point having coordinates (u, v) belonging to [umin, Umax] X [Vmin, vmax], with a corresponding pixel ( i, j) belonging to [0, W-1] x [0, H-1], where W and H respectively mean the width and height of the frame 12.
    [027] A general form for the mapping operator 11 can be given by [eq. 1]
    [028] Where Mo is a normalized mapping operator which is the identity function in the case of Mo = IMO 11. Such an operator has the advantage of being simple to perform with a direct inverse operator and simple to calculate. This operator has no need to be coded in the stream as it is used as the predefined mapping operator and undoed the mapping in the prior art. IMO 11 stores point information in pixels 13 of table 12. For dense parts of the point cloud, it is highly likely that two or more points are mapped to the same pixel. The respective associated values are then combined into a single value (for example, mediated). A large portion 14 of the frame's pixels is set to 'unused' value. This area 14 is useless considering that it could be used to store information from different points that are combined in the same pixel by IMO 11.
    [029] Figure 2 illustrates a dense mapping operator 21 (DMO) that maps the points of the point cloud 10 of figure 1 in a frame 22 in a non-orthogonal manner. DMO 21 is an operator that maps points in a frame differently according to the local density of points in the cloud. Such an operator optimizes the distribution of the points on the board, in order to minimize the number of 'unused' pixels and maximize the number of information pixels. This technical effect is illustrated in figure 2: the useful part 13 of table 22 is wider than the useful part of table 12 of figure 1 while the 'unused' part 14 of the table
    Petition 870180036897, of 05/04/2018, p. 18/44
    10/25 is less than the 'unused' part 14 of table 12. There are techniques to calculate such a dense mapping operator 21 and to execute the adaptive base. Among these techniques, a category of grid generation algorithms (as, for example, in “A Practical Guide to Direct Optimization for Planar Grid-Generation”, JE CASTILLO, JS OTTO) has the advantage of guaranteeing some essential properties that such an operator must have. In particular, the grid generation algorithm has to be chosen to generate the exposed projection of the point cloud space in the frame space. This is a required property of the selected dense mapping operator, in order to allow the computation of a simplified mapping operator, for example, as an interpolator in bilinear parts.
    [030] An optimized dense mapping operator 21 can be calculated for a frame or a sequence of frames, such as a group of images (GOP) of the video compression standards. Computing such a DMO is time consuming and resource consuming. The computation of the reverse operator is also complex. In some cases, the inverse operator is not even mathematically definable. In any case, coding such an algorithm would require a long list of parameters that must be compressed without loss (or even not compressed at all). It would not be efficient to associate the coding of a DMO or the coding of its reverse DMO ' 1 with a frame or a GOP. Computing the DMO optimized for an image or a sequence of images is outside the scope of the present disclosure. The DMO is calculated to be optimal for a point cloud or group of point clouds. DMO 21 can be different for different point clouds to be encoded. The mapping of point cloud 10 in table 22 follows the equation [eq. 1], where Mo = DMO 21.
    [031] Figure 3 shows diagrammatically the calculation of a simplified mapping operator 31 of a dense mapping operator 21, according to the present principles. First, a regular grid of 30 K * L points is
    Petition 870180036897, of 05/04/2018, p. 19/44
    11/25 defined. In figure 3, the grid points are joined by segments for visual purposes only. The grid is a set of points regularly distributed over the point cloud space. The coordinates of these points are regularly distributed in the space reference frame. The information associated with these points is an existence value (that is, a Boolean value). This regular point cloud (which is still called 'the grid') 30 is mapped in Table 32 using the DMO 21 calculated for the point cloud (or group of point clouds) 10 in Figure 2. As a result, the K * L points in grid 30 are mapped to pixels in frame 32. Frame 32 is the same size as frame 22. In Figure 3, the points in frame 32 are joined by segments for visual purposes only. The segments illustrate the key exposure characteristic of dense mapping operators, such as DMO 21: the result of the projection for DMO 21 of grid 30 is a grid; that is, ‘no segment of the mapped grid crosses another segment of the mapped grid’. Points in grid 30 remain in the same structural order after mapping by DMO 21. Both points A and B in grid 30 can be projected onto the same pixel in image 32. If DMO 21 leads to such a situation, it arises from the fact that there is a point in the point cloud (or group of point clouds) 10 in figure 2 between (and / or within the proximity of) the coordinates of points A and B.
    [032] According to the present principles, in this stage, two objects are generated. First, the K * L coordinates of the mapped points in grid 30 are stored and encoded in data 33. Data 33 is encoded as a two-dimensional coordinate K * L matrix. In one variation, data 33 is encoded as two K * L arrays of integers or floating numbers. In a variation, the K * L coordinates are normalized in the image's width and height and are stored in data 33. These matrices will be used to define, for example, by parameterization, the inverse operator. Second, one
    Petition 870180036897, of 05/04/2018, p. 20/44
    12/25 simplified mapping operator 31 is calculated as an interpolator between the grid points 32. For example, the interpolator can be a bilinear interpolator between projected points. In one variation, the interpolator can be a bicubic part interpolator, as described, for example, in Fritsch, FN, & Carlson, RE (1980) “Monotone piecewise cubic interpolation” ·, SIAM Journal on Numerical Analysis, 17 (2) , 238-246. In another variation, the interpolator may be a Lanczos filter that is well known to those skilled in the art.
    [033] The bilinear part interpolation illustrated in figure 6 is a technique well known to those skilled in the image processing field. A point in point cloud 10 mapped by IMO 11 would have coordinates 60a. This position can be expressed in the reference frame of a rectangle defined by four points 61a, 62a, 63a and 64a of the regular grid 30. This position is normalized in the rectangle. For example, the coordinates 60a have an x value equal to a percentage 65 of the segment length [61a - 62a] (and thus the segment [64a - 63a]) and a y value equal to a percentage 66 of the segment length [61a - 64a] (and so from the segment [62a - 63a]). An interpolation in bilinear parts maintains these percentages according to the segments corresponding to the DMO 21 mapping of the four points 61a, 62a, 63a and 64a of the regular grid 30. The coordinates 61b, 62b, 63b and 64b (stored in data 33) are the DMO 21 mapping of the respective coordinates 61a, 62a, 63a and 64a. The interpolation in bilinear parts of the coordinates 60a is then 60b. The percentage values 65 and 66 are maintained by this transformation in the segments [61b - 62b] [64b - 63b] for the x value respectively and in the segments [61b - 64b] and [62b - 63b] for the y value in the reference frame mapped. This operator is simple to be determined according to the coordinates 61b, 62b, 63b and 64b and its application in a point cloud is very fast to calculate. The inverse function is based on the same principles and is simple to determine according to
    Petition 870180036897, of 05/04/2018, p. 21/44
    13/25 with coordinates 61b, 62b, 63b and 64b stored in data 33 and its application in a table is very fast to calculate. Furthermore, a bilinear interpolator in parts has the advantage of being totally determined by a sublist of possibly small and easy to transfer and store node points.
    [034] Figure 4 shows diagrammatically the coding of a point cloud 10 in a rectangular frame 42 using SMO 31 as described in figure 3, according to a particular embodiment of the present principles. If the coordinates of a point in the point cloud 10 correspond to one of the points in grid 30 in Figure 3, the mapping of that point is the corresponding point in the exposed grid 32 stored in data 33. In other cases, the coordinates of a point in the point cloud 10 belong to a “rectangle” of grid 30; that is, the coordinates are located between four of the points of grid 30. The mapping is the interpolation in bilinear parts between the corresponding four points of the projected grid 32. The projection of the point cloud 10 using SMO 21 generates a frame 42, in which the useful part 13 is wider than the useful part of table 12 of figure 1, while the 'unused' part 14 is smaller than the 'unused' part 14 of table 12. The application of SMO 21 follows the equation [eq. 1] in which Mo = SMO 21.
    [035] Frame 42 is encoded and associated with data 33. Frame 42 can be compressed using a standard image or video compression algorithm. The compression algorithm can be chosen to be lossless. In a variation, a key map can be generated according to table 42 and associated with table 42 and data 33. A key map is a table containing Boolean values (for example, 0 or 1), in which useful pixels are defined for a given value (for example, 1) and non-useful pixels are defined for the other value (for example, 0). Key maps are usually not compressed or compressed without loss. Associated encoded data is transmitted to a decoder
    Petition 870180036897, of 05/04/2018, p. 22/44
    14/25 as one or more flows. In one variation, associated encrypted data is stored in local memory or storage medium.
    [036] Figure 5 illustrates diagrammatically the decoding of a point cloud 50 of a frame 52 associated with data 53 respectively corresponding to table 42 in figure 4 and data 33 in figure 3. Table 52 and data 53 are obtained from a source. In this document, obtaining the table and / or the data has to be understood as receiving the data from a source or reading the data from a source. For example, the source belongs to a set that comprises:
    - a local memory, for example, a video memory or a RAM (or random access memory), a flash memory, a ROM (or read memory), a hard disk;
    - a storage interface, for example, an interface with mass storage, RAM, flash memory, ROM, optical disc or magnetic media, and
    - a communication interface, for example, a wired interface (for example, a bus interface, a remote network interface, a LAN interface) or a wireless interface (such as an IEEE 802.11 interface or a Bluetooth interface ®).
    [037] In a variation, data 53 and table 52 are obtained from different sources. Data 53 corresponds to data 33 calculated as described with reference to figure 3. The compaction and transmission may have changed them slightly. In one variation, a lossless compression method was used and safe transmission guaranteed and data 53 is exactly the same as data 33. Data 53 comprises the coordinates of the K * L points of a grid 54. In a variation, the data 33 were normalized in the coding. According to this variation, the coordinates included in data 53 are expressed in a normalized space. A simplified mapping operator
    Petition 870180036897, of 05/04/2018, p. 23/44
    Inverse 15/25 51 (called SMO ' 1 51) is calculated according to data 53 (that is, according to the corresponding exposed grid 54). For normalized pixel coordinates (a, b), the SMO ' 1 rating (a, b) goes down to a bilinear interpolation between the next four values:
    Where A = floor (Ka) and B = floor (Lb).
    [038] The result of applying SMO -1 51 is a cloud of two dimensions 50. SMO 31 has the advantages of DMO 21 without its disadvantages.
    The use of a bilinear part interpolation based on the DMO mapping of a regular grid 30 generates a frame 42, in which the use of pixels is optimized (that is, the number of unused pixels is minimized; the number of points of the point cloud 10 mapped to the same pixel is minimized). A different SMO can be associated with each point cloud in a sequence of point clouds (for example, a video). In one variation, an optimized SMO can be calculated for a group of point clouds in a sequence of point clouds. At the same time, SMO 21 coding requires a small amount of data as a two-dimensional coordinate K * L matrix. On the decoding side, the computation of SMO -1 of data 53 is straightforward and its application, an interpolation in bilinear parts, is very efficient, requires limited processing time and resources.
    [039] Figure 7 illustrates a method 70 for encoding two-dimensional point clouds according to a non-restrictive modality of the present principles. In a step 71, the point cloud (or the sequence of point clouds) to be encoded is obtained from a source. The source can be, for example, a local memory, a storage interface or a communication interface.
    Petition 870180036897, of 05/04/2018, p. 24/44
    16/25
    A two-dimensional point cloud can be obtained by projecting a n-dimensional point cloud onto a surface, n being greater than 2. In the case of a sequence of point clouds, point clouds can be grouped as images are grouped in GOP for standard video compression methods. The frame size (width and height of the target frames) is determined. For example, the frame size determined is a standard image size such as 640 x 480 (VGA) or 1920 x 1200 (widescreen). A dense mapping operator is determined that optimizes the use of pixels from a frame of the given size for the point cloud (or the group of point clouds). The determined DMO minimizes the number of unused pixels.
    [040] In step 72, the determined BMD is applied in a regular grid of K * L points. The result of this calculation is an exposed grid of K * L coordinates of two mapped dimensions that are stored in a K * L array of pairs of floating numbers. In a variation, these coordinates are stored in two matrices of floating numbers. In another variation, these coordinates are mapped in the frame and the indexes of the mapped pixels are stored in an array of pairs of integers or in two arrays of integers. These coordinates are used to parameterize a simplified mapping operator that is an interpolator in bilinear parts.
    [041] In step 73, the parameterized SMO is applied to the point cloud (or each of the group of point clouds) to be coded. A frame (or a group of frames) of the size determined in step 71 is generated. In step 74, the generated frame (or group of frames) is encoded, for example, using a standard image or video compression method such as MPEG 2 or H264. Data representative of the matrix (or matrices) of the mapped grid calculated in step 72 are encoded. The coded frame (or group of frames) and the data are joined in a stream. In one variation, coded frames and data are
    Petition 870180036897, of 05/04/2018, p. 25/44
    17/25 encoded in different streams. The flow is generated at this stage and sent to a destination. The destination belongs to a set that comprises, for example:
    - a local memory, for example, a video memory or a RAM (or random access memory), a flash memory, a ROM (or read memory), a hard disk;
    - a storage interface, for example, an interface with mass storage, RAM, flash memory, ROM, optical disc or magnetic media, and
    - a communication interface, for example, a wiring interface (for example, a bus interface, a remote network interface, a LAN interface) or a wireless interface (such as an IEEE 802.11 interface or a Bluetooth interface ®).
    [042] Figure 8 illustrates a method 80 of decoding a two-dimensional point cloud of a stream comprising a frame and data representative of an exposed two-dimensional grid. In step 81, a pixel frame and associated data representative of the exposed grid (ie, representative of a two-dimensional coordinate matrix) are obtained from a source. The source belongs to a set comprising a local memory, a storage interface, and a communication interface. The data coordinates are representative of an exposed grid of points. A simplified inverse mapping operator (SMO -1 ) is parameterized according to the data coordinates and the size (width and height in pixels) of the frame. SMO ' 1 is a bilinear part interpolator. In step 83, the parameterized SMO ' 1 is applied to the frame obtained in step 81. This application generates a two-dimensional point cloud. The same data can be associated with a frame group and the same SMO ' 1 is used to decode frames in the frame group. Parameterization and application of such an operator is simple and quick and
    Petition 870180036897, of 05/04/2018, p. 26/44
    18/25 require limited processing time and resources for a processor or tablet.
    [043] Figure 9 shows an exemplary architecture of a device 90 that can be configured to carry out a method described in connection with figures 7 and / or 8.
    [044] Device 90 comprises the following elements that are joined by a data bus and addresses 91:
    - a microprocessor 92 (or CPU), which is, for example, a DSP (or digital signal processor);
    - a ROM (or reading memory) 93;
    - a RAM (or random access memory) 94;
    - a storage interface 95;
    - an l / O 96 interface for receiving data to be transmitted from an application and
    - a power supply, for example, a battery.
    [045] According to an example, the power supply is external to the device. In each of the aforementioned memories, the word «recorder» used in the specification can correspond to a small capacity area (a few bits) or a very large area (for example, a complete program or a large amount of received or decoded data) . ROM 93 comprises at least one program and parameters. ROM 93 can store algorithms and instructions for executing techniques in accordance with these principles. When turned on, CPU 92 loads the program into RAM and executes the corresponding instructions.
    [046] RAM 94 comprises, in a register, the program executed by
    CPU 92 and loaded after device 90 is turned on, data input in a register, intermediate data in different method states in a
    Petition 870180036897, of 05/04/2018, p. 27/44
    19/25 register and other variables used to execute the method in a register.
    [047] According to an encoding example or an encoder, the point cloud (or the sequence of point clouds) is obtained from a source. For example, the source belongs to a set that comprises:
    - a local memory (93 or 94), for example, a video memory or a RAM (or random access memory), a flash memory, a ROM (or read memory), a hard disk;
    - a storage interface (95), for example, an interface with mass storage, a RAM, a flash memory, a ROM, an optical disc or a magnetic medium;
    - a communication interface (96), for example, a wired interface (for example, a bus interface, a remote network interface, a LAN interface) or a wireless interface (such as an IEEE 802.11 interface, or Bluetooth® interface) and
    - a user interface, such as a graphical user interface allowing a user to enter data.
    [048] According to examples of the decoding or decoder (s), the frame and the representative data of a coordinate matrix are sent to a destination; specifically, the destination belongs to a set comprising:
    - a local memory (93 or 94), for example, a video memory or a RAM, a flash memory, a hard disk;
    - a storage interface (95), for example, an interface with mass storage, RAM, flash memory, ROM, optical disc or magnetic medium, and
    - a communication interface (96), for example, a wired interface (for example, a bus interface (for example, USB (or serial bus)
    Petition 870180036897, of 05/04/2018, p. 28/44
    20/25 universal)), a remote network interface, a LAN interface, an HDMI interface (high definition multimedia interface)) or a wireless interface (such as an IEEE 802.11, WiFi® interface or a Bluetooth interface ®).
    [049] According to coding or encoder examples, a bit stream comprising the frame and data representative of a two-dimensional coordinate array is sent to a destination. As an example, the bit stream is stored in a local or remote memory, for example, a video memory (94) or a RAM (94), a hard disk (93). In a variation, the bit stream is sent to a storage interface (95), for example, an interface with mass storage, a flash memory, ROM, an optical disc or a magnetic medium and / or transmitted through a communication interface (96), for example, an interface for a point to point connection, a communication bus, a point connection to multiple points, or a broadcast network.
    [050] According to examples of decoding or decoding or rendering, the bit stream is obtained from a source. In an exemplary way, the bit stream is read from a local memory, for example, a video memory (94), a RAM (94), a ROM (93), a flash memory (93) or a hard disk (93 ). In a variation, the bit stream is received from a storage interface (95), for example, an interface with mass storage, RAM, ROM, flash memory, optical disk or magnetic media and / or received from a communication interface (95), for example, an interface for a point-to-point connection, a bus, a point-to-multiple connection or a broadcast network.
    [051] According to examples, device 90 is configured to carry out a method described in relation to figure 7 and belongs to a set comprising:
    Petition 870180036897, of 05/04/2018, p. 29/44
    21/25
    - a mobile device,
    - a communication device,
    - a gaming device,
    - a tablet (or tablet computer),
    - a laptop,
    - a photo camera,
    - a video camera,
    - an encoding chip,
    - a server (for example, a broadcast server, a video on demand server or a network server).
    [052] According to examples, device 90 is configured to perform a rendering method described in relation to figure 8 and belongs to a set comprising:
    - a mobile device,
    - a communication device,
    - a gaming device,
    - a frequency converter,
    - a TV set,
    - a tablet (or tablet computer),
    - a laptop and
    - a monitor (such as an HMD, for example).
    [053] According to an example illustrated in figure 10, in a transmission context between two remote devices 101 and 102 (of the device type 90) over a NET 100 communication network, the device 101 comprises means that are configured to performing a method for encoding point clouds and generating a flow as described in connection with figure 7, and device 102 comprises means that are configured to perform a method for decoding
    Petition 870180036897, of 05/04/2018, p. 30/44
    22/25 point clouds as described in relation to figure 8.
    [054] According to an example, network 100 is a LAN or WLAN network, adapted to broadcast images of photos or videos with associated data from device 101 to decoding devices including device 102.
    [055] Figure 11 shows an example of a syntax modality for such a flow when data is transmitted through a packet-based transmission protocol. Fig. 11 shows an exemplary structure 110 of a bit stream. The structure comprises a container that organizes the flow into independent syntax elements. The structure may comprise a header part 111 which is a common data set for each syntax element of the stream. For example, the header part contains metadata about syntax elements, describing the nature and function of each one. The structure may comprise a payload comprising elements of syntax 112 and 113, the first element of syntax 112 being relative to the representative data of the exposed grids, for example, as two-dimensional coordinate arrays and the second element of syntax 113 being relative to frames generated by the simplified mapping operator. The representative data of an exposed grid comprises information associating the exposed grid with at least one frame. For example, frames are identified by a GOP number and a frame number within the GOP. The GOP number is associated with the exposed grid of the SMO that generated the frames for that GOP.
    [056] Naturally, the present disclosure is not limited to the previously described modalities.
    [057] In particular, the present disclosure is not limited to a method of encoding or decoding a two-dimensional point cloud, but also extends to any method of transmitting a representative stream of point clouds, to an encoding method and decoding sequences of
    Petition 870180036897, of 05/04/2018, p. 31/44
    23/25 point clouds from two mansions and any device that performs these methods. Performing the calculations necessary to generate the tables and representative data for the SMO is not limited to an achievement in shadow type microprograms, but also extends to an achievement in any type of program, for example, programs that can be executed by a CPU type microprocessor. The use of the methods of the present disclosure is not limited to live use, but also extends to any other use, for example, for processing known as post-production processing in a recording studio.
    [058] The achievements described here can be accomplished, for example, in a method or a process, an apparatus, a software program, a data flow or a signal. Even if only discussed in the context of a single embodiment (for example, discussed only as a method or a device), the realization of the features discussed can also be carried out in other forms (for example, a program). A device can be made, for example, in appropriate hardware, software and firmware. The methods can be carried out, for example, on an apparatus such as, for example, a processor, which refers to processing devices in general including, for example, a computer, a microprocessor, an integrated circuit or a programmable logic device. Processors also include communication devices, such as, for example, Smartphones, tablets, computers, mobile phones, portable / personal digital assistants (“PDAs”) and other devices that facilitate the communication of information between end users.
    [059] The achievements of the various processes and characteristics described here can be embodied in a variety of different equipment or applications, particularly, for example, equipment or applications associated with data encoding, data decoding, visualization generation,
    Petition 870180036897, of 05/04/2018, p. 32/44
    24/25 texture processing and other image processing and related texture information and / or depth information. Examples of such equipment include an encoder, a decoder, a post-processor output from a decoder, a pre-processor providing input to an encoder, a video encoder, a video decoder, a video codec, a server network, a frequency converter, a laptop, a personal computer, a cell phone, a PDA and other communication devices. As must be evident, the equipment can be mobile and even installed in a mobile vehicle.
    [060] Additionally, the methods can be performed by instructions being executed by a processor and such instructions (and / or data values produced by an achievement) can be stored in a processor-readable medium such as, for example, an integrated circuit , a software carrier or other storage device such as, for example, a hard disk, a compact floppy disk (“CD”), an optical disk (such as, for example, a DVD, often referred to as a digital versatile disk or a digital video disc), a random access memory (“RAM”) or a read memory (“ROM”). Instructions can form a personified application program in a tangible way on a processor-readable medium. The instructions can be, for example, in hardware, firmware, software or a combination. Instructions can be found, for example, on an operating system, a separate application, or a combination of the two. The processor can therefore be characterized, for example, either as a device configured to perform a process or as a device that includes a processor-readable medium (such as a storage device) having instructions for executing a process. In addition, a readable medium per processor can store, in addition to or in place of instructions, data values produced by an achievement.
    Petition 870180036897, of 05/04/2018, p. 33/44
    25/25 [061] As will be apparent to one skilled in the art, achievements can produce a variety of signals formatted to carry information that can, for example, be stored or transmitted. The information may include, for example, instructions for executing a method or data produced by one of the described achievements. For example, a signal can be formatted to carry, as data, the rules for writing or reading the syntax of a described modality, or to carry, as data, the actual syntax values written by a described modality. Such a signal can be formatted, for example, as an electromagnetic wave (for example, using a portion of the radio frequency spectrum) or as a baseband signal. Formatting can include, for example, encoding a data stream and modulating a carrier with the encoded data stream. The information that the signal carries can be, for example, analog or digital information. The signal can be transmitted over a variety of different wired or wireless connections, as it is known. The signal can be stored in a readable medium per processor.
    [062] Several achievements have been described. However, it will be understood that several modifications can be made. For example, elements of different achievements can be combined, supplemented, modified or removed to produce other achievements. Additionally, a skilled person will understand that other structures and processes can replace those disclosed and the resulting accomplishments will perform at least substantially the same function (s), in at least substantially the same way (s), to achieve at least substantially the same result (s) ) than revealed achievements. Thus, these and other achievements are considered by this request.
    权利要求:
    Claims (15)
    [1]
    1. Method (80) of decoding a two-dimensional point cloud (50) of a bit stream (110), said method CHARACTERIZED by the fact that it comprises:
    - obtaining (81), from the bit stream, a pixel frame (52) and a two-dimensional coordinate matrix (53) representative of an exposed grid (54), said matrix being associated with said frame (52) ;
    - determining (82) an operator to undo the mapping (51) according to said matrix (53) and
    - decoding (83) the point cloud (50) applying said operator to undo the mapping (51) in said frame (52).
    [2]
    2. Method, according to claim 1, CHARACTERIZED by the fact that the operator to undo the mapping comprises a bilinear interpolator in parts parameterized with said matrix.
    [3]
    3. Method, according to claim 1 or 2, CHARACTERIZED by the fact that the matrix (53) is associated with a group of frames in the bit stream (110); the method further comprising decoding the frames of said group of frames by applying the operator to undo the mapping (51) determined according to said matrix in said frames.
    [4]
    4. Method (70) of encoding a two-dimensional point cloud (10) in a bit stream, the method CHARACTERIZED by the fact that it comprises:
    - generate (71) a two-dimensional coordinate matrix (33) representative of an exposed grid (32) by mapping a regular grid (30) according to a dense mapping operator (21); a dense mapping operator being a mapping operator optimizing the use of a frame's pixels for a point cloud;
    Petition 870180036897, of 05/04/2018, p. 35/44
    2/4
    - generating (73) a pixel frame (42) by applying a mapping operator (31) to the point cloud (10), said mapping operator (31) being determined according to said matrix (33) and
    - generating (74) the bit stream (110) by encoding said frame (42) associated with said two-dimensional coordinate matrix (33) representative of said exposed grid (32) in the bit stream.
    [5]
    5. Method, according to claim 4, CHARACTERIZED by the fact that said mapping operator comprises a bilinear interpolator in parts parameterized with said matrix.
    [6]
    6. Method, according to claim 4 or 5, CHARACTERIZED by the fact that the dense mapping operator (21) is determined for a group of point clouds, and further comprising generating a frame for each point cloud of said group of point clouds and associate said matrix (33) with the generated group of frames in the bit stream (110).
    [7]
    7. Method according to one of claims 4 to 6, CHARACTERIZED by the fact that the cloud of points of two dimensions is a projection of a cloud of points of n dimensions on a surface, n being greater than two.
    [8]
    8. Device (90), CHARACTERIZED by the fact that it comprises a memory associated with at least one processor configured for:
    - obtaining, from a bit stream (110), a pixel frame (52) and a two-dimensional coordinate matrix (53) representative of an exposed grid (54), said data being associated with said frame (52 );
    - determining an operator to undo the mapping (51) according to said matrix (54) and
    - decoding the point cloud (50) applying said operator to undo the mapping (51) in said frame (52).
    [9]
    9. Device, according to claim 8, CHARACTERIZED by
    Petition 870180036897, of 05/04/2018, p. 36/44
    3/4 the fact that said operator to undo the mapping comprises a bilinear interpolator in parts parameterized with said matrix.
    [10]
    10. Device according to claim 8 or 9, CHARACTERIZED by the fact that the matrix (53) is associated with a group of frames in the bit stream (110) and in which said at least one processor is still configured for decode frames of said group of frames by applying the operator to undo the mapping (51) determined according to said matrix (53) in said frames.
    [11]
    11. Device (90), CHARACTERIZED by the fact that it comprises a memory associated with at least one processor configured for:
    - generate a two-dimensional coordinate matrix (33) representative of an exposed grid (32) by mapping a regular grid (30) according to a dense mapping operator (21); a dense mapping operator being a mapping operator optimizing the use of pixels in a frame for a point cloud;
    - generating a pixel frame (42) by applying a mapping operator (31) to the two-dimensional point cloud (10), said mapping operator (31) being determined according to said matrix (33) and
    - generating a bit stream (110) by encoding said frame (42) associated with said matrix (33) in the bit stream (110).
    [12]
    12. Device, according to claim 11, CHARACTERIZED by the fact that said mapping operator comprises a bilinear interpolator in parts parameterized with said matrix.
    [13]
    13. Device, according to claim 11 or 12, CHARACTERIZED by the fact that the dense mapping operator (21) is determined for a group of point clouds, and in which said at least one processor is still configured to generate one frame for each point cloud of said group of point clouds and associate said matrix (33) with the group of frames generated in the
    Petition 870180036897, of 05/04/2018, p. 37/44
    4/4 bit stream (110).
    [14]
    14. Device according to one of claims 11 to 13, CHARACTERIZED by the fact that the two-dimensional point cloud (10) is a projection of a n-dimensional point cloud on a surface, n being greater than two .
    [15]
    15. Flow (110) carrying data representing a two-dimensional point cloud (10, 50), CHARACTERIZED by the fact that the data comprises:
    - a first syntax element (112) carrying a two-dimensional coordinate array (33, 53) representative of an exposed grid (32, 54) generated by mapping a regular grid (30) according to a dense mapping operator (21); a dense mapping operator being a mapping operator optimizing the use of pixels from a frame to a point cloud and
    - a second syntax element (113) relative to at least one pixel frame (42, 52) generated by the application of a mapping operator (31, 51) in the two-dimensional point cloud (10, 50), said mapping operator (31, 51) being determined according to said matrix (33, 53);
    wherein the matrix is associated with said at least one frame.
    类似技术:
    公开号 | 公开日 | 专利标题
    BR102018009143A2|2019-03-19|METHOD AND APPARATUS FOR ENCODING AND DECODING TWO-DIMENSION POINT CLOUDS
    US20200244986A1|2020-07-30|Picture prediction method and related apparatus
    WO2016138779A1|2016-09-09|Intra-frame codec method, coder and decoder
    BR112013000963B1|2019-12-03|video encoding and decoding methods and video encoder and decoder
    US10681367B2|2020-06-09|Intra-prediction video coding method and device
    US20170353720A1|2017-12-07|Prediction mode selection method, apparatus and device
    BR112020027024A2|2021-04-06|POINT CLOUD CODING METHOD, POINT CLOUD DECODING METHOD, ENCODER, AND DECODER
    US20130121573A1|2013-05-16|Hybrid codec for compound image compression
    EP3813377A1|2021-04-28|Point cloud coding and decoding method and coder-decoder
    WO2019199513A1|2019-10-17|A method and apparatus for encoding and decoding metadata associated with patched projection of point clouds
    US9661334B2|2017-05-23|Method and apparatus for constructing an epitome from an image
    US20180184096A1|2018-06-28|Method and apparatus for encoding and decoding lists of pixels
    WO2021164323A1|2021-08-26|Coding mode selection method and apparatus, and electronic device and computer readable medium
    WO2020063718A1|2020-04-02|Point cloud encoding/decoding method and encoder/decoder
    WO2020063294A1|2020-04-02|Point cloud encoding and decoding method and encoder/decoder
    BR112021005937A2|2021-06-29|image processing apparatus and method.
    US20200374537A1|2020-11-26|Methods and apparatuses for encoding and decoding a bytestream
    BR112021005904A2|2021-07-27|image processing device and method
    WO2019184489A1|2019-10-03|Method for transformation in image block coding, and method and apparatus for inverse transformation in image block decoding
    TW201601515A|2016-01-01|Method for decoding a color mapping table and corresponding device
    CN106251374B|2019-05-10|MapReduce data processing method in Hadoop based on Zigzag
    Qiu et al.2019|Rich-information watermarking scheme for 3D models of oblique photography
    He et al.2021|Adaptive Compression for Online Computer Vision: An Edge Reinforcement Learning Approach
    EP3079081A1|2016-10-12|Method and device for coding the coordinates of a set of points
    Nandi et al.2014|Fractal image compression using Fast Context Independent HV partitioning scheme and MRBH, MRBHM coding
    同族专利:
    公开号 | 公开日
    RU2018115810A|2019-10-30|
    RU2762005C2|2021-12-14|
    EP3399757A1|2018-11-07|
    US20180324240A1|2018-11-08|
    KR20180122947A|2018-11-14|
    EP3399758B1|2020-09-16|
    EP3399758A1|2018-11-07|
    CN108810571A|2018-11-13|
    JP2018198421A|2018-12-13|
    CA3003608A1|2018-11-04|
    RU2018115810A3|2021-06-23|
    US11122101B2|2021-09-14|
    MX2018005496A|2018-11-09|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US6701028B1|2000-01-11|2004-03-02|Applied Materials, Inc.|Method and apparatus for fast signal convolution using spline kernel|
    US7236613B2|2003-08-29|2007-06-26|The Boeing Company|Methods and apparatus for infrared resolution of closely-spaced objects|
    WO2005025404A2|2003-09-08|2005-03-24|Vanderbilt University|Apparatus and methods of cortical surface registration and deformation tracking for patient-to-image alignment in relation to image-guided surgery|
    US8042056B2|2004-03-16|2011-10-18|Leica Geosystems Ag|Browsers for large geometric data visualization|
    EP2300982B1|2008-06-17|2015-07-22|NTT DoCoMo, Inc.|Image/video quality enhancement and super-resolution using sparse transformations|
    US8179393B2|2009-02-13|2012-05-15|Harris Corporation|Fusion of a 2D electro-optical image and 3D point cloud data for scene interpretation and registration performance assessment|
    US9445072B2|2009-11-11|2016-09-13|Disney Enterprises, Inc.|Synthesizing views based on image domain warping|
    US8525848B2|2009-11-16|2013-09-03|Autodesk, Inc.|Point cloud decimation engine|
    US20110222757A1|2010-03-10|2011-09-15|Gbo 3D Technology Pte. Ltd.|Systems and methods for 2D image andspatial data capture for 3D stereo imaging|
    CN103261912B|2010-07-29|2016-01-20|威凯托陵科有限公司|For the distance of measuring object and/or the equipment of strength characteristics and method|
    GB201103822D0|2011-03-07|2011-04-20|Isis Innovation|System for providing locality information and associated devices|
    TWI538474B|2011-03-15|2016-06-11|杜比實驗室特許公司|Methods and apparatus for image data transformation|
    WO2013133648A1|2012-03-07|2013-09-12|엘지전자 주식회사|Method and device for processing video signal|
    RU2013100160A|2013-01-09|2014-07-20|ЭлЭсАй Корпорейшн|PROCESSING IMAGES WITH SUPER RESOLUTION USING THE REVERSIBLE RARE SPARED MATRIX|
    CN103761744B|2014-02-21|2016-06-15|哈尔滨工业大学|Universal two-dimensional points cloud group technology|
    EP2947630A1|2014-05-23|2015-11-25|My Virtual Reality Software AS|Method for compressing coordinate data|
    US9940727B2|2014-06-19|2018-04-10|University Of Southern California|Three-dimensional modeling from wide baseline range scans|
    US9864583B2|2014-11-14|2018-01-09|Cavium, Inc.|Algorithm to derive logic expression to select execution blocks for programmable network devices|
    CN106339208B|2015-07-15|2019-12-13|腾讯科技（深圳）有限公司|Multimedia information popup window processing method and equipment|
    US10193904B2|2015-07-16|2019-01-29|Qatar University|Data-driven semi-global alignment technique for masquerade detection in stand-alone and cloud computing systems|
    CN105427239B|2015-12-05|2018-05-22|刘忠艳|A kind of two-dimensional points cloud matching process|
    WO2017120552A1|2016-01-06|2017-07-13|Meta Company|Apparatuses, methods and systems for pre-warping images for a display system with a distorting optical component|
    US20170214943A1|2016-01-22|2017-07-27|Mitsubishi Electric Research Laboratories, Inc.|Point Cloud Compression using Prediction and Shape-Adaptive Transforms|
    CN106228570B|2016-07-08|2019-04-09|百度在线网络技术（北京）有限公司|A kind of Truth data determines method and apparatus|
    US10417533B2|2016-08-09|2019-09-17|Cognex Corporation|Selection of balanced-probe sites for 3-D alignment algorithms|
    US20180053324A1|2016-08-19|2018-02-22|Mitsubishi Electric Research Laboratories, Inc.|Method for Predictive Coding of Point Cloud Geometries|
    US20180268570A1|2017-03-16|2018-09-20|Samsung Electronics Co., Ltd.|Point cloud and mesh compression using image/video codecs|
    US20190116372A1|2017-10-16|2019-04-18|Mitsubishi Electric Research Laboratories, Inc.|Systems and Methods for Compressing, Representing and Processing Point Clouds|KR102256913B1|2018-12-03|2021-05-27|고려대학교 산학협력단|3d dynamic point cloud compression method|
    CN111327902A|2018-12-13|2020-06-23|华为技术有限公司|Point cloud encoding and decoding method and device|
    CN111327897B|2018-12-13|2022-01-14|华为技术有限公司|Point cloud encoding method and encoder|
    US11039115B2|2018-12-21|2021-06-15|Samsung Electronics Co., Ltd.|Low complexity color smoothing of reconstructed point clouds|
    WO2020146341A1|2019-01-07|2020-07-16|Futurewei Technologies, Inc.|Point cloud bitstream structure and auxiliary information differential coding|
    KR20200087543A|2019-01-11|2020-07-21|삼성전자주식회사|Method and apparatus for encoding 3-dimensional data and method and apparatus for decoding 3-dimensional data|
    WO2020163038A1|2019-02-07|2020-08-13|Interdigital Vc Holdings, Inc.|Processing a point cloud|
    WO2020189903A1|2019-03-20|2020-09-24|엘지전자 주식회사|Point cloud data transmitting device, point cloud data transmitting method, point cloud data receiving device, and point cloud data receiving method|
    WO2020190097A1|2019-03-20|2020-09-24|엘지전자 주식회사|Point cloud data reception device, point cloud data reception method, point cloud data processing device and point cloud data processing method|
    WO2020190114A1|2019-03-21|2020-09-24|엘지전자 주식회사|Point cloud data transmission device, point cloud data transmission method, point cloud data reception device, and point cloud data reception method|
    WO2020235979A1|2019-05-23|2020-11-26|삼성전자 주식회사|Method and device for rendering point cloud-based data|
    WO2020256244A1|2019-06-21|2020-12-24|엘지전자 주식회사|Point cloud data transmission device, point cloud data transmission method, point cloud data reception device, point cloud data reception method|
    WO2021002558A1|2019-07-03|2021-01-07|엘지전자 주식회사|Point cloud data transmission device, point cloud data transmission method, point cloud data reception device, and point cloud data reception method|
    WO2021066626A1|2019-10-03|2021-04-08|Lg Electronics Inc.|Point cloud data transmission device, point cloud data transmission method, point cloud data reception device, and point cloud data reception method|
    CN110730364B|2019-10-14|2021-08-31|广州微算互联信息技术有限公司|Cloud mobile phone data transmission method and device and storage medium|
    WO2021117859A1|2019-12-13|2021-06-17|ソニーグループ株式会社|Image processing device and method|
    WO2021206365A1|2020-04-11|2021-10-14|엘지전자 주식회사|Point cloud data transmission device, point cloud data transmission method, point cloud data reception device and point cloud data reception method|
    法律状态:
    2019-03-19| B03A| Publication of a patent application or of a certificate of addition of invention [chapter 3.1 patent gazette]|
    2019-07-16| B25G| Requested change of headquarter approved|Owner name: THOMSON LICENSING (FR) |
    2019-07-30| B25A| Requested transfer of rights approved|Owner name: INTERDIGITAL VC HOLDINGS, INC. (US) |
    优先权:
    申请号 | 申请日 | 专利标题
    EP17305504.7A|EP3399757A1|2017-05-04|2017-05-04|Method and apparatus to encode and decode two-dimension point clouds|
    EP17305504.7|2017-05-04|
    [返回顶部]