BR112018003247B1 - MOTION VECTOR FIELD CODING AND DECODING METHOD, CODING APPARATUS AND DECODING APPARATUS - Google Patents

Abstract

MOTION VECTOR FIELD CODING AND DECODING METHOD, CODING APPARATUS AND DECODING APPARATUS. These are a motion vector field encoding and decoding method, an encoding apparatus and a decoding apparatus. The method in embodiments of the present invention includes: obtaining (101) an original signal from a current motion vector field block, wherein the current motion vector field block is obtained by dividing a current motion vector field in blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining (102) a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required to obtain the prediction signal; calculating (103) a residual prediction signal from the current motion vector field block according to the prediction signal and the original signal, wherein the residual prediction signal is used to indicate a residual between the signal (... ).

Description

FIELD OF TECHNIQUE

[001] The present invention relates to the field of image processing technologies, and in particular to a motion vector field encoding and decoding method, an encoding apparatus and a decoding apparatus.

BACKGROUND

[002] Whereas the International Telegraph Union (in English: international telegraph union, for short, ITU) released the first international video coding standard H.120 in 1984, a video coding technology that developed rapidly and vigorously , and has become an indispensable and important part of modern information technologies. The rapid development of the Internet, wireless communication networks and digital broadcasting networks results in people's increasing demands for multimedia information. Video encoding technology is one of the key technologies for efficient transmission and storage of video information.

[003] Motion vector field compression is an important part of most video encoding solutions. In video coding, a video corresponding to a moving situation includes a series of video frames, and each video frame includes a still image. The series of video frames causes an illusion of movement by displaying consecutive images relatively quickly, for example by displaying consecutive images at a rate of 15 frames at 30 frames per second. Due to a relatively high frame rate, the images in all of the video frame series are very similar. A video frame is selected from the series of video frames as a reference image, and a motion vector field of another video frame in the series of video frames indicates information about the displacement of the video frame with respect to the reference image. It should be noted that the video frame may be an image adjacent to the reference image, or it may not be an image adjacent to the reference image.

[004] Specifically, a video frame includes multiple pixels. An image in the video frame can be divided into multiple image units. Each image unit includes at least one pixel, and motion vectors of all pixels in each image unit are the same, that is, one image unit has one motion vector. A video frame motion vector field includes motion vectors of all picture units.

[005] However, there is currently no method to effectively compress a motion vector field.

SUMMARY

[006] Embodiments of the present invention provide a motion vector field encoding and decoding method, an encoding apparatus and a decoding apparatus, so as to improve the efficiency of motion vector field compression.

[007] A first aspect of embodiments of the present invention provides a video encoding method that includes: obtaining an original signal from a current motion vector field block, wherein the current motion vector field block is obtained by dividing -if a current motion vector field is in blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required to obtain the prediction signal; calculate a residual prediction signal from the current motion vector field block according to the prediction signal and the original signal, where the residual prediction signal is used to indicate a residual between the original signal and the prediction signal ; and recording the prediction information and the residual prediction signal into a bit stream.

[008] Referring to the first aspect, in a first possible implementation of the first aspect, obtaining a prediction signal of the current motion vector field block and the prediction information of the current motion vector field block includes : get a first reference motion vector field from the current motion vector field block, where the first reference motion vector field is a encoded and reconstructed motion vector field, the first motion vector field reference is a motion vector field corresponding to a video frame at time t1 and the video frame at time t1 is a video frame next to the video frame at time t; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used to perform interframe prediction on the video frame corresponding to the first reference motion vector field; obtaining the prediction signal according to the second reference motion vector field, the prediction signal including a motion vector field block of the second reference motion vector field, and a block coordinate range of motion vector field of the second reference motion vector field in the second reference motion vector field is equal to a coordinate range of the current motion vector field block in the current motion vector field; and obtaining the prediction information in accordance with the first reference motion vector field, the prediction information including information used to indicate the first reference motion vector field.

[009] Referring to the first possible deployment of the first aspect, in a second possible deployment of the first aspect, obtaining a second reference motion vector field, according to the first reference motion vector field, the moment t, the moment t1 and a moment t2 includes: obtaining a motion vector of a second sampling point from the second reference motion vector field by calculating according to a formula , being that is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as offset is equal to a location of the second sampling point.

[010] Referring to the first possible deployment of the first aspect, in a third possible deployment of the first aspect, obtaining a second reference motion vector field, according to the first reference motion vector field, the time t, time t1, and time t2 includes: determining a second sampling point from the second reference motion vector field, where a location of the second sampling point is equal to a location to which motion is performed from a location of each of the first two sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as an offset, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[011] Referring to the first possible deployment of the first aspect, in a fourth possible deployment of the first aspect, obtaining a second reference motion vector field, according to the first reference motion vector field, the time t, time t1, and time t2 includes: obtaining at least a second sampling point near a second sampling target point of the second reference motion vector field, where a location to which motion is performed from a location of any first sampling point of the first reference motion vector field as a starting point I according to as displacement is different from a location of the second sampling target point 1, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[012] Referring to the first aspect, in a fifth possible implementation of the first aspect, obtaining a prediction signal of the current motion vector field block and the prediction information of the current motion vector field block includes : Obtain a direction coefficient of the current motion vector field block, where the direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the motion field block current motion vector and a value of a second component of the sampling point motion vector; obtaining a reconstruction value of the first component; calculating the prediction signal according to the first component reconstruction value and the direction coefficient, the prediction signal including a second component predictor; and obtaining the prediction information according to the direction coefficient, the prediction information including the direction coefficient information used to indicate the direction coefficient.

[013] Referring to the fifth possible deployment of the first aspect, in a sixth possible deployment of the first aspect, obtaining a direction coefficient of the current motion vector field block includes: obtaining motion vectors of at least two points sampling of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables of a predefined function, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and function values; and use a coefficient from the predefined function obtained through fitting as the direction coefficient.

[014] Referring to the fifth possible deployment of the first aspect, in a seventh possible deployment of the first aspect, obtaining a direction coefficient of the current motion vector field block includes: obtaining motion vectors of at least two points sampling of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and the function values; obtaining a direction coefficient of at least one encoded motion vector field block that is in the current motion vector field and that is close to the current motion vector field block; using a function coefficient obtained through fitting and the direction coefficient of the at least one encoded motion vector field block as candidate direction coefficients from a set of candidate direction coefficients of the current motion vector field block; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient.

[015] Referring to the fifth possible deployment of the first aspect, in an eighth possible deployment of the first aspect, obtaining a direction coefficient of the current motion vector field block includes: when the direction coefficients of at least two encoded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, use the direction coefficients of the at least two encoded motion vector field blocks as the direction coefficient of the current motion vector field block.

[016] Referring to the fifth possible deployment of the first aspect, in a ninth possible deployment of the first aspect, obtaining a direction coefficient of the current motion vector field block includes: when the direction coefficients of at least two encoded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, and the direction coefficients of the at least two encoded motion vector field blocks are used to indicate a ratio between a first component of a motion vector of a sampling point of the at least two coded motion vector field blocks and a second component of the motion vector of the sampling point, perform the following steps: using the direction coefficients of the at least two coded motion vector field blocks as a candidate direction coefficient from a set of candidate direction coefficients; obtaining a motion vector of at least one sampling point in the original signal; when the at least one sampling point is a sampling point, use a ratio of a first component of the sampling point motion vector to a second component of the sampling point motion vector as a candidate direction coefficient of the set of candidate direction coefficients; or when the at least one sampling point is at least two sampling points, use an average value of ratios between the first motion vector components of the at least two sampling points and the second motion vector components of the at least two sampling points as a candidate direction coefficient from the set of candidate direction coefficients; and obtaining residual candidate prediction signals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients, and using a candidate direction coefficient corresponding to a candidate residual prediction signal with lower signal energy or lower rate distortion as the current motion vector field block direction coefficient.

[017] A second aspect of embodiments of the present invention provides a motion vector field decoding method including: obtaining prediction information and a residual prediction signal from a current motion vector field block, wherein the block of current motion vector field is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining a prediction signal from the current motion vector field block according to the prediction information; and calculating a reconstruction signal of the current motion vector field block in accordance with the prediction signal and the residual prediction signal.

[018] Referring to the second aspect, in a first possible implementation of the second aspect, the prediction information includes information used to indicate a first reference motion vector field of the motion vector field block; and obtaining a prediction signal from the current motion vector field block in accordance with the prediction information comprises: obtaining the first reference motion vector field in accordance with the prediction information, wherein the first motion vector field reference motion vector is a motion vector field of one video frame at time t1; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used for the video frame corresponding to the first reference motion vector field; and obtaining a motion vector field block of the second reference motion vector field, wherein a range of coordinates of the motion vector field block of the second reference motion vector field in the second motion vector field reference is equal to a coordinate range of the current motion vector field block in the current motion vector field, and the prediction signal includes the motion vector field block of the second reference motion vector field.

[019] Referring to the first possible deployment of the second aspect, in a second possible deployment of the second aspect, obtaining a second reference motion vector field according to the first reference motion vector field, the moment t, the moment t1 and a moment t2 includes: obtaining a motion vector of a second sampling point from the second reference motion vector field by calculating according to a formula , being that is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as offset is equal to a location of the second sampling point.

[020] Referring to the first possible deployment of the second aspect, in a third possible deployment of the second aspect, obtaining a second reference motion vector field according to the first reference motion vector field, the moment t, time t1, and time t2 includes: determining a second sampling point from the second reference motion vector field, where a location of the second sampling point is equal to a location to which motion is performed from one location of each of the at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as an offset, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[021] Referring to the first possible deployment of the second aspect, in a fourth possible deployment of the second aspect, obtaining a second reference motion vector field according to the first reference motion vector field, the moment t, time t1, and time t2 includes: obtaining at least a second sampling point near a second sampling target point of the second reference motion vector field, where a location to which motion is performed from a location of any first sampling point of the first reference motion vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[022] Referring to the second aspect, in a fifth possible implementation of the second aspect, the prediction information includes direction coefficient information used to indicate a direction coefficient of the current motion vector field block, and the coefficient of direction is used to indicate the relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the motion vector of the sampling point; and obtaining a prediction signal of the current motion vector field block in accordance with the prediction information comprises: obtaining a reconstruction value of the first component and calculating a predictor of the second component according to the direction coefficient and the reconstruction value of the first component, wherein the predictor signal includes the predictor of the second component.

[023] Referring to the fifth possible deployment of the second aspect, in a sixth possible deployment of the second aspect, the direction coefficient information includes information used to indicate a reconstructed motion vector field block in the current motion vector field , and the direction coefficient includes a direction coefficient of the reconstructed motion vector field block; or the bearing coefficient information includes a bearing coefficient value.

[024] A third aspect of embodiments of the present invention provides a coding apparatus including: a first acquisition module configured to obtain an original signal from a current motion vector field block, wherein the current motion vector field block current motion is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; a second acquisition module configured to obtain a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required to obtain the forecast signal; a calculation module configured to calculate a residual prediction signal of the current motion vector field block according to the prediction signal obtained by the second acquisition module and the original signal obtained by the first acquisition module, wherein the residual signal prediction is used to indicate a residual between the original signal and the prediction signal; and an encoding module configured to record, in a bit stream, the prediction information obtained by the second acquisition module and the residual prediction signal obtained by means of calculation by the calculation module.

[025] Referring to the third aspect, in a first possible implementation of the third aspect, the second acquisition module is configured to: obtain a first reference motion vector field from the current motion vector field block, where the first reference motion vector field is a encoded and reconstructed motion vector field, the first reference motion vector field is a motion vector field corresponding to a video frame at time t1, and the frame of video at time t1 is a video frame close to the video frame at time t; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used to perform interframe prediction on the video frame corresponding to the first reference motion vector field; obtaining the prediction signal according to the second reference motion vector field, wherein the prediction signal comprises a motion vector field block of the second reference motion vector field, and a block coordinate range of motion vector field of the second reference motion vector field in the second reference motion vector field is equal to a coordinate range of the current motion vector field block in the current motion vector field; and obtaining the predictive information in accordance with the first reference motion vector field, wherein the predictive information comprises information used to indicate the first reference motion vector field.

[026] Referring to the first possible deployment of the third aspect, in a second possible deployment of the third aspect, the second acquisition module is configured to obtain a motion vector of a second sampling point from the second motion vector field of reference by means of calculation according to a formula , being that is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as displacement is the same as a location of the second sampling point.

[027] Referring to the first possible deployment of the third aspect, in a third possible deployment of the third aspect, the second acquisition module is configured to: determine a second sampling point of the second reference motion vector field, where a location of the second sampling point is the same as a location to which movement is performed from a location of each of at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as displacement, and the motion vector of each of the first two sampling points is a product of a motion vector of each among at least the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[028] Referring to the first possible deployment of the third aspect, in a fourth possible deployment of the third aspect, the second acquisition module is configured to: obtain at least a second sampling point close to a second target sampling point of the second reference motion vector field, wherein a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[029] Referring to the third aspect, in a fifth possible implementation of the third aspect, the second obtaining module is configured to: obtain a direction coefficient of the current motion vector field block, where the direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the motion vector of the sampling point; obtaining a reconstruction value of the first component; calculating the prediction signal according to the first component reconstruction value and the direction coefficient, the prediction signal including a second component predictor; and obtaining the prediction information according to the direction coefficient, the prediction information including the direction coefficient information used to indicate the direction coefficient.

[030] Referring to the fifth possible deployment of the third aspect, in a sixth possible deployment of the third aspect, the second acquisition module is configured to: obtain motion vectors from at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables of a predefined function, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and function values; and use a coefficient from the predefined function obtained through fitting as the direction coefficient.

[031] Referring to the fifth possible deployment of the third aspect, in a seventh possible deployment of the third aspect, the second acquisition module is configured to: obtain motion vectors from at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and the values of function; and obtaining a direction coefficient of at least one encoded motion vector field block that is in the current motion vector field and that is close to the current motion vector field block; using a function coefficient obtained through fitting and the direction coefficient of the at least one encoded motion vector field block as candidate direction coefficients from a set of candidate direction coefficients of the current motion vector field block; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient.

[032] Referring to the fifth possible deployment of the third aspect, in an eighth possible deployment of the third aspect, the second acquisition module is set to: when the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and which are adjacent to the current motion vector field block are equal, use the direction coefficients of the at least two coded motion vector field blocks as the direction coefficient of the field block drive vector icon.

[033] Referring to the fifth possible deployment of the third aspect, in a ninth possible deployment of the third aspect, the second acquisition module is set to: when the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and which are adjacent to the current motion vector field block are equal, and the direction coefficients of the at least two encoded motion vector field blocks are used to indicate a ratio between a first component of a motion vector of a sampling point of the at least two coded motion vector field blocks and a second component of the motion vector of the sampling point, perform the following steps: use the direction coefficients of the at least two motion vector field blocks encoded as a candidate direction coefficient from a set of candidate direction coefficients; obtaining a motion vector of at least one sampling point in the original signal; when the at least one sampling point is a sampling point, use a ratio of a first component of the sampling point motion vector to a second component of the sampling point motion vector as a candidate direction coefficient of the set of candidate direction coefficients; or when the at least one sampling point is at least two sampling points, use an average value of ratios between the first motion vector components of the at least two sampling points and the second motion vector components of the at least two sampling points as a candidate direction coefficient from the set of candidate direction coefficients; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient.

[034] A fourth aspect of embodiments of the present invention provides a decoding apparatus, including: a first acquisition module, configured to obtain prediction information and a residual prediction signal from a current motion vector field block, in that the current motion vector field block is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at a time t ; a second obtaining module configured to obtain a prediction signal from the current motion vector field block in accordance with the prediction information; and a calculation module configured to calculate a current motion vector field block reconstruction signal in accordance with the prediction signal obtained by the second acquisition module and the residual prediction signal obtained by the first acquisition module.

[035] Referring to the fourth aspect, in a first possible implementation of the fourth aspect, the prediction information includes information used to indicate a first reference motion vector field of the motion vector field block; and the second get module is configured to: get the first reference motion vector field according to the prediction information, wherein the first reference motion vector field is a motion vector field from a frame of video at a time t1; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used for the video frame corresponding to the first reference motion vector field; and obtaining a motion vector field block of the second reference motion vector field, wherein a range of coordinates of the motion vector field block of the second reference motion vector field in the second motion vector field reference is equal to a range of coordinates of the current motion vector field block in the current motion vector field, and the prediction signal comprises the motion vector field block of the second reference motion vector field.

[036] Referring to the first possible deployment of the fourth aspect, in a second possible deployment of the fourth aspect, the second acquisition module is configured to: obtain a motion vector from a second sampling point of the second motion vector field reference by means of calculation according to a formula , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as offset is equal to a location of the second sampling point.

[037] Referring to the first possible deployment of the fourth aspect, in a third possible deployment of the fourth aspect, the second acquisition module is configured to: determine a second sampling point of the second reference motion vector field, where a location of the second sampling point is equal to a location to which movement is performed from a location of each of at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as displacement, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the at least minus first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[038] Referring to the first possible deployment of the\ fourth aspect, in a fourth possible deployment of the fourth aspect, the second acquisition module is configured to: obtain at least a second sampling point close to a second sampling target point of the second reference motion vector field, wherein a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[039] Referring to the fourth aspect, in a fifth possible implementation of the fourth aspect, the prediction information includes direction coefficient information used to indicate a direction coefficient of the current motion vector field block, and the coefficient of direction is used to indicate the relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the motion vector of the sampling point; and the second obtaining module is configured to obtain a reconstruction value of the first component, and calculate a predictor of the second component according to the direction coefficient and the reconstruction value of the first component, wherein the prediction signal includes the predictor of the second component.

[040] Referring to the fifth possible deployment of the fourth aspect, in a sixth possible deployment of the fourth aspect, the direction coefficient information includes information used to indicate a reconstructed motion vector field block in the current motion vector field , and the direction coefficient includes a direction coefficient of the reconstructed motion vector field block; or the bearing coefficient information includes a bearing coefficient value.

[041] It can be learned from the above technical solutions that the embodiments of the present invention have the following advantages:

[042] In the embodiments of the present invention, during coding of the current block, the original signal of the current motion vector field block does not need to be encoded; instead, the prediction information and the residual prediction signal are encoded. Therefore, the efficiency of motion vector field compression is improved.

BRIEF DESCRIPTION OF THE DRAWINGS

[043] Figure 1 is a flowchart of an embodiment of a motion vector field encoding method, according to the present invention.

[044] Figure 2 is a schematic diagram of continuous division and distinct division.

[045] Figure 3 is a flowchart of a method of obtaining a prediction signal of the current block in the mode shown in Figure 1.

[046] Figure 4 is a flowchart of an embodiment of determining a weight of a motion vector of each first sampling point in a first set in the embodiment shown in Figure 3.

[047] Figure 5 is a flowchart of a method of obtaining a prediction signal of the current block in the mode shown in Figure 1.

[048] Figure 6 is a flowchart of a method of obtaining a direction coefficient of a current block in the mode shown in Figure 5.

[049] Figure 7 is a flowchart of another method of obtaining a direction coefficient of a current block in the mode shown in Figure 5.

[050] Figure 8 is a flowchart of another method of obtaining a direction coefficient of a current block in the mode shown in Figure 5.

[051] Figure 9 is a flowchart of an embodiment of a motion vector field decoding method according to the present invention.

[052] Figure 10 is a schematic structural diagram of one embodiment of an encoding apparatus according to the present invention.

[053] Figures 11 is a schematic structural diagram of one embodiment of a decoding apparatus according to the present invention.

[054] Figure 12 is a structural block diagram of another embodiment of an encoding apparatus, according to the present invention.

[055] Figures 13 is a structural block diagram of another embodiment of a decoding apparatus, according to the present invention.

DESCRIPTION OF MODALITIES

[056] To make persons skilled in the art better understand the solutions in the present invention, the following clearly and completely describes the technical solutions in the embodiments of the present invention with reference to the attached drawings in the embodiments of the present invention. Apparently, the described embodiments are merely a part rather than all embodiments of the present invention. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[057] The following separately provides detailed descriptions.

[058] The terms "first", "second", "third", "fourth" and the like in the specification, claims and accompanying drawings of the present invention are intended to distinguish between different objects, but do not indicate a specific sequence. Furthermore, the terms “includes”, “has” and any other variant thereof are intended to encompass non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes a step or unit that is not listed, or optionally includes plus another step or unit that is inherent to the process, method, product or device.

[059] The following first describes a motion vector field encoding method provided in embodiments of the present invention. The motion vector field encoding method provided in the embodiments of the present invention is performed by an encoding apparatus. The encoding device can be any device that needs to output or store a video, for example, a device such as a notebook computer, a tablet computer, a personal computer, a mobile phone, or a video server.

[060] In an embodiment of the motion vector field encoding method in the present invention, wherein a motion vector field encoding method includes: obtaining an original signal from a current motion vector field block, in that the current motion vector field block is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at a time t ; obtaining a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required to obtain the prediction signal; calculating a prediction residual signal from the current motion vector field block according to the prediction signal and the original signal, wherein the prediction residual signal is used to indicate a residual between the original signal and the prediction signal; and recording the prediction information and the residual prediction signal into a bit stream.

[061] First, referring to Figure 1, Figure 1 is a schematic flowchart of a motion vector field encoding method provided in an embodiment of the present invention. As shown in Figure 1, the motion vector field encoding method provided in this embodiment of the present invention can include the following content.

[062] 101: Get an original signal from a current motion vector field block.

[063] A motion vector field is information about the movement of an image with respect to another image. In video compression, a motion vector field is used to describe motion information of a target video frame relative to a video reference frame of the target video frame. The target video frame includes multiple image blocks, and each image block has a corresponding compatible block in the video reference frame. There is a one-to-one correspondence between sample points in the motion vector field and image blocks in the target video frame. A value of each sampling point is a motion vector of a picture block corresponding to the sampling point, and the motion vector is information about the displacement of the picture block with respect to a compatible block of the picture block in the sampling frame. video reference.

[064] In this mode, during compression of a motion vector corresponding to a video frame at time t, the motion vector field is divided into different motion vector field blocks, and compression encoding is performed in the motion vector field by performing compression encoding on each motion vector field block. A motion vector field block includes at least one sampling point. In the following, for ease of description, a motion vector field block to be current compressed is called a current block, and a motion vector field in which the current block is located is called a current field.

[065] It should be noted that when a motion vector field is divided into different motion vector field blocks, the motion vector field blocks are not necessarily corresponding to image blocks obtained by dividing a frame video corresponding to the motion vector field. For a method of dividing a motion vector field into blocks, see Method of dividing a video frame into blocks. This is not limited in this document.

[066] 102: Get a prediction signal from the current motion vector field block and prediction information from the current motion vector field block.

[067] There are multiple methods to obtain the prediction signal. For example, 35 interframe preview modes of a video frame are provided in a current standard HEVC. The 35 interframe prediction modes include 33 directional prediction modes, an Intra_DC mode and an Intra_Planar mode. In this embodiment, while getting the prediction signal from the motion vector field block, the interframe prediction mode is applied to the interframe prediction of a current field.

[068] Specifically, an interframe prediction mode is determined, and at least one reconstructed encoded motion vector field block that is in the current field and that is close to the current block is used as a motion vector field block current block reference; the current block prediction signal is obtained according to the interframe prediction mode and the reference motion vector field block.

[069] The motion vector field block next to the current block can be a motion vector field block adjacent to the motion vector field block (ie, contiguous to the motion vector field block), or can be a motion vector field block separated from the current block by a predefined number of motion vector field blocks. This is not limited in this document. In the current application, all motion vector field blocks in the current field are coded in a left-to-right, top-to-bottom sequence. Therefore, in general, an adjacent motion vector field block that is encoded and reconstructed and that is to the left of the current block, to the lower left of the current block, above the current block, or to the upper right of the previous block is used as the reference move vector field block.

[070] There are multiple methods to calculate the prediction signal according to a prediction mode and a reference motion vector field block. For example, the obtained interframe prediction mode is a horizontal prediction mode in the 33 directional prediction modes. In this case, the current block's reference motion vector field block is a motion vector field block that is located on the same row as the current block and which is the first motion vector field block to the left of the block current. A reference motion vector field block reconstruction signal is used as the current block prediction signal.

[071] As another example, the obtained interframe prediction mode is the Intra_DC mode. After the reference motion vector field block of the current block is obtained, an average value of reconstruction pixels of the reference motion vector field block is used as the prediction signal of the current block.

[072] Then, correspondingly, the prediction information is an interframe prediction mode index and a reference motion vector field block index.

[073] Alternatively, the forecast information may not include an index of the reference motion vector field block. An encoding apparatus and a decoding apparatus have predetermined a location, with respect to the current block, of a reference motion vector field block which is corresponding to each interframe prediction mode and which is used for the current block. In that embodiment, the encoding apparatus and the decoding apparatus further predetermine a prediction signal calculation method corresponding to each interframe prediction mode. Thus, after receiving the prediction information, the decoding apparatus calculates the prediction signal according to the prediction mode indices and the reference motion vector field block in the prediction information using the calculation method default.

[074] In the current application, while obtaining the prediction signal of the current block, an interframe prediction mode can be directly determined. The current block's prediction signal is calculated using the interframe prediction mode, and an interframe prediction mode index is added to the prediction information. Alternatively, all interframe prediction modes can be cycled through, the current block prediction signal is calculated using all interframe prediction modes, and an index of an interframe prediction mode corresponding to a prediction signal whose energy has a Minor difference of the original signal energy of the current block (ie a residual prediction signal mentioned below) is added to the prediction information. The subsequent forecast residual signal is calculated using the forecast signal.

[075] 103: Calculate a residual prediction signal from the current motion vector field block according to the prediction signal and the original signal.

[076] The residual prediction signal is used to indicate a difference between the original signal and the prediction signal. The prediction signal is a prediction signal obtained according to the prediction information. After the current block prediction signal is obtained, the difference between the original signal and the current block prediction signal is calculated, so as to obtain the residual current block prediction signal.

[077] The method for calculating the difference between the original signal and the prediction signal of the current block belongs to the prior art, and the details are not described in this document.

[078] 104: Write the prediction information and the residual prediction signal into a bit stream.

[079] The decoding apparatus can obtain the prediction signal of the current block by performing the prediction in the current block according to the prediction information, and then it can calculate an original signal of the current field with reference to the residual signal of prediction of the current block. Therefore, when performing compression encoding on the current block, the encoding apparatus only needs to write the prediction information and the residual prediction signal into the bitstream and send the bitstream to the decoding apparatus, so that the apparatus decoding can obtain the original signal of the current block.

[080] In this mode, to encode the residual prediction signal, refer to the encoding of a residual prediction signal of a video frame in a standard video. In fact, during encoding of the predictive residual signal, the predictive residual signal is first compressed, and then the compressed data is written into the bit stream.

[081] The compression of the predictive residual signal is generally classified into lossy compression and lossless compression. Lossless compression means: After compression, a reconstructed motion vector field signal is exactly the same as an original signal and there is no loss of information. Lossy compression means: After compression, a reconstructed motion vector field signal is not exactly the same as an original signal and there is a loss of specific information. A lossless compression process can include transformation and entropy encoding. A lossy compression process can include transformation, quantization, and entropy encoding. This belongs to the prior art, and details are not described in the present document.

[082] In this mode, during the coding of the current block, the original signal of the current block does not need to be codified; instead, the prediction information and the residual prediction signal are encoded. Therefore, an amount of information that needs to be encoded is greatly reduced, and the efficiency of motion vector field encoding is improved. Furthermore, because the prediction information is required information to obtain the prediction signal of the current block, the prediction signal that can be obtained by the decoding apparatus according to the prediction information is the same as the prediction signal. prediction obtained by the coding apparatus. Therefore, the decoding apparatus can reconstruct, according to the prediction information and the residual prediction signal, a signal which is highly similar to the original signal of the current block.

[083] In this modality, there are multiple types of forecast information. Correspondingly, there are multiple methods to obtain the prediction signal of the current block according to the prediction information. The following describes several of the methods using examples.

[084] In a first example, an intrafield prediction method can be used, that is, the prediction is performed in the current block according to the current field. Due to the fact that there is a specific correlation between the current block and a motion vector field that is spatially adjacent to the current block, prediction can be performed on the current block according to the intrafield prediction method.

[085] In the current application, there are multiple methods of intra-field forecasting. For example, intra-field prediction methods include angle prediction and intra-frame area division prediction. In an angle prediction method, the prediction signal of the current block can be obtained according to the interframe prediction mode and the reference motion vector field block of the current block, as described in the explanation of step 102.

[086] An intraframe area division prediction method is specifically as follows: The current block is divided into at least two areas, and in each area, a motion vector is obtained as an area prediction signal. Therefore, in the intraframe area division prediction method, the prediction information includes information about an area division method and information about a method for determining a prediction signal of each area. Area division method information is used to indicate the area division method, for example, it is an area division method index. Information about the method for determining the predictive signal of each area is used to indicate the method for determining the predictive signal of each area, for example, it is an index of the method for determining the predictive signal of each area .

[087] Alternatively, the prediction information may not include information about a method for determining a prediction signal for each area, however, the encoding apparatus and the decoding apparatus store the same predefined method for determining the prediction signal for each area area. This is not limited in this document.

[088] In this modality, there are multiple methods to determine the forecast signal of each area. For example, an average value of motion vectors from all sample points in an area can be used as a prediction signal for the area, a motion vector from one of the sample points in the area can be used as a area forecast, or another method can be used. This is not limited in this document. When a motion vector of one of the sampling points in an area is used as an area prediction signal, all sampling points in the area can be traversed once, and a motion vector of a sampling point that makes with which the energy of a predictive residual signal to be the smallest is determined as the predictive signal.

[089] In this mode, there are multiple methods to divide the current block. For example, the current block is first divided into two areas, and then at least one of the two areas is divided into two areas, and so on. Area division is terminated when a predefined condition is met.

[090] Preset condition has multiple types of settings. For example, a skew value of the current block can be calculated after each split, where the skew value is a maximum value of differences between predictor signals and original signals from all areas in the current block after splitting. When the distortion value is less than a predefined value, the predefined condition is fulfilled. Of course, the above is merely a description of the predefined condition with the use of an example and does not constitute any limitation.

[091] As another example, a preset value can be predetermined, and a number of areas into which the current block is divided is calculated after each division. When the quantity reaches the preset value, the preset condition is met.

[092] When an area is divided into two areas, one or two of the areas can be in the shape of a trapezoid or a triangle. Of course, an area shape can alternatively have another shape. This is not limited in this document.

[093] Optionally, in this mode, during division of each area, all division methods of the area are cycled through, the forecast signals of two subareas in the area when each division method is used are calculated, and an optimal division method is determined from the division methods according to the forecast signals of the two subareas. Specifically, the optimal splitting method can be determined using a rate distortion optimization principle.

[094] In this mode, when the current block is divided into different areas, continuous division or distinct division can be performed in the current block. The following separately explains continuous division and distinct division with respect to Figure 2.

[095] As shown in a left-hand figure in Figure 2, continuous division means that an actual block S is directly divided into two areas, P1 and P2, using a straight line L. Distinct division means that the block current S is made up of multiple pixel blocks, and during division, the current block is divided along the edges of pixel blocks in a current field, to divide the current block S into two areas: P1 and P2.

[096] Alternatively, when the current block is divided into different areas, contour division can be performed on the current block. That is, the current block is divided according to an outline of an object represented by an image in the current block. This is not limited in this document.

[097] In a second example, a cross-field prediction method can be used. That is, prediction is performed on the current block using a reference motion vector field and according to a correlation between the current block and a motion vector field that is temporally close to the current block. The reference motion vector field is another motion vector field that is temporarily close to the current field.

[098] It should be noted that the motion vector field that is next to the current field may be a motion vector field adjacent to the current field (i.e., a motion vector field of a video frame that temporarily follows or precedes a video frame corresponding to the current field), or may be separated from the current field by at least one motion vector field (that is, it may be a motion vector field corresponding to a video frame that is temporarily separated , for at least one video frame, of a video frame corresponding to the current field). The following describes two embodiments of the cross-field prediction method with the use of examples.

[099] Modality 1: Due to the fact that there is a specific coherence of object movement in time, there is a specific correlation between the vector fields of movement in time. That is, at least a part of the current block appears in the reference motion vector field. However, a current block location in the current field is not necessarily the same as a current block location in the reference motion vector field.

[100] For the sake of clear description, in this specification, a location of a motion vector field block in a motion vector field is a range of coordinates of the motion vector field block in the motion vector field. Therefore, when a location of a motion vector field block p in a motion vector field P is the same as a location of a motion vector field block q in a motion vector field Q, it means that a coordinate range of p motion vector field block in P motion vector field is the same as a coordinate range of q motion vector field block in Q motion vector field.

[101] Therefore, in this embodiment, obtaining a prediction signal of the current block specifically includes: determining the reference motion vector field of the current block, searching for a compatible block of the current block in the reference motion vector field , and use a compatible block reconstruction signal as the current block prediction signal. There are multiple methods to search for compatible blocks. For example, all motion vector field blocks in the reference motion vector field can be traversed, the differences between the motion vector field blocks and the current block are calculated, and a motion vector field block movement with a minor difference from the current block is used as the compatible block of the current block.

[102] Obtaining prediction information of the current block specifically includes: using information about the reference motion vector field and information about the compatible block as the prediction information, where information about the motion vector field of reference is used to indicate the reference motion vector field, and information about the compatible block is used to indicate the compatible block. Specifically, the reference motion vector field information may be an index of the reference motion vector field, and the compatible block information may be offset information from a location of the compatible block relative to a locating a first motion vector field block in the reference motion vector field. The location of the first motion vector field block in the reference motion vector field is the same as the location of the current block in the current field. Alternatively, the compatible block information may be an index of the compatible block. This is not limited in this document.

[103] Modality 2: In this modality, it is assumed that a state of movement of the same object that is filmed in a video sequence remains unchanged, that is, a direction and magnitude of movement do not change, in a relatively short time . In this case, the prediction signal of the current block can be deduced according to the reference motion vector field.

[104] Specifically, in this modality, as shown in Figure 3, obtaining a prediction signal of the current block specifically includes the following steps:

[105] S11: Get a first reference motion vector field from the current motion vector field block.

[106] The first reference motion vector field is a motion vector field encoded and reconstructed next to the current field. For ease of description, the video frame corresponding to the current field is called a video frame at time t, and a video frame corresponding to the first reference motion vector field is called a video frame at time t1. Time t1 can be before time t, or it can be after time t. This is not limited in this document.

[107] S12: Obtain a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2.

[108] For ease of understanding, below, the motion vector fields and video frames to be described are explained first: A location of a target object in the video frame at time t1 is A, a frame of video reference used to perform interframe prediction in the video frame is a video frame at time t2, and a location of the target object in the video frame at time t2 is B. In this case, a motion vector from a first sampling point corresponding to the reference is used to indicate the displacement from location B to location A.

[109] It is assumed that a state of motion (including a speed and a direction) of the target object remains unchanged, i.e. the corresponding displacement of the target object within a time from tl to t2 is . In this case, it can be deduced that the displacement of the target object within a time from t to tl must be . That is, it is assumed that a location of the target object in the video frame at time t is C, in which case the offset from location A to location C must be

[110] According to the above method, each sampling point in the first reference motion vector field is considered as a target object, and a location to which each sampling point is moved at time t can be deduced. Each sampling point in the first reference motion vector field is moved. The offset between a location of each sampling point after the sampling point is moved and a sampling point location before the sampling point is moved is , a vector of moving the sampling point before the sampling point is moved is , and the sampling point motion vector changes to after the sampling point is moved. For the sake of ease of description, a new motion vector field that is formed after all sampling points in the first reference motion vector field are moved according to the above rule and vectors of all sampling points change is called the second reference motion vector field.

[111] Therefore, during obtaining the second reference motion vector field, a motion vector of a second sampling point of the second reference motion vector field is obtained by calculating according to a formula , being that is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as displacement is the same as a location of the second sampling point.

[112] In this embodiment, when all sampling points in the first reference motion vector field are moved so as to obtain the second reference motion vector field, at least two sampling points in the first motion vector field reference points will be moved to the same location in the motion vector field at time t.

[113] In other words, the current speeds and directions of all sampling points in the first reference motion vector field remain unchanged, and in the motion vector field formed at time t (i.e., the second motion vector field (reference motion vector field), at least two sampling points in the first reference motion vector field are possibly moved to the same location. There are multiple ways to determine a value of a sampling point at the location in the second reference motion vector field. For example, a product of a motion vector of one of the sampling points and (t-t1)/(t1-t2) can be used as the motion vector of the sampling point in the location.

[114] Alternatively, preferably, in this case, a motion vector at the location is determined using the following method:

[115] A second sampling point of the second reference motion vector field is determined, wherein a location of the second sampling point is equal to a location to which motion is performed from one location of each of at least first two sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the at least two first sampling points as displacement, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2).

[116] The product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) is used as a motion vector of the second sampling point. To facilitate the description below, the second sampling point is referred to as the second target sampling point, and the first two sampling points of the first reference motion vector field are referred to as the first set.

[117] When determining the motion vector weights of all first sample points in the first set, the motion vector weights of all first sample points in the first set can be preset equal, i.e. an average value of the motion vectors of all the first sampling points in the first set is used as a motion vector for the second sampling target point.

[118] Alternatively, as shown in Figure 4, determining the motion vector weights of all the first sampling points in the first set specifically includes the following steps:

[119] S21: Obtain a motion vector of at least one second sampling point that is in the second reference motion vector field and that is located around the second sampling target point.

[120] For ease of description, the at least one second sampling point located around the second target sampling point is called the second set, and each second sampling point is an element of the second set.

[121] Optionally, the second set may include at least one of four second sampling points that are located around the second sampling target point and that are adjacent to the second sampling target point.

[122] S22: Compute a degree of similarity between each of the motion vectors of the first sample points in the first set and a motion vector from the second set.

[123] For each first sampling point in the first set, a degree of similarity between a motion vector of the first sampling point and a motion vector of each sampling point in the second set is calculated. There are multiple methods for calculating degree of similarity. For example, the differences between the motion vector of the first sampling point and the motion vectors of all elements in the second set can be calculated, and then a sum or average value of all the differences is used as the degree of similarity between the motion vector of the first sampling point and the motion vector of the second set. In this case, a smaller sum or average value of the differences indicates a greater degree of similarity.

[124] Of course, the above descriptions are only an example and do not constitute any limitation.

[125] S23: Determine a weight of a motion vector of each first sampling point in the first set according to the degree of similarity, where a greater degree of similarity between a motion vector of a first sampling point and the second set motion vector indicates that the first sample point has a larger motion vector weight.

[126] After the degree of similarity corresponding to the motion vector of each element in the first set is determined, the weight of the motion vector of each element in the first set is determined according to a magnitude of the degree of similarity. An element with a higher degree of similarity has a higher motion vector weight. Specifically, weights corresponding to different classifications can be predefined. After a rank of one degree of similarity of each element in the first set is determined, a weight corresponding to the rank of the element is used as a weight of a move vector of the element.

[127] The foregoing describes a case in which at least two first sampling points in the first reference motion vector field are moved to the same location in the reference motion vector field at time t.

[128] In a real application possibly also there is a special location in the second reference motion vector field, and no sampling point the first reference motion vector field is moved to the special location at time t.

[129] In that case, at least the second sampling point close to a second sampling target point of the second reference motion vector field is obtained, where a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; It is

[130] When the at least one second sampling point is a second sampling point, an average value of the second sampling point is used as a motion vector of the second sampling target point.

[131] When the at least one second sampling point is at least two second sampling points, a weighted average value of motion vectors of the at least two second sampling points is used as a motion vector of the second target point of sampling.

[132] For ease of description, the at least one second sampling point is called the third set below.

[133] When determining the motion vector weights of all second sample points in the third set, the motion vector weights of all second sample points in the third set can be preset to be equal, i.e., a value average of the motion vectors of all second sampling points in the third set is used as the motion vector of the second sampling target point.

[134] Alternatively, determining the motion vector weights of all second sampling points in the third set specifically includes: obtaining a distance between each second sampling point in the third set and the second target sampling point, determining a weight of every second sampling point in the third set according to distance, where a second sampling point at a distance from the second sampling target point has a greater motion vector weight.

[135] Of course, when there is only a second sampling point in the third set, a weighted average value of a motion vector of the second sampling point in the third set is the motion vector of the second sampling point.

[136] Specifically, for example, a distance between a second left sampling point (whose motion vector is ) closest to the second sampling point and the second sampling point is m, and a distance between a second right sampling point (whose motion vector is ) closest to the second target sampling point and the second target sampling point is n. In this case, the motion vector of the target motion vector is Of course, the above descriptions are only an example and do not constitute any limitation.

[137] S13: Obtaining the prediction signal according to the second reference motion vector field, wherein the prediction signal includes a motion vector field block of the second reference motion vector field.

[138] A motion vector field block coordinate range of the second reference motion vector field in the second reference motion vector field is the same as a coordinate range of the current motion vector field block in the current move vector field.

[139] Because the second reference motion vector field is a predictive motion vector field at time t, the second reference motion vector field is a predictive signal of the current field. A prediction method used for a different block in the current field is not necessarily the method described in this embodiment. Therefore, when the method described in this embodiment is used for the current block, a signal from a motion vector field block in an area in which the current block is located in the second reference motion vector field is used as the signal current block forecast.

[140] S14: Obtaining the prediction information according to the first reference motion vector field, wherein the prediction information comprises information used to indicate the first reference motion vector field.

[141] While getting the current block prediction signal, the current block prediction signal can be determined according to contents of the first reference motion vector field only. Therefore, the decoding apparatus can calculate the prediction signal of the current block using only information about the first reference motion vector field as the prediction information. Information about the first reference motion vector field block is used to indicate the first reference motion vector field. For example, information about the first reference motion vector field is an index of the first reference motion vector field.

[142] In this embodiment, due to the fact that the current block prediction information only includes the information about the first reference motion vector field, a bit quantity required to encode the current block is greatly reduced.

[143] It should be noted that although all of the methods described above are described using the motion vector as an example, the motion vector can alternatively be replaced by a motion vector component in the actual application.

[144] In a third example, an intermotion vector field component prediction method. That is, prediction is performed on one component of a motion vector of each sampling point in the current block according to a direction coefficient of the current block and another component of the motion vector of the sampling point in the current block.

[145] A motion vector includes a direction and a magnitude, and the motion vector can be broken down into two components: a horizontal component and a vertical component. For example, for a motion vector , an included angle between the motion vector and a horizontal direction is θ. In this case, the magnitudes of horizontal and vertical components of the motion vector deducted. In a real application, the encoding apparatus will generally store a motion vector of each sampling point by storing a horizontal component and a vertical component of the motion vector of the sampling point. In this embodiment, a prediction signal of one component of a motion vector is calculated using a magnitude of another component of the motion vector and a ratio of the other component to the component.

[146] Specifically, referring to Figure 5, obtaining a current block prediction signal specifically includes the following steps.

[147] S31: Get a direction coefficient from the current motion vector field block.

[148] A motion vector of a sampling point can be decomposed into a vertical component and a horizontal component. For ease of description, one of the vertical and horizontal components of a motion vector is called the first component, and the other component is called the second component of the motion vector. Due to the fact that the current block includes at least one sampling point, the current block can be decomposed into a first component block and a second component block. The first component block includes a first component from each sample point in the current block, and the second component block includes a second component from each sample point in the current block.

[149] The current block direction coefficient is used to indicate a relationship between a first component value and a second component value for each sampling point. That is, during prediction of the current block, it is assumed that the function relationships between the first components and second components of all sample points in the current block are equal.

[150] S32: Get a rebuild value of the first component.

[151] In this mode, compression of the current block includes compression of the first component block and second component block of the current block. During encoding of the first component block, the first component block can be encoded using the method in the embodiment shown in the Figure. 1 or using the interfield forecasting method or the interfield forecasting method described above. This is not limited in this document.

[152] During coding of the second component block, due to the fact that the function relationships between the first components and second components of all sample points are assumed to be equal, a predictor of each sample point in the second component block can be calculated using a first component of the sampling point and the direction coefficient.

[153] The decoding apparatus computes a second component predictor of each sample point in the current block according to the direction coefficient and a first component reconstruction value of the sample point. Therefore, before calculating the prediction signal of the current block, the encoding apparatus first obtains the direction coefficient and reconstruction value of the first component of each sampling point in the current block. It should be noted that, due to the fact that an amount of direction coefficient information is relatively small, the coding apparatus generally performs lossless coding on the direction coefficient. Therefore, the coding apparatus does not need to obtain a reconstruction value of the direction coefficient, but directly uses an original value of the direction coefficient.

[154] S33: Calculate the prediction signal according to the reconstruction value of the first component and the direction coefficient, and the prediction signal includes a predictor of the second component.

[155] A decoder can obtain the reconstruction value of the first component. Therefore, when the current block prediction signal includes the second component predictor, the decoder can obtain the current block prediction signal according to the second component predictor and the first component reconstruction value.

[156] For each sample point in the current block, the direction coefficient of the current block is used as a direction coefficient of the sample point, i.e. a second component of the sample point is calculated according to a first component of the sample point. sampling point and the direction coefficient of the current block. In this way, a predictor signal from each sampling point in the current block includes a predictor of the second component.

[157] S34: Get the prediction information according to the direction coefficient, where the prediction information includes the direction coefficient information used to indicate the direction coefficient.

[158] In this embodiment, due to the fact that the current block prediction information includes a direction coefficient, the amount of bits required to encode the current block is relatively small.

[159] In the realization shown in Figure 5, there are multiple methods to obtain the current block's direction coefficient. The following describes several of the methods using examples.

[160] In the first example, as shown in Figure. 6, obtaining the current block direction coefficient includes the following steps:

[161] S41: Obtain motion vectors from at least two sampling points of the original signal.

[162] S42: Use first components of the motion vectors of the at least two sampling points as independent variables of a predefined function, use second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, adjust the independent variables and function values, and use a coefficient from the predefined function obtained through adjustment as the direction coefficient.

[163] In this mode, the motion vectors of at least two sampling points in the current block are nested. During the fit, the first components of the motion vectors of all sampling points are used as the independent variables of the predefined function, and the second components of them are used as the function values of the predefined function, to obtain a function relationship between the first components and second components of all sample points in the current block.

[164] Specifically, all sample points can be fitted as a straight line. When all sample points are fitted as a straight line, a predefined straight loop function equation is y=kx, where k does not equal 0, or y=ax+b, where a does not equal 0 .When the predefined function equation is y=kx, k is the direction coefficient of the current block. This indicates that all points in the current block move in a straight line in the same direction. When the predefined function equation is y=ax+b, a and b are direction coefficients of the current block.

[165] Alternatively, all points can be fitted as a curve. In this case, a coefficient in a function equation corresponding to the curve is the direction coefficient of the current block. For example, the predefined function equation corresponding to the curve y=ax2+bx+c. In this case, a, b, and c are the current block's direction coefficients.

[166] In this embodiment, as a part of the prediction information, the current motion vector field direction coefficient information specifically includes a direction coefficient value. Of course, the encoding apparatus and the decoding apparatus need to predefine the function equations corresponding to different amounts of coefficients. In this way, the coding apparatus only needs to record the direction coefficient in the bit stream, so that the decoding apparatus can determine a corresponding function equation according to a quantity of values included in the direction coefficients.

[167] In a second example, as shown in Figure. 7, obtaining the current block direction coefficient includes the following steps:

[168] S51: obtain motion vectors of at least two sampling points of the original signal.

[169] S52: Use the first components of the motion vectors of the at least two sampling points as independent variables, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and function values.

[170] For detailed explanation, refer to a portion of the descriptions of step S42 in the first example. Details will not be described again in this document.

[171] S53: Obtain a direction coefficient of at least one encoded motion vector field block that is in the current motion vector field and that is close to the current motion vector field block.

[172] S54: use a function coefficient obtained through fitting and the direction coefficient of the at least one encoded motion vector field block as candidate direction coefficients from a set of candidate direction coefficients of the motion field block current drive vector.

[173] S55: Obtain candidate residuals prediction signals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients, and use a candidate direction coefficient corresponding to a residual prediction signal candidate with lower signal energy or lower rate distortion as the direction coefficient of the current motion vector field block.

[174] A difference from the first example is that, in this embodiment, after the coefficient of an adjustment function of the sampling points in the current block is calculated, the coefficient of the adjustment function is not used directly as the direction coefficient. of the current block. In all the coded blocks that are located around the current block and that are adjacent to the current block, an image corresponding to the current block and an image corresponding to one of all the coded blocks is possibly the same object. In this case, the direction coefficient of the current block is possibly equal to a direction coefficient of one of the coded blocks.

[175] Therefore, in this embodiment, the direction coefficients of all coded blocks are additionally obtained, and the coefficient of the fit function and the direction coefficients of all coded blocks are used as the candidate direction coefficients of the coefficient set. of leadership candidates of the current block. A candidate prediction signal and a residual candidate prediction signal of the current block that are corresponding to each candidate direction coefficient are calculated. The candidate direction coefficient corresponding to the candidate residual prediction signal with the lowest energy is used as the current block direction coefficient.

[176] In this mode, when the direction coefficient of the current block is the function coefficient, the direction coefficient information will include the direction coefficient value. When the current block direction coefficient is the coded block direction coefficient, the direction coefficient information will be used to indicate the coded block or will include a coded block direction coefficient value. This is not limited in this document.

[177] In a third example, obtaining the direction coefficient of the current block includes: when the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, use the direction coefficients of the at least two encoded motion vector field blocks as the direction coefficient of the current motion vector field block.

[178] When the direction coefficients of at least two coded blocks that are adjacent to the current block are equal, it can be deduced that the images corresponding to the at least two coded blocks belong to the same object. It is additionally assumed that an image corresponding to the current block and the images corresponding to at least two coded blocks belong to the same object. Therefore, it can be directly determined that the direction coefficient of the current block is equal to the direction coefficients of at least two coded blocks.

[179] In this embodiment, the direction coefficient information from the current motion vector field is used specifically to indicate the coded block or includes the direction coefficient value of the coded block. For example, the direction coefficient information includes a coded block index. This is not limited in this document.

[180] In a fourth example, as shown in Figure. 8, obtaining the current block direction coefficient includes the following steps:

[181] S61: When the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, and the coefficients direction of the at least two coded motion vector field blocks are used to indicate a ratio between a first component of a motion vector of a sampling point of the at least two coded motion vector field blocks and a second component of the sampling point motion vector, perform the following steps (S61):

[182] S62: Use the direction coefficients of the at least two encoded motion vector field blocks as a candidate direction coefficient from a set of candidate direction coefficients.

[183] S63: Obtain a motion vector of at least one sampling point in the original signal.

[184] When the at least one sampling point is a sampling point, use a ratio of a first component of the sampling point motion vector to a second component of the sampling point motion vector as a candidate direction coefficient of the set of candidate direction coefficients. when the at least one sampling point is at least two sampling points, use an average value of ratios between the first motion vector components of the at least two sampling points and the second motion vector components of the at least two points sampling as a candidate direction coefficient from the set of candidate direction coefficients.

[185] S64: Obtain candidate residuals prediction signals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients, and use a candidate direction coefficient corresponding to a residual prediction signal candidate with lower signal energy or lower rate distortion as the direction coefficient of the current motion vector field block.

[186] A difference from the third example is that, in this embodiment, when the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and that are adjacent to the motion vector block current motion vector field are the same, the direction coefficients of the at least two coded motion vector field blocks will not be used directly as the direction coefficient of the current block. In coded blocks adjacent to the current block, the images corresponding to all sample points in the current block are possibly the same object. In this case, the direction coefficient of the current block is possibly equal to a direction coefficient of one of the sampling points.

[187] Therefore, in this embodiment, the motion vector of the at least one sampling point in the original signal is additionally obtained, and the coded block direction coefficient and the average value of the ratio between the first component of the hair motion vector at least one sampling point and the second component of the motion vector of the at least one sampling point are used with the candidate direction coefficients from the set of candidate direction coefficients, or the coded block direction coefficient and the ratio of the first motion vector component and the second motion vector component are used as the candidate direction coefficients from the set of candidate direction coefficients.

[188] In this mode, when the direction coefficient of the current block is the function coefficient, the direction coefficient information will include the direction coefficient value. When the current block direction coefficient is the coded block direction coefficient, the direction coefficient information will be used to indicate the coded block or will include a coded block direction coefficient value. This is not limited in this document.

[189] The above describes several methods to obtain the prediction signal and prediction information of the current block. In a real application, during compression encoding of all motion vector field blocks into the current motion vector field, the same method or different methods can be used to obtain prediction signals and prediction information from different blocks of motion. drive vector field.

[190] When different methods are used to obtain prediction signals and prediction information from different blocks of motion vector field, there are multiple methods to determine a method used to obtain the prediction signal and prediction information of the current block . For example, an index corresponding to each obtaining method can be preset in the encoding apparatus and the decoding apparatus. While getting the prediction signal and the prediction information of the current block, the coding apparatus cycles through all the acquisition methods, calculates a residual prediction signal when each acquisition method is used, selects, as the block acquisition method current, a fetch method corresponding to a residual prediction signal with the lowest energy, and adds an index of the fetch method to the prediction information for the current block.

[191] The foregoing describes the motion vector field encoding method in the embodiments of the present invention. In the following, the first motion vector field encoding method provided in embodiments of the present invention is described. The motion vector field decoding method provided in the embodiments of the present invention is performed by a decoding apparatus. The decoding device can be any device that needs to output or play back a video, for example, a device such as a mobile phone, a notebook computer, a tablet computer or a personal computer.

[192] In an embodiment of the motion vector field decoding method in the present invention, a motion vector field decoding method includes: obtaining prediction information and a residual prediction signal from a motion vector field block current motion, where the current motion vector field block is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at a time t; obtaining a prediction signal from the current motion vector field block according to the prediction information; and calculating a reconstruction signal of the current motion vector field block in accordance with the prediction signal and the residual prediction signal.

[193] Referring to Figure 9, Figure 9 is a schematic flowchart of a motion vector field decoding method provided in another embodiment of the present invention. As shown in Figure 9, the motion vector field decoding method provided in the other embodiment of the present invention can include the following content.

[194] 901: Get prediction information and a residual prediction signal from a current motion vector field block.

[195] Upon receiving a video bitstream, a decoding apparatus decodes the video bitstream to recover all the video images in an original video stream. During the decoding of each video frame, the video frame is decoded using a video frame reference frame and a video frame motion vector field.

[196] Therefore, the decoding apparatus needs to decode the reference frame and motion vector field of the video frame first. In this embodiment, a motion vector field currently being decoded is called a current motion vector field. The current motion vector field block is obtained by dividing the current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to one video frame at time t.

[197] During current motion vector field decoding, a current motion vector field reconstruction signal is obtained by sequentially reconstructing all motion vector field blocks in the current motion vector field. During current block decoding, the prediction information and the residual prediction signal of the current block are first obtained from the video bit stream.

[198] 902: Obtain a prediction signal from the current motion vector field block according to the prediction information.

[199] For prediction information that includes different content, a method to get the prediction signal of the current block according to the prediction information is also different.

[200] For example, when the prediction information is an index of an intraframe prediction mode and an index of a reference motion vector field block, obtaining a prediction signal from the motion vector field block current according to the prediction information specifically includes: determining the intraframe prediction mode according to the intraframe prediction mode index; determining the reference motion vector field block in accordance with the index of the reference motion vector field block; and then obtaining the prediction signal of the current block according to the intraframe prediction mode and a signal of the reference motion vector field block.

[201] An encoding apparatus and a decoding apparatus predefine a prediction signal calculation method corresponding to each intraframe prediction mode. Thus, after receiving the prediction information, the decoding apparatus calculates the prediction signal according to the prediction mode indices and the reference motion vector field block in the prediction information using a method of prediction. default calculation.

[202] For example, the intraframe predictive mode obtained is a horizontal predictive mode in 35 intraframe predictive modes of a video frame that are provided in an HEVC standard. In this case, a block reconstruction signal of the current block's reference motion vector field is used as the current block's prediction signal.

[203] As another example, the intraframe prediction mode obtained is an Intra_DC mode in the 35 intraframe prediction modes. In this case, an average value of the reference motion vector field block reconstruction pixels is used as the current block prediction signal.

[204] In a real application, the prediction information obtained by the decoding apparatus may alternatively not include an index of a reference motion vector field block. The decoding apparatus and the encoding apparatus pre-specify a location, relative to the current block, of a reference motion vector field block which corresponds to each intraframe prediction mode and which is used for the current block. In this way, after receiving the prediction information, the decoding apparatus determines a reference motion vector field of the current block according to an intraframe prediction mode in the prediction information.

[205] 903: Calculating a reconstruction signal of the current motion vector field block according to the prediction signal and the residual prediction signal.

[206] The residual prediction signal is used to indicate a difference between an original signal from the current block and the prediction signal. After obtaining the prediction signal of the current block, the decoding apparatus modifies the prediction signal using the residual prediction signal, so as to obtain the reconstruction signal of the current block.

[207] In this embodiment, because the prediction information is required information to obtain the prediction signal of the current block, the prediction signal that can be obtained by the decoding apparatus according to the prediction information is the same as a prediction signal obtained by the coding apparatus. Therefore, the decoding apparatus can reconstruct, according to the prediction information and the residual prediction signal, a signal which is highly similar to the original signal of the current block.

[208] In this embodiment, there are multiple types of forecast information. Correspondingly, there are multiple methods to obtain the prediction signal of the current block according to the prediction information. The following describes several of the methods using examples.

[209] In the first example, the forecast information includes information about an area division method and information about a method for determining a forecast signal for each area.

[210] The area division method information is used to indicate the area division method, for example, it is an index of the area division method. Information about the method for determining the predictive signal of each area is used to indicate the method for determining the predictive signal of each area, for example, it is an index of the method for determining the predictive signal of each area .

[211] Obtaining a prediction signal of the current motion vector field block according to the prediction information includes: determining the area division method according to the area division method information, and dividing the area division method. current block in different areas with the use of area division method; and obtaining, in accordance with the information about the method for determining the predictive signal of each area, the method for determining the predictive signal of each area, and obtaining the predictive signal of each area using the method.

[212] For example, information on the method for determining the prediction signal for each area indicates that the area prediction signal is an average value of motion vectors from all sample points in the area. In this case, while obtaining the area prediction signal, the average value of the motion vectors of all sample points in the area is calculated, and the average value is used as the area prediction signal.

[213] Alternatively, information about the method for determining the prediction signal for each area indicates that the area prediction signal is a vector of motion of a sampling point in the area. In this case, while obtaining the area prediction signal, the sampling point motion vector is obtained according to a sampling point index, and the sampling point motion vector is used as the sampling point prediction signal. area.

[214] Of course, the above descriptions are examples only and do not constitute any limitation in this document.

[215] Alternatively, the prediction information may not include information about a method for determining a prediction signal for each area, but the encoding apparatus and the decoding apparatus store the same predefined method for determining the prediction signal for each area. area. This is not limited in this document.

[216] In this mode, there are multiple methods to split the current block. For more details, see example descriptions of the method for splitting the current block in the descriptions of the first example of obtaining the prediction information and the prediction signal in the motion vector field encoding method. Details will not be described again in this document.

[217] In a second example, the prediction information includes information about a reference motion vector field and information about a compatible block, where information about the reference motion vector field is used to indicate the field of reference. reference move vector, and the compatible block information is used to indicate the compatible block.

[218] Specifically, the reference motion vector field information may be an index of the reference motion vector field, and the compatible block information may be offset information from a location of the compatible block in relative to a location of a first motion vector field block in the reference motion vector field. The location of the first motion vector field block in the reference motion vector field is the same as the location of the current block in the current field. Alternatively, the compatible block information may be an index of the compatible block. This is not limited in this document.

[219] Obtaining a prediction signal from the current motion vector field block according to the prediction information includes: determining the reference motion vector field according to the information about the motion vector field from reference; fetch the compatible block in the reference motion vector field according to the information about the compatible block; and use a compatible block reconstruction signal as the current block prediction signal.

[220] In a third example, the prediction information includes the information used to indicate a first reference motion vector field from the motion vector field block.

[221] Obtaining a prediction signal from the current motion vector field block according to the prediction information includes: obtaining the first reference motion vector field according to the prediction information, where the first reference motion vector field is a motion vector field of one video frame at time t1; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used for the video frame corresponding to the first reference motion vector field; and obtaining a motion vector field block of the second reference motion vector field, wherein a range of coordinates of the motion vector field block of the second reference motion vector field in the second motion vector field reference is equal to a coordinate range of the current motion vector field block in the current motion vector field, and the prediction signal includes the motion vector field block of the second reference motion vector field.

[222] In this modality, in order to facilitate understanding, below, the motion vector fields and video frames to be described are explained first: A location of a target object in the video frame at time t1 is A, a video reference frame used to perform interframe prediction in the video frame is a video frame at time t2, and a target object location in the video frame at time t2 is B. In this case, a motion vector of a sample corresponding to the target object in the first reference motion vector field is used to indicate the displacement from location B to location A.

[223] It is assumed that a state of motion (including a speed and a direction) of the target object remains unchanged, i.e. the corresponding displacement of the target object within a time from t1 to t2 is . In this case, it can be deduced that the displacement of the target object within a time from t to t1 must be . That is, it is assumed that a video location at time t is C, and at that location A to location C must be

[224] According to the above method, each sampling point in the first reference motion vector field is considered as a target object, and a location to which each sampling point is moved at time t can be deduced. Each sampling point in the first reference motion vector field is moved. The offset between a location of each sampling point after the sampling point is moved and a sampling point location before the sampling point is moved is , a vector of moving the sampling point before the sampling point is moved is , and the sampling point motion vector changes to after the sampling point is moved. For the sake of ease of description, a new motion vector field that is formed after all sampling points in the first reference motion vector field are moved according to the above rule and vectors of all sampling points change is called the second reference motion vector field.

[225] Therefore, optionally obtaining a second reference motion vector field according to the first reference motion vector field, the moment t, the moment t1, and a moment t2 specifically includes: obtaining a motion vector of a second sampling point of the second reference motion vector field by calculating according to a formula , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as offset is equal to a location of the second sampling point.

[226] In this embodiment, when all sampling points in the first reference motion vector field are moved so as to obtain the second reference motion vector field, at least two sampling points in the first motion vector field reference points will be moved to the same location in the motion vector field at time t.

[227] In other words, the current speeds and directions of all sampling points in the first reference motion vector field remain unchanged, and in the motion vector field formed at time t (i.e., the second motion vector field (reference motion vector field), at least two sampling points in the first reference motion vector field are possibly moved to the same location. There are multiple ways to determine a value of a sampling point at the location in the second reference motion vector field. For example, a product of a motion vector of one of the sampling points and (t-t1)/(t1-t2) can be used as a motion vector of the sampling point in the location.

[228] Alternatively, optionally, obtaining a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2 specifically includes: determining a second sampling point of the second reference motion vector field, wherein a location of the second sampling point is equal to a location to which movement is performed from a location of each of the at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as displacement, and the motion vector of each of the first two points sampling is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[229] To facilitate the description below, the second sampling point is referred to as the second target sampling point, and the first two sampling points of the first reference motion vector field are referred to as the first set.

[230] During the determination of the weights of the motion vectors of all the first sampling points in the first set, the weights of the motion vectors of all the first sampling points in the first set can be set equal, i.e. an average value of the motion vectors of all the first sampling points in the first set is used as the motion vector of the second sampling target point.

[231] Alternatively, determining the motion vector weights of all the first sampling points in the first set specifically includes the following steps:

[232] S71: Obtain a motion vector of at least one second sampling point that is in the second reference motion vector field and that is located around the second sampling target point.

[233] For ease of description, the at least one second sampling point located around the second target sampling point is called the second set, and each second sampling point is an element of the second set.

[234] Optionally, the second set may include at least one of four second sampling points that are located around the second sampling target point and that are adjacent to the second sampling target point.

[235] S72: Compute a degree of similarity between each of the motion vectors of the first sample points in the first set and a motion vector from the second set.

[236] For each first sampling point in the first set, a degree of similarity between a motion vector of the first sampling point and a motion vector of each sampling point in the second set is calculated. There are multiple methods for calculating degree of similarity. For example, the differences between the motion vector of the first sampling point and the motion vectors of all elements in the second set can be calculated, and then a sum or average value of all the differences is used as the degree of similarity between the motion vector of the first sampling point and the motion vector of the second set. In this case, a smaller sum or average value of the differences indicates a greater degree of similarity.

[237] Of course, the above descriptions are only an example and do not constitute any limitation.

[238] S73: Determining a weight of a motion vector of each first sampling point in the first set according to the degree of similarity, where a greater degree of similarity between a motion vector of a first sampling point and the second set motion vector indicates that the first sample point has a larger motion vector weight.

[239] After the degree of similarity corresponding to the motion vector of each element in the first set is determined, the weight of the motion vector of each element in the first set is determined according to a magnitude of the degree of similarity. An element with a higher degree of similarity has a higher motion vector weight. Specifically, weights corresponding to different classifications can be predefined. After a rank of one degree of similarity of each element in the first set is determined, a weight corresponding to the rank of the element is used as a weight of a move vector of the element.

[240] The foregoing describes a case in which at least two first sampling points in the first reference motion vector field are moved to the same location in the reference motion vector field at time t.

[241] In a real application possibly also there is a special location in the second reference motion vector field, and no sampling point the first reference motion vector field is moved to the special location at time t.

[242] Therefore, optionally obtaining a second reference motion vector field according to the first reference motion vector field, the moment t, the moment t1, and a moment t2 specifically includes: obtaining at least the second sampling point near a second sampling target point of the second reference motion vector field, wherein a location to which movement is performed from a location of any first sampling point of the first vector field reference movement as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[243] For a specific explanation of the solution in the third example, refer to the embodiment descriptions shown in Figure 3 and Figure 4 in the motion vector field encoding method. Details will not be described again in this document.

[244] In a fourth example, the prediction information includes direction coefficient information used to indicate a direction coefficient of the current motion vector field block. The direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the motion vector of the sampling point. sampling.

[245] Obtaining a prediction signal from the current motion vector field block according to the prediction information includes: obtaining a reconstruction value of the first component and calculating a predictor of the second component according to the direction coefficient and the reconstruction value of the first component, wherein the prediction signal includes the second component predictor.

[246] There are multiple types of direction coefficient information from the current motion vector field block. For example, the direction coefficient information includes information used to indicate a reconstructed motion vector field block in the current motion vector field, and the direction coefficient includes a direction coefficient of the reconstructed motion vector field block . Alternatively, the bearing coefficient information includes a bearing coefficient value.

[247] For a specific explanation of the solution in the fourth example, refer to the embodiment descriptions shown in Figure 5 in the motion vector field encoding method. Details will not be described again in this document.

[248] The foregoing describes the motion vector field encoding method and the motion vector field decoding method in the present invention. Next, an encoding apparatus is described in the present invention.

[249] Referring to Figure 10, an encoding apparatus provided in an embodiment of the present invention includes: a first acquisition module 1001, configured to obtain an original signal from a current motion vector field block, wherein the block of current motion vector field is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; a second acquisition module 1002, configured to obtain a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required for get the prediction signal; a calculation module 1003 configured to calculate a residual prediction signal of the current motion vector field block in accordance with the prediction signal obtained by the second acquisition module and the original signal obtained by the first acquisition module, wherein the residual prediction signal is used to indicate a residual between the original signal and the prediction signal; and an encoding module (1004) configured to record, in a bit stream, the prediction information obtained by the second acquisition module and the residual prediction signal obtained by means of calculation by the calculation module.

[250] In this embodiment, during current block encoding, the encoding apparatus does not need to encode the original signal of the current motion vector field block, but encodes the prediction information and the residual prediction signal. Therefore, the motion vector field compression efficiency is improved.

[251] In some possible implementations of the present invention, the second acquisition module 1002 is configured to: determine an intraframe prediction mode, use at least one block of encoded and reconstructed motion vector field that is in the motion vector field and which is close to the current motion vector field block as a reference motion vector field block of the current block, and obtain the prediction signal of the current block according to the intraframe prediction mode and the current block. reference move vector field.

[252] For example, the intraframe predictive mode obtained is a horizontal predictive mode in the 33 directional predictive modes. In this case, the current block's reference motion vector field block is a motion vector field block that is located on the same row as the current block and which is the first motion vector field block to the left of the block current. In that case, the second acquisition module 1002 is configured to use a reference motion vector field block reconstruction signal as the current block prediction signal.

[253] As another example, the intraframe prediction mode obtained is Intra_DC mode. After obtaining the reference motion vector field block of the current block, the second obtain module 1002 is configured to use an average value of reconstruction pixels of the reference motion vector field block as the prediction signal of the block current.

[254] In some possible implementations of the present invention, the second obtaining module 1002 is configured to: obtain a first reference motion vector field from the current motion vector field block, wherein the first motion vector field The first reference motion vector field is a motion vector field corresponding to a video frame at time t1, and the video frame at time t1 is a motion vector field. video frame next to video frame at time t; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, and that the time t2 is a time corresponding to a reference frame video used to perform interframe prediction on the video frame corresponding to the first reference motion vector field; obtaining the prediction signal according to the second reference motion vector field, wherein the prediction signal comprises a motion vector field block of the second reference motion vector field, and a block coordinate range of motion vector field of the second reference motion vector field in the second reference motion vector field is equal to a coordinate range of the current motion vector field block in the current motion vector field; and obtaining the predictive information in accordance with the first reference motion vector field, wherein the predictive information comprises information used to indicate the first reference motion vector field.

[255] In some possible implementations of the present invention, the second acquisition module 1002 is configured to obtain a motion vector of a second sampling point from the second reference motion vector field by means of calculation according to a formula , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the motion vector moving the second sampling point as displacement is equal to one location of the second sampling point.

[256] In some possible implementations of the present invention, the second acquisition module 1002 is configured to: determine a second sampling point from the second reference motion vector field, where a location of the second sampling point is equal to a location to which movement is performed from a location of each of the at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the at least first two sampling points as displacement, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1 )/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[257] In some possible implementations of the present invention, the second acquisition module 1002 is configured to: obtain at least one second sampling point close to a second sampling target point of the second reference motion vector field, where a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[258] In some possible implementations of the present invention, the second obtaining module 1002 is configured to: obtain a direction coefficient from the current motion vector field block, whereby the direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the motion vector of the sampling point; obtaining a reconstruction value of the first component; calculating the prediction signal according to the first component reconstruction value and the direction coefficient, the prediction signal including a second component predictor; and obtaining the prediction information according to the direction coefficient, the prediction information including the direction coefficient information used to indicate the direction coefficient.

[259] In some possible implementations of the present invention, the second acquisition module 1002 is configured to: obtain motion vectors from at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables of a predefined function, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and function values; and use a coefficient from the predefined function obtained through fitting as the direction coefficient.

[260] In some possible implementations of the present invention, the second acquisition module 1002 is configured to: obtain motion vectors from at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and the values of function; and obtaining a direction coefficient of at least one encoded motion vector field block that is in the current motion vector field and that is close to the current motion vector field block; using a function coefficient obtained through fitting and the direction coefficient of the at least one encoded motion vector field block as candidate direction coefficients from a set of candidate direction coefficients of the current motion vector field block; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient.

[261] In some possible implementations of the present invention, the second acquisition module 1002 is configured to: when the direction coefficients of at least two encoded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, use the direction coefficients of the at least two coded motion vector field blocks as the direction coefficient of the current motion vector field block.

[262] In some possible implementations of the present invention, the second acquisition module 1002 is configured to: when the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, and the direction coefficients of the at least two encoded motion vector field blocks are used to indicate a ratio of a first component of a motion vector to a sampling point of the at least two coded motion vector field blocks and a second motion vector component of the sampling point, perform the following steps: use the direction coefficients of the at least two coded motion vector field blocks as a coefficient candidate direction from a set of candidate direction coefficients; obtain a motion vector of at least one sampling point in the original signal, and when the at least one sampling point is a sampling point, use a ratio between a first component of the motion vector of the sampling point and a second component of the sampling point motion vector as a candidate direction coefficient from the set of candidate direction coefficients; or when the at least one sampling point is at least two sampling points, use an average value of ratios between the first motion vector components of the at least two sampling points and the second motion vector components of the at least two sampling points as a candidate direction coefficient from the set of candidate direction coefficients; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient.

[263] The foregoing describes the motion vector field encoding method, the motion vector field decoding method and the encoding apparatus in the present invention. Next, a decoding apparatus is described in the present invention.

[264] Referring to Figure 11, a decoding apparatus provided in an embodiment of the present invention includes: a first acquisition module 1101, configured to obtain prediction information and a residual prediction signal from a motion vector field block current motion, where the current motion vector field block is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame in a moment t; a second acquisition module 1102 configured to obtain a prediction signal from the current motion vector field block in accordance with the prediction information; and a calculation module 1103 configured to calculate a reconstruction signal of the current motion vector field block in accordance with the prediction signal obtained by the second acquisition module and the residual prediction signal obtained by the first acquisition module.

[265] In some possible implementations of the present invention, the prediction information includes information used to indicate a first reference motion vector field of the motion vector field block, and the second acquisition module 1102 is configured to: obtain the first reference motion vector field according to the prediction information, wherein the first reference motion vector field is a motion vector field of one video frame at time t1; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used for the video frame corresponding to the first reference motion vector field; and obtaining a motion vector field block of the second reference motion vector field, wherein a range of coordinates of the motion vector field block of the second reference motion vector field in the second motion vector field reference is equal to a range of coordinates of the current motion vector field block in the current motion vector field, and the prediction signal comprises the motion vector field block of the second reference motion vector field.

[266] In some possible implementations of the present invention, the second acquisition module 1202 is configured to: obtain a motion vector of a second sampling point from the second reference motion vector field by means of calculation according to a formula , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as offset is equal to a location of the second sampling point.

[267] In some possible implementations of the present invention, the second acquisition module 1202 is configured to: determine a second sampling point from the second reference motion vector field, where a location of the second sampling point is equal to a location to which motion is performed from a location of each of the at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the at least first two sampling points as displacement, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1 )/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[268] In some possible implementations of the present invention, the second acquisition module 1202 is configured to: obtain at least one second sampling point close to a second sampling target point of the second reference motion vector field, where a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[269] In some possible implementations of the present invention, the prediction information includes direction coefficient information used to indicate a direction coefficient of the current motion vector field block, and the direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block, and a value of a second component of the motion vector of the sampling point.

[270] The second obtaining module 1102 is configured to obtain a reconstruction value of the first component and calculate a predictor of the second component according to the direction coefficient and the reconstruction value of the first component, where the prediction signal includes the predictor of the second component.

[271] In some possible implementations of the present invention, the direction coefficient information includes information used to indicate a motion vector field block reconstructed into the current motion vector field, and the direction coefficient includes a direction coefficient of the reconstructed drive vector field block. Alternatively, the bearing coefficient information includes a bearing coefficient value.

[272] Referring to Figure 12, Figure 12 is a structural block diagram of an encoding apparatus 1200 provided in another embodiment of the present invention. The encoding apparatus 1200 can include at least one processor 1201, one memory 1205 and at least one communication bus 1202. Optionally, the video encoding apparatus 1200 can additionally include at least one network interface 1204 and/or one user interface 1203. User interface 1203 includes, for example, a display (e.g., a touchscreen, an LCD, holographic imaging, a CRT, or a projector (Projector)), a click device (for example, a mouse, trackball, touchpad, or touchscreen), a camera, and/or a sound pickup device.

[273] Memory 1205 may include read-only memory and random access memory and may provide an instruction and data to processor 1201. A portion of memory 1205 may further include non-volatile random access memory.

[274] In some deployments, memory 1205 will store the following elements, i.e., an executable module or a data structure or a subset thereof, or a set of extensions thereof: an operating system 12051 including various system and configured to deploy several basic services and process a hardware-based job; and an application program module 12052, which includes various application programs and configured to implement various application services.

[275] In this embodiment of the present invention, by invoking the program or an instruction stored in memory 1205, the processor 1201 is configured to: obtain an original signal from a current motion vector field block, where the field block current motion vector field is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required to obtain the prediction signal; calculate a residual prediction signal from the current motion vector field block according to the prediction signal and the original signal, where the residual prediction signal is used to indicate a residual between the original signal and the prediction signal ; and recording the prediction information and the residual prediction signal into a bit stream.

[276] In this embodiment of the present invention, during current block encoding, the original current motion vector field block signal need not be encoded; instead, the prediction information and the residual prediction signal are encoded. Therefore, the motion vector field compression efficiency is improved.

[277] Optionally, obtaining a current motion vector field block prediction signal and current motion vector field block prediction information includes: obtaining a first reference motion vector field from the current motion vector field block, wherein the first reference motion vector field is a reconstructed encoded motion vector field, the first reference motion vector field is a motion vector field corresponding to a video frame at time t1 and the video frame at time t1 is a video frame close to the video frame at time t; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used to perform interframe prediction on the video frame corresponding to the first reference motion vector field; obtaining the prediction signal according to the second reference motion vector field, the prediction signal including a motion vector field block of the second reference motion vector field, and a block coordinate range of motion vector field of the second reference motion vector field in the second reference motion vector field is equal to a coordinate range of the current motion vector field block in the current motion vector field; and obtaining the prediction information in accordance with the first reference motion vector field, the prediction information including information used to indicate the first reference motion vector field.

[278] Optionally, obtaining a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2 includes: obtaining a motion vector from a second sampling point of the second reference motion vector field by means of calculation according to a formula , being that is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as offset is equal to a location of the second sampling point.

[279] Optionally, obtaining a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2 includes: determining a second point of reference sampling the second reference motion vector field, wherein a location of the second sampling point is equal to a location to which movement is performed from a location of each of the at least two first sampling points of the first motion field reference motion vector as a starting point according to a motion vector of each of the first at least two sampling points as displacement, and the motion vector of each of the first at least two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[280] Optionally, obtaining a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2 includes: obtaining at least one second sampling point near a second target sampling point of the second reference motion vector field, wherein a location to which movement is performed from a location of any first sampling point of the first motion vector field of reference reference as a starting point I according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[281] Optionally, obtaining a current motion vector field block prediction signal and current motion vector field block prediction information includes: obtaining a vector field block direction coefficient of the current motion vector, where the direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the current motion vector. sampling point motion vector; obtaining a reconstruction value of the first component; calculating the prediction signal according to the first component reconstruction value and the direction coefficient, the prediction signal including a second component predictor; and obtaining the prediction information according to the direction coefficient, the prediction information including the direction coefficient information used to indicate the direction coefficient.

[282] Optionally, obtaining a direction coefficient of the current motion vector field block includes: obtaining motion vectors from at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables of a predefined function, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and function values; and use a coefficient from the predefined function obtained through fitting as the direction coefficient.

[283] Optionally, obtaining a direction coefficient of the current motion vector field block includes: obtaining motion vectors from at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and the function values; obtaining a direction coefficient of at least one encoded motion vector field block that is in the current motion vector field and that is close to the current motion vector field block; using a function coefficient obtained through fitting and the direction coefficient of the at least one encoded motion vector field block as candidate direction coefficients from a set of candidate direction coefficients of the current motion vector field block; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient.

[284] Optionally, obtaining a direction coefficient of the current motion vector field block includes: when the direction coefficients of at least two coded motion vector field blocks that are in the motion vector field that are adjacent to the current motion vector field block are the same, use the direction coefficients of the at least two coded motion vector field blocks as the direction coefficient of the current motion vector field block.

[285] Optionally, obtaining a direction coefficient of the current motion vector field block includes: when the direction coefficients of at least two coded motion vector field blocks that are in the motion vector field that are adjacent to the current motion vector field block are equal, and the direction coefficients of the at least two coded motion vector field blocks are used to indicate a ratio between a first component of a motion vector of a sampling point of the at least two coded motion vector field blocks and a second motion vector component of the sampling point, perform the following steps: use the direction coefficients of the at least two motion vector field blocks encoded as a candidate direction coefficient from a set of candidate direction coefficients; obtain a motion vector of at least one sampling point in the original signal, and when the at least one sampling point is a sampling point, use a ratio between a first component of the motion vector of the sampling point and a second component of the sampling point motion vector as a candidate direction coefficient from the set of candidate direction coefficients; or when the at least one sampling point is at least two sampling points, use an average value of ratios between the first motion vector components of the at least two sampling points and the second motion vector components of the at least two sampling points as a candidate direction coefficient from the set of candidate direction coefficients; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient.

[286] Referring to Figure 13, Figure 13 is a structural block diagram of a decoding apparatus 1300 provided in another embodiment of the present invention. The decoding apparatus 1300 may include at least one processor 1301, a memory 1305 and at least one communication bus 1302. Optionally, the video decoding apparatus 1300 may additionally include at least one network interface 1304 and/or a user interface 1303. User interface 1303 includes, for example, a display (e.g., a touchscreen, an LCD, holographic imaging, a CRT, or a projector (Projector)), a click device (for example, a mouse, trackball, touchpad, or touchscreen), a camera, and/or a sound pickup device.

[287] Memory 1305 may include read-only memory and random access memory and may provide an instruction and data to processor 1001. A portion of memory 1305 may further include non-volatile random access memory.

[288] In some deployments, memory 1305 will store the following elements, namely, an executable module or a data structure or a subset thereof, or a set of extensions thereof: an operating system 13051 including various system and configured to deploy several basic services and process a hardware-based job; and an application program module 13052, which includes various application programs and configured to implement various application services.

[289] In this embodiment of the present invention, by invoking the program or an instruction stored in memory 1305, processor 1301 is configured to: obtain prediction information and a residual prediction signal from a current motion vector field block, wherein the current motion vector field block is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to one video frame at a time t; obtaining a prediction signal from the current motion vector field block according to the prediction information; and calculating a reconstruction signal of the current motion vector field block in accordance with the prediction signal and the residual prediction signal.

[290] Optionally, the prediction information includes information used to indicate a first reference motion vector field from the motion vector field block.

[291] Obtaining a prediction signal from the current motion vector field block according to the prediction information includes: obtaining the first reference motion vector field according to the prediction information, where the first reference motion vector field is a motion vector field of one video frame at time t1; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used for the video frame corresponding to the first reference motion vector field; and obtaining a motion vector field block of the second reference motion vector field, wherein a range of coordinates of the motion vector field block of the second reference motion vector field in the second motion vector field reference is equal to a coordinate range of the current motion vector field block in the current motion vector field, and the prediction signal includes the motion vector field block of the second reference motion vector field.

[292] Optionally, obtaining a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2 includes: obtaining a motion vector from a second sampling point of the second reference motion vector field by means of calculation according to a formula , being that is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the move vector of the second sampling point as offset is equal to a location of the second sampling point.

[293] Optionally, obtaining a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2 includes: determining a second point of reference sampling the second reference motion vector field, wherein a location of the second sampling point is equal to a location to which movement is performed from a location of each of the at least two first sampling points of the first motion field reference motion vector as a starting point according to a motion vector of each of the first at least two sampling points as displacement, and the motion vector of each of the first at least two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point.

[294] Optionally, obtaining a second reference motion vector field according to the first reference motion vector field, time t, time t1, and time t2 includes: obtaining at least one second sampling point near a second target sampling point of the second reference motion vector field, wherein a location to which movement is performed from a location of any first sampling point of the first motion vector field of reference reference as a starting point I according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of the second sampling point as a motion vector of the second sampling target point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point.

[295] Optionally, the prediction information includes direction coefficient information used to indicate a direction coefficient of the current motion vector field block. The direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the motion vector of the sampling point. sampling;

[296] Obtaining a prediction signal from the current motion vector field block according to the prediction information includes: obtaining a reconstruction value of the first component and calculating a predictor of the second component according to the direction coefficient and the reconstruction value of the first component, wherein the prediction signal includes the second component predictor.

[297] Optionally, the direction coefficient information includes information used to indicate a reconstructed motion vector field block into the current motion vector field, and the direction coefficient includes a direction coefficient of the motion vector field block rebuilt drive. Alternatively, the bearing coefficient information includes a bearing coefficient value.

[298] It may be understood by persons skilled in the art that, for the sake of ease and brevity of description, for the detailed working processes of the aforesaid systems, apparatus and units, reference may be made to the corresponding processes in the aforesaid method embodiments. Details will not be described again in this document.

[299] In the various embodiments provided in this application, it is to be understood that the disclosed systems, apparatus, and methods may be implemented in other ways. For example, the device embodiment described is simply an example. For example, unit division is merely a logic function division and may be another division in a real implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be overlooked or may not be realized. In addition, the mutual couplings or direct couplings or communication connections shown or discussed can be implemented using some interfaces. Couplings or indirect communication connections between appliances or units may be implanted in electrical, mechanical, or other forms.

[300] Units depicted as separate parts may or may not be physically separate, and parts shown as units may or may not be physical units, may be located in one location, or may be distributed across multiple network units. Some or all units can be selected according to actual requirements to achieve the objectives of the solutions in the modalities.

[301] Additionally, the functional units in the embodiments of the present invention can be integrated into a processing module, or each of the units can physically exist by itself, or two or more units can be integrated into a module. The integrated unit may be implemented in a hardware form, or it may be implemented in a form of a software functional unit.

[302] When the integrated unit is deployed in the form of a functional software unit and marketed or used as a stand-alone product, the integrated unit may be stored on computer-readable storage media. Based on such an understanding, the technical solutions of the present invention essentially, or the part contributing to the prior art, or all or a part of the technical solutions can be implemented in a software product form. The computer software product is stored on storage media and includes various instructions for instructing a computer device (which may be a personal computer, a server, a network device, or the like) to perform all or part of the steps of the methods described in the embodiments of the present invention. The aforementioned storage media include: any media that can store program code, such as a USB flash drive, a removable hard disk, read-only memory (ROM), random access memory (RAM , Random Access Memory), a magnetic disk or an optical disk.

[303] The aforementioned embodiments are intended only to describe the technical solutions of the present invention, however, they are not intended to limit the present invention. Although the present invention is described in detail with reference to the aforementioned embodiments, people of common skill in the art should understand that it is still possible to make modifications to the technical solutions described in the aforementioned embodiments or to perform equivalent replacements in some technical resources, without that there is a departure from the spirit and scope of the technical solutions of the modalities of this application.

Claims (34)

1. A motion vector field encoding method comprising: obtaining (101) an original signal from a current motion vector field block, CHARACTERIZED in that the current motion vector field block is obtained by dividing it if a current motion vector field is in blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining (102) a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required to obtain the prediction signal and includes an index of the interframe prediction mode, where the interframe prediction mode is applied to the interframe prediction of the current motion vector field, and the prediction signal of the current motion vector field block is obtained from a block of reference motion vector field according to the interframe prediction mode indicated by the index; calculating (103) a residual prediction signal from the current motion vector field block according to the prediction signal and the original signal, wherein the residual prediction signal is used to indicate a residual between the original signal and the original signal. forecast signal; and recording (104) the prediction information and the residual prediction signal into a bit stream. 2. Motion vector field encoding method, according to claim 1, CHARACTERIZED by the fact that obtaining (102) a prediction signal from the current motion vector field block and the block prediction information of current motion vector field comprising: obtaining (S11) a first reference motion vector field from the current motion vector field block, wherein the first reference motion vector field is a motion vector field encoded and reconstructed, where the first reference motion vector field is a motion vector field that corresponds to a video frame at time t1, and the video frame at time t1 is a video frame next to frame of video at time t; obtaining (S12) a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a video reference frame used to perform interframe prediction on the video frame corresponding to the first reference motion vector field; obtaining (S13) the prediction signal according to the second reference motion vector field, wherein the prediction signal comprises a motion vector field block of the second reference motion vector field, and a range of motion vector field block coordinates of the second reference motion vector field in the second reference motion vector field is equal to one range of coordinates of the current motion vector field block in the current motion vector field ; and obtaining (S14) predictive information in accordance with the first reference motion vector field, wherein the predictive information comprises information used to indicate the first reference motion vector field. 3. Motion vector field encoding method, according to claim 2, CHARACTERIZED by the fact that obtaining a second reference motion vector field in accordance with the first reference motion vector field, the moment t, the moment t1 and a moment t2 comprises: obtaining a motion vector of a second sampling point from the second reference motion vector field by means of calculation according to a formula , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the motion vector of the second sampling point as the offset is equal to a location of the second sampling point. 4. Motion vector field encoding method, according to claim 2, CHARACTERIZED by the fact that obtaining a second reference motion vector field in accordance with the first reference motion vector field, the time t, time t1, and time t2 comprises: determining a second sampling point of the second reference motion vector field, where a location of the second sampling point is equal to a location to which motion is performed from a location of each of the first two sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as an offset, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point. 5. Motion vector field encoding method, according to claim 2, CHARACTERIZED by the fact that obtaining a second reference motion vector field in accordance with the first reference motion vector field, the time t, time t1 and a time t2 comprises: obtaining at least a second sampling point near a second sampling target point of the second reference motion vector field, where a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point UUUUR according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of a motion vector of the second sampling point as a motion vector of the second target sampling point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point. 6. Motion vector field encoding method, comprising: obtaining an original signal from a current motion vector field block, wherein the current motion vector field block is obtained by dividing a current motion vector field in blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information necessary to obtain the prediction signal; calculating a prediction residual signal from the current motion vector field block according to the prediction signal and the original signal, wherein the prediction residual signal is used to indicate a residual between the original signal and the prediction signal; and writing the prediction information and the residual prediction signal into a bit stream, the method CHARACTERIZED in that obtaining a prediction signal from the current motion vector field block and the prediction information from the field block current motion vector field comprises: obtaining (S31) a direction coefficient of the current motion vector field block, wherein the direction coefficient is used to indicate a relationship between a value of a first component of a current motion vector field a sampling point of the current motion vector field block and a value of a second motion vector component of the sampling point; obtaining (S32) a reconstruction value of the first component; calculating (S33) the prediction signal according to the first component reconstruction value and the direction coefficient, wherein the prediction signal comprises a second component predictor; and obtaining (S34) the prediction information according to the direction coefficient, wherein the prediction information comprises direction coefficient information used to indicate the direction coefficient. 7. Motion vector field encoding method, according to claim 6, CHARACTERIZED by the fact that obtaining a direction coefficient of the current motion vector field block comprises: obtaining (S41) motion vectors of at least two sampling points of the original signal; using (S42) the first components of the motion vectors of the at least two sampling points as independent variables of a predefined function, using the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust independent variables and function values; and use a coefficient from the predefined function obtained through fitting as the direction coefficient. 8. Motion vector field encoding method, according to claim 6, CHARACTERIZED by the fact that obtaining a direction coefficient of the current motion vector field block comprises: obtaining (S51) motion vectors of at least two sampling points of the original signal; using (S52) the first components of the motion vectors of the at least two sampling points as independent variables, using the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjusting the variables independent and function values; obtaining (S53) a direction coefficient of at least one encoded motion vector field block that is in the current motion vector field and that is close to the current motion vector field block; using (S54) a function coefficient obtained by fitting and the direction coefficient of the at least one encoded motion vector field block as candidate direction coefficients from a set of candidate direction coefficients of the motion vector field block current drive; and obtaining (S55) candidate residual prediction signals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and using a candidate direction coefficient corresponding to a residual prediction signal of candidate with lower signal energy or lower rate distortion as the current motion vector field block direction coefficient. 9. Motion vector field encoding method, according to claim 6, CHARACTERIZED by the fact that obtaining a direction coefficient of the current motion vector field block comprises: when the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, use the direction coefficients of the at least two motion vector field blocks encoded as the direction coefficient of the current motion vector field block. 10. Motion vector field encoding method, according to claim 6, CHARACTERIZED by the fact that obtaining a direction coefficient of the current motion vector field block comprises: when the direction coefficients of at least two coded motion vector field blocks that are in the current motion vector field and that are adjacent to the current motion vector field block are equal, and the direction coefficients of the at least two motion vector field blocks encoded are used to indicate a ratio of a first component of a motion vector of a sampling point of the at least two coded motion vector field blocks to a second component of the motion vector of the sampling point, perform the following steps (S61): using (S62) the direction coefficients of the at least two coded motion vector field blocks as a candidate direction coefficient from a set of candidate direction coefficients; obtaining (S63) a motion vector of at least one sampling point in the original signal; when the at least one sampling point is a sampling point, use a ratio of a first component of the sampling point motion vector to a second component of the sampling point motion vector as a candidate direction coefficient of the set of candidate direction coefficients; or when the at least one sampling point is at least two sampling points, use an average value of ratios between the first motion vector components of the at least two sampling points and the second motion vector components of the at least two sampling points as a candidate direction coefficient from the set of candidate direction coefficients; and obtaining (S64) candidate residual prediction signals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and using a candidate direction coefficient corresponding to a residual prediction signal of candidate with lower signal energy or lower rate distortion as the current motion vector field block direction coefficient. 11. Motion vector field decoding method, FEATURED in that it comprises: obtaining (901) prediction information and a residual prediction signal from a current motion vector field block, wherein the motion vector field block current motion vector is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining (902) a prediction signal from the current motion vector field block in accordance with the prediction information, wherein the prediction information includes an index of the interframe prediction mode, wherein the interframe prediction mode is applied to current motion vector field interframe prediction and the current motion vector field block prediction signal is obtained from a reference motion vector field block according to the interframe prediction mode indicated by the index; and calculating (903) a current motion vector field block reconstruction signal in accordance with the prediction signal and the residual prediction signal. 12. Motion vector field decoding method, according to claim 11, CHARACTERIZED by the fact that the prediction information comprises information used to indicate a first reference motion vector field of the motion vector field block ; and obtaining a prediction signal from the current motion vector field block in accordance with the prediction information comprises: obtaining the first reference motion vector field in accordance with the prediction information, wherein the first motion vector field reference motion vector is a motion vector field of one video frame at time t1; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used for the video frame corresponding to the first reference motion vector field; and obtaining a motion vector field block of the second reference motion vector field, wherein a range of coordinates of the motion vector field block of the second reference motion vector field in the second motion vector field reference is equal to a range of coordinates of the current motion vector field block in the current motion vector field, and the prediction signal comprises the motion vector field block of the second reference motion vector field. 13. Motion vector field decoding method, according to claim 12, CHARACTERIZED by the fact that obtaining a second reference motion vector field according to the first reference motion vector field, the moment t, the moment t1 and a moment t2 comprises: obtaining a motion vector of a second sampling point from the second reference motion vector field by means of calculation according to a formula , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the motion vector moving the second sampling point as displacement is equal to one location of the second sampling point. 14. Motion vector field decoding method, according to claim 12, CHARACTERIZED by the fact that obtaining a second reference motion vector field according to the first reference motion vector field, the time t, time t1, and time t2 comprises: determining a second sampling point of the second reference motion vector field, where a location of the second sampling point is equal to a location to which motion is performed from a location of each of the first two sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as an offset, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point. 15. Motion vector field decoding method, according to claim 12, CHARACTERIZED by the fact that obtaining a second reference motion vector field according to the first reference motion vector field, the time t, time t1 and a time t2 comprises: obtaining at least a second sampling point near a second sampling target point of the second reference motion vector field, where a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point in accordance with as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of a motion vector of the second sampling point as a motion vector of the second target sampling point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point. 16. A motion vector field decoding method, comprising: obtaining prediction information and a residual prediction signal from a current motion vector field block, wherein the current motion vector field block is obtained by dividing a current motion vector field in blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; obtaining a prediction signal from the current motion vector field block according to the prediction information; and calculating a reconstruction signal of the current motion vector field block in accordance with the prediction signal and the residual prediction signal, the method being CHARACTERIZED by the fact that the prediction information comprises direction coefficient information used to indicate a direction coefficient of the current motion vector field block, and the direction coefficient is used to indicate a relationship between a value of a first component of a motion vector to a sampling point of the motion vector field block current and a value of a second component of the sampling point motion vector; and obtaining a prediction signal of the current motion vector field block in accordance with the prediction information comprises: obtaining a reconstruction value of the first component and calculating a predictor of the second component according to the direction coefficient and the reconstruction value of the first component, wherein the predictor signal comprises the predictor of the second component. 17. Motion vector field decoding method according to claim 16, CHARACTERIZED by the fact that the direction coefficient information comprises information used to indicate a motion vector field block reconstructed into the motion vector field current, and the direction coefficient comprises a direction coefficient of the reconstructed motion vector field block; or the direction coefficient information comprises a direction coefficient value. 18. Encoding apparatus FEATURED in that it comprises: a first obtaining module (1001) configured to obtain an original signal from a current motion vector field block, wherein the current motion vector field block is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; a second obtaining module (1002), configured to obtain a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate information required to obtain the prediction signal and include an index of the interframe prediction mode, where the interframe prediction mode is applied to either the interframe prediction of the current motion vector field and the prediction signal of the motion vector field block current is taken from a reference motion vector field block according to the interframe prediction mode indicated by the index; a calculation module (1003), configured to calculate a residual prediction signal of the current motion vector field block in accordance with the prediction signal obtained by the second acquisition module and the original signal obtained by the first acquisition module, wherein the predictive residual signal is used to indicate a residual between the original signal and the predictive signal; and an encoding module (1004) configured to record, in a bit stream, the prediction information obtained by the second acquisition module and the residual prediction signal obtained by means of calculation by the calculation module. 19. Encoding apparatus, according to claim 18, CHARACTERIZED by the fact that the second obtaining module is configured to: obtain a first reference motion vector field from the current motion vector field block, in which the first reference motion vector field is a encoded and reconstructed motion vector field, the first reference motion vector field is a motion vector field corresponding to a video frame at time t1, and the reference motion vector field video at time t1 is a video frame next to the video frame at time t; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used to perform interframe prediction on the video frame corresponding to the first reference motion vector field; obtaining the prediction signal according to the second reference motion vector field, wherein the prediction signal comprises a motion vector field block of the second reference motion vector field, and a block coordinate range of motion vector field of the second reference motion vector field in the second reference motion vector field is equal to a coordinate range of the current motion vector field block in the current motion vector field; and obtaining the predictive information in accordance with the first reference motion vector field, wherein the predictive information comprises information used to indicate the first reference motion vector field. 20. Coding apparatus, according to claim 19, CHARACTERIZED by the fact that the second acquisition module is configured to obtain a movement vector of a second sampling point from the second reference movement vector field by means of calculation according to a formula , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the motion vector moving the second sampling point as displacement is equal to one location of the second sampling point. 21. Coding apparatus, according to claim 19, CHARACTERIZED by the fact that the second acquisition module is configured to: determine a second sampling point of the second reference motion vector field, where a location of the second point of sampling is equal to a location to which movement is performed from a location of each of the at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as a displacement, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points. sampling and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point. 22. Coding apparatus, according to claim 19, CHARACTERIZED by that the second acquisition module is configured to: obtain at least a second sampling point close to a second sampling target point of the second motion vector field reference, wherein a location to which movement is performed from a location of any first sampling point of the first reference motion vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of a motion vector of the second sampling point as a motion vector of the second target sampling point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point. 23. Coding apparatus, comprising: a first obtaining module, for obtaining an original signal from a current motion vector field block, wherein the current motion vector field block is obtained by dividing a current motion vector field current in blocks, and the current motion vector field is a motion vector field corresponding to one video frame at time t; a second obtaining module for obtaining a prediction signal from the current motion vector field block and prediction information from the current motion vector field block, wherein the prediction information is used to indicate the information required to obtain the prediction signal; a calculation module, for calculating a residual prediction signal of the current motion vector field block in accordance with the prediction signal obtained by the second acquisition module and the original signal obtained by the first acquisition module, wherein the residual signal of prediction is used to indicate a residual between the original signal and the prediction signal; and a coding module, for recording, in a bit stream, the prediction information obtained by the second acquisition module and the residual prediction signal obtained by means of calculation by the calculation module, the apparatus being CHARACTERIZED by the fact that the second get module is configured to: get a direction coefficient from the current motion vector field block, where the direction coefficient is used to indicate a relationship between a value of a first component of a point's motion vector sampling point of the current motion vector field block and a value of a second motion vector component of the sampling point; obtaining a reconstruction value of the first component; calculating the prediction signal according to the reconstruction value of the first component and the direction coefficient, wherein the prediction signal comprises a predictor of the second component; and obtaining the prediction information according to the direction coefficient, wherein the prediction information comprises direction coefficient information used to indicate the direction coefficient. 24. Coding apparatus, according to claim 23, CHARACTERIZED by the fact that the second obtaining module is configured to: obtain movement vectors from at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables of a predefined function, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and function values; and use a coefficient from the predefined function obtained through fitting as the direction coefficient. 25. Coding apparatus, according to claim 23, CHARACTERIZED by the fact that the second obtaining module is configured to: obtain movement vectors of at least two sampling points of the original signal; use the first components of the motion vectors of the at least two sampling points as independent variables, use the second components of the motion vectors of the at least two sampling points as function values corresponding to the independent variables, and adjust the independent variables and the values of function; and obtaining a direction coefficient of at least one encoded motion vector field block that is in the current motion vector field and that is close to the current motion vector field block; using a function coefficient obtained through fitting and the direction coefficient of the at least one encoded motion vector field block as candidate direction coefficients from a set of candidate direction coefficients of the current motion vector field block; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient. 26. Coding apparatus, according to claim 23, CHARACTERIZED by the fact that the second obtaining module is configured for: when the direction coefficients of at least two coded motion vector field blocks that are in the vector field that are adjacent to the current motion vector field block are equal, use the direction coefficients of the at least two coded motion vector field blocks as the direction coefficient of the current motion vector field block . 27. Coding apparatus, according to claim 23, CHARACTERIZED by the fact that the second obtaining module is configured for: when the direction coefficients of at least two coded motion vector field blocks that are in the vector field that are adjacent to the current motion vector field block are equal, and the direction coefficients of the at least two encoded motion vector field blocks are used to indicate a ratio between a first component of a motion vector movement of a sampling point of the at least two coded motion vector field blocks and a second component of the sampling point movement vector, perform the following steps: use the direction coefficients of the at least two vector field blocks of motion encoded as a candidate direction coefficient from a set of candidate direction coefficients; obtaining a motion vector of at least one sampling point in the original signal; when the at least one sampling point is a sampling point, use a ratio of a first component of the sampling point motion vector to a second component of the sampling point motion vector as a candidate direction coefficient of the set of candidate direction coefficients; or when the at least one sampling point is at least two sampling points, use an average value of ratios between the first motion vector components of the at least two sampling points and the second motion vector components of the at least two sampling points as a candidate direction coefficient from the set of candidate direction coefficients; and obtain candidate residuals from the current motion vector field block that are corresponding to candidate direction coefficients from the set of candidate direction coefficients and use a candidate direction coefficient corresponding to a candidate residual signal with less signal energy or least rate distortion as the current motion vector field block direction coefficient. 28. Decoding apparatus FEATURED in that it comprises: a first acquisition module (1101) configured to obtain prediction information and a residual prediction signal from a current motion vector field block, wherein the field block current motion vector field is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; a second acquisition module (1102), configured to obtain a prediction signal from the current motion vector field block in accordance with the prediction information obtained by the first acquisition module, wherein the prediction information includes a mode index interframe prediction mode, where the interframe prediction mode is applied to the interframe prediction of the current motion vector field and the prediction signal of the current motion vector field block is obtained from a motion vector field block reference according to the interframe prediction mode indicated by the index; and a calculation module (1103) configured to calculate a reconstruction signal of the current motion vector field block in accordance with the prediction signal obtained by the second acquisition module and the residual prediction signal obtained by the first acquisition module. obtaining. 29. Decoding apparatus according to claim 28, characterized in that the predictive information comprises information used to indicate a first reference motion vector field of the motion vector field block; and the second get module is configured to: get the first reference motion vector field according to the prediction information, wherein the first reference motion vector field is a motion vector field from a frame of video at a time t1; obtain a second reference motion vector field according to the first reference motion vector field, the time t, the time t1, and a time t2, where the time t2 is a time corresponding to a reference frame video used for the video frame corresponding to the first reference motion vector field; and obtaining a motion vector field block of the second reference motion vector field, wherein a range of coordinates of the motion vector field block of the second reference motion vector field in the second motion vector field reference is equal to a range of coordinates of the current motion vector field block in the current motion vector field, and the prediction signal comprises the motion vector field block of the second reference motion vector field. 30. Decoding apparatus, according to claim 29, CHARACTERIZED by the fact that the second acquisition module is configured to: obtain a motion vector from a second sampling point of the second reference motion vector field by means of calculation according to a formula UUUUR UUUUR , on what is a motion vector of a first sampling point of the first reference motion vector field, and a location to which movement is performed from a location of the first sampling point as a starting point according to the motion vector moving the second sampling point as displacement is equal to one location of the second sampling point. 31. Decoding apparatus, according to claim 29, CHARACTERIZED by the fact that the second acquisition module is configured to: determine a second sampling point of the second reference motion vector field, where a location of the second point of sampling is equal to a location to which movement is performed from a location of each of the at least two first sampling points of the first reference motion vector field as a starting point according to a motion vector of each of the first two sampling points as a displacement, and the motion vector of each of the first two sampling points is a product of a motion vector of each of the first two sampling points. sampling and (t-t1)/(t1-t2); and use a product of a weighted average value of the motion vectors of at least the first two sampling points and (t-t1)/(t1-t2) as a motion vector of the second sampling point. 32. Decoding apparatus, according to claim 29, CHARACTERIZED by the fact that the second acquisition module is configured to: obtain at least a second sampling point close to a second sampling target point of the second vector field of reference movement, wherein a location to which movement is performed from a location of any first sampling point of the first reference movement vector field as a starting point according to as displacement is different from a location of the second sampling target point, and is a motion vector of the first sampling point; and when the at least one second sampling point is a second sampling point, using a weighted value of a motion vector of the second sampling point as a motion vector of the second target sampling point; or when the at least one second sampling point is at least two second sampling points, use a weighted average value of the motion vectors of the at least two second sampling points as a motion vector of the second target sampling point. 33. Motion vector field decoding apparatus, comprising: a first obtaining module for obtaining prediction information and a residual prediction signal from a current motion vector field block, wherein the motion vector field block current motion vector field is obtained by dividing a current motion vector field into blocks, and the current motion vector field is a motion vector field corresponding to a video frame at time t; a second acquisition module for obtaining a prediction signal from the current motion vector field block in accordance with the prediction information obtained by the first acquisition module; and a calculation module configured to calculate a current motion vector field block reconstruction signal in accordance with the prediction signal obtained by the second acquisition module and the residual prediction signal obtained by the first acquisition module, the apparatus being CHARACTERIZED by the fact that the prediction information comprises direction coefficient information used to indicate a direction coefficient of the current motion vector field block, and the direction coefficient is used to indicate a relationship between a value of a first component of a motion vector of a sampling point of the current motion vector field block and a value of a second component of the motion vector of the sampling point; and the second obtaining module is configured to obtain a reconstruction value of the first component and calculate a predictor of the second component according to the direction coefficient and the reconstruction value of the first component, wherein the prediction signal comprises the predictor of the second component. 34. Decoding apparatus, according to claim 33, CHARACTERIZED by the fact that the direction coefficient information comprises information used to indicate a motion vector field block reconstructed into the current motion vector field, and the direction coefficient direction comprises a direction coefficient of the reconstructed motion vector field block; or the direction coefficient information comprises a direction coefficient value.