Coding optimization method of multi-view spliced video

Abstract

The invention discloses a coding optimization method for a multi-viewpoint spliced video, which comprises the following steps of: S10, inputting a multi-viewpoint video sequence to be coded, determining the number N of viewpoints to be spliced, a key viewpoint K and other viewpoints except the key viewpoint K as auxiliary viewpoints Ai, and calculating the average offset Offset of the viewpoints; s20, according to the number of the viewpoints and the types of the viewpoints, on the basis of the shortest distance principle, all the viewpoints of the previous frame in the time domain are spliced into a large viewpoint frame, the number of the key viewpoints is only one generally, the other N-1 viewpoints are all auxiliary viewpoints, and the shortest distance means that the distance d between the pixel in any auxiliary viewpoint and the pixel of the main viewpoint is minimum; s30, inputting a video sequence formed by the spliced large viewpoint frames into an HEVC / H.265 standard encoder for encoding; and S40, after all the large viewpoint frames are coded, outputting a code stream.

Description

Technical Field

[0001] The invention belongs to the technical field of video encoding, and in particular relates to an encoding optimization method for multi-viewpoint spliced ​​video. Background Art

[0002] In recent years, with the rapid development of multimedia technology, people are no longer satisfied with simple high-definition video applications, but hope to receive and experience video content in a multi-faceted manner through their own senses. This demand has promoted the development of ordinary single-view video content applications to multi-view video applications. Multi-view video refers to video that is recorded by multiple cameras at different angles at the same time. In this type of video application, multi-view video compression is an important research and component. The international standards group has expanded on the basis of the single-view compression standard and released standards for multi-view video. The most advanced one is the MV-HEVC standard based on the HEVC standard extension. Compared with multi-channel simulcast encoding, the technology used in the MV-HEVC standard can reduce the bit rate by 20% to 30% under the same picture quality.

[0003] Unfortunately, in the current industry, standard codecs for single-channel video still occupy most of the market. For cost and robustness considerations, people prefer to use these standard codecs to process various video contents, rather than purchasing additional codecs for corresponding encoding and decoding standards for multi-view video.

[0004] Therefore, it is crucial to establish a method that can process multi-viewpoint videos with a standard codec. The conventional idea in the industry is to splice the viewpoints to be encoded (it may be left-right splicing or top-down splicing), and then treat the spliced ​​video as a single-channel video for encoding and decoding. However, this method is too complex and does not take advantage of the similarity between different viewpoints. In order to further reduce the bit rate of multi-viewpoint videos compressed by this method, current research has reduced the information redundancy between different viewpoints by using the intra-frame block copy algorithm of screen content encoding to reduce the video bit rate as much as possible. However, this method has two disadvantages: First, the intra-frame block copy algorithm is not accurate and has a large complexity, which is not suitable for application in industrial scenarios. Second, it only provides a solution for dual-viewpoint splicing, but does not provide an optimization method when more viewpoint splicing content appears. Summary of the invention

[0005] In view of the above problems, the present invention provides a coding optimization method for multi-view spliced ​​video, which is used to utilize a reasonable viewpoint splicing scheme combined with an optimized intra-frame block copy mode to significantly save the bit rate of encoding multi-view spliced ​​content using a standard encoder.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A coding optimization method for multi-view splicing video, comprising the following steps:

[0008] S10, input the multi-view video sequence to be encoded, determine the number of viewpoints N to be spliced, and the key viewpoint K, and the remaining viewpoints other than the key viewpoint K are auxiliary viewpoints A i , calculate the average viewpoint offset Offset;

[0009] S20, according to the number and type of viewpoints, based on the shortest distance principle, all viewpoints of the previous frame in the time domain are spliced ​​into a large viewpoint frame. The number of key viewpoints is usually only one, and the remaining N-1 viewpoints are all auxiliary viewpoints. The shortest distance means that the distance d between the pixel in any auxiliary viewpoint and the pixel in the main viewpoint is the smallest;

[0010] S30, inputting the video sequence composed of the spliced ​​large viewpoint frames into the HEVC / H.265 standard encoder for encoding, no matter how many viewpoints are spliced, encoding starts from the starting position of the key viewpoint frame first; the encoding order is based on raster scanning, and rate distortion optimization is performed in units of CTU in each viewpoint. For the area where the main viewpoint frame is located, the intra-frame and inter-frame predictions in the encoding process are consistent with the standard encoding process; for the area where the auxiliary viewpoint frame is located, the intra-frame and inter-frame predictions involved in the standard encoding process are performed, and the intra-frame block copy mode based on the average viewpoint offset Offset is performed to use the similarity between the main viewpoint and the auxiliary viewpoint for predictive encoding; for the boundary area where the main viewpoint and the auxiliary viewpoint are spliced, after encoding, sample adaptive compensation and deblocking filter technology are additionally performed to reduce edge discontinuity and block effects;

[0011] S40: After all large viewpoint frames are encoded, the code stream is output.

[0012] In a possible implementation manner, the calculation of the average viewpoint offset Offset in S10 specifically includes:

[0013] Perform discrete Fourier transform DFT on all viewpoints of the first frame of the video sequence to convert the image from the spatial domain to the frequency domain. In the frequency domain, take the key viewpoint K and the auxiliary viewpoint A i , calculate the Fourier transform results F of these two images K (u,v) and Calculate the ratio of their phase spectra:

[0014]

[0015] Among them, F represents the frequency domain representation of the signal, and u and v in F(u,v) represent the image signal in the horizontal and vertical directions respectively. yes The complex conjugate of , the symbol |·| represents the magnitude of a complex number;

[0016] Then, the ratio is subjected to a two-dimensional inverse Fourier transform IDFT to obtain the phase correlation function, also known as the dot product graph r i (x,y):

[0017]

[0018] The peak position of the dot product image represents the relative displacement between the two images. In order to find this position, traverse the entire dot product image and find the position of its maximum value to obtain the relative offset of the two images:

[0019]

[0020] The relative offset of the i-th auxiliary viewpoint relative to the key viewpoint K is calculated, and the average relative offset Offset of the current frame is:

[0021]

[0022] This value reflects the average absolute offset between the key viewpoint and the auxiliary viewpoints, in pixels.

[0023] In a possible implementation manner, executing the intra-block copy mode based on the viewpoint average offset Offset in S30 specifically includes:

[0024] The coded content in the current coding frame is used as the reference area to perform inter-frame predictive coding, which can be specifically described as follows: Assume that the coded image frame is F, where each block to be coded can be marked as B i (x,y), then the goal of intra-block copying is to find B . (x′,y′), such that:

[0025] B i (x,y)=B . (x′,y′),(x′,y′)∈(F 已编码 )

[0026] Since the pixel differences of image blocks cannot be completely consistent, the above formula is equivalent to:

[0027]

[0028] That is, find the block that makes Dif as close to 0 as possible. If there are multiple such blocks, further consider the distance between the two blocks, and use the block with the shortest distance as the final selected copy block, or the one with the least bits required for the encoding block vector, which can be further expressed as the following optimization goal:

[0029]

[0030] Where BV represents the block vector (xx′, yy′), bit(BV) represents the bits required to encode the block vector; δ is a trade-off parameter used to determine the best block among multiple candidate blocks;

[0031] Considering that at the end of the coded frame, the coded area will accumulate to a relatively large area, which increases the complexity of subsequent search, a fast search method is adopted as follows: according to the position of the area processed by the current intra-block copy mode, the position corresponding to the main viewpoint is found; based on the position information, it is added to the offset to obtain a widened area, which is the reference area for fast search optimization of intra-block copy;

[0032] After the intra-frame block copy is completed, that is, after obtaining the optimal block to be copied and the corresponding BV as described above, the predicted coding residual R and the searched block vector BV are encoded into the bitstream; the coding residual refers to the pixel difference between the coding block to be coded and the copied block, that is:

[0033] R(x,y)=B i (x,y)-B . (x′,y′)

[0034] In terms of storage method, for each auxiliary viewpoint, only the first BV is fully encoded, and the remaining BVs are stored in raster order, and only the difference between its BV and the first stored BV is encoded to minimize the overall bit rate of the scheme.

[0035] In a possible implementation, based on the position information, adding the Offset to obtain a widened area specifically includes: if the main viewpoint and the auxiliary viewpoint are spliced ​​in a vertical manner, then the horizontal component x is added to the Offset.

[0036] In a possible implementation, based on the position information, the information is added to the Offset to obtain a widened area. Specifically, if the main viewpoint and the auxiliary viewpoint are spliced ​​in a vertical manner, the vertical component y needs to be added to the Offset.

[0037] In a possible implementation, based on the position information, the information is added to the Offset to obtain a widened area. Specifically, if the main viewpoint and the auxiliary viewpoint are diagonally spliced, the horizontal component x and the vertical component y are both added to the Offset.

[0038] The present invention has the following beneficial effects:

[0039] (1) The high-precision fast intra-frame block copy mode can fully utilize the correlation between different viewpoints to significantly reduce the complexity and bit rate of encoding auxiliary viewpoints, while reducing the complexity of encoding, making it more suitable for industrial applications.

[0040] (2) It can adapt to the splicing of N viewpoints and provides a variety of suitable splicing methods to minimize the complexity of encoding and save bit rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The present invention is a flowchart of a method for optimizing encoding of multi-view spliced ​​videos according to an embodiment of the present invention.

[0042] Figure 2 Schematic diagram of the optimal stitching method under 2 viewpoints.

[0043] Figure 3 Schematic diagram of the optimal stitching method under 3 viewpoints.

[0044] Figure 4 Schematic diagram of the optimal stitching method under 4 viewpoints.

[0045] Figure 5 Schematic diagram of the optimal stitching method under 5 viewpoints.

[0046] Figure 6 Schematic diagram of the optimal stitching method under 7 viewpoints.

[0047] Figure 7 Schematic diagram of search area limitation for fast intra block copy using dual viewpoints as an example.

[0048] Figure 8 Schematic diagram of parallel encoding strategy. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] See also Figure 1 , which is a flowchart of a method for optimizing encoding of a multi-view spliced ​​video according to an embodiment of the present invention, comprising the following steps:

[0051] S10, input the multi-view video sequence to be encoded, according to the preset number of viewpoints N to be spliced, and the key viewpoint K, the remaining viewpoints other than the key viewpoint K are all auxiliary viewpoints A i, calculate the average viewpoint offset Offset; the key viewpoint serves as a reference viewpoint for the auxiliary viewpoint, and the relevant analysis and prediction of the auxiliary viewpoint are all carried out around the key viewpoint. The number of viewpoints to be spliced ​​is pre-set manually. The viewpoint types include key viewpoints and auxiliary viewpoints. The key viewpoint is also called the main viewpoint, which is also determined manually, and the rest of the viewpoints are auxiliary viewpoints.

[0052] S20, according to the number and type of viewpoints, based on the shortest distance principle, all viewpoints of the previous frame in the time domain are spliced ​​into a large viewpoint frame. The number of key viewpoints is usually only one, and the remaining N-1 viewpoints are all auxiliary viewpoints. The shortest distance means that the distance d between the pixel in any auxiliary viewpoint and the pixel in the main viewpoint is the smallest;

[0053] S30, inputting the video sequence composed of the spliced ​​large viewpoint frames into the HEVC / H.265 standard encoder for encoding, no matter how many viewpoints are spliced, encoding starts from the starting position of the key viewpoint frame first; the encoding order is based on raster scanning, and rate distortion optimization is performed in units of CTU in each viewpoint. For the area where the main viewpoint frame is located, the intra-frame and inter-frame predictions in the encoding process are consistent with the standard encoding process; for the area where the auxiliary viewpoint frame is located, the intra-frame and inter-frame predictions involved in the standard encoding process are performed, and the intra-frame block copy mode based on the average viewpoint offset Offset is performed to use the similarity between the main viewpoint and the auxiliary viewpoint for predictive encoding; for the boundary area where the main viewpoint and the auxiliary viewpoint are spliced, after encoding, sample adaptive compensation and deblocking filter technology are additionally performed to reduce edge discontinuity and block effects;

[0054] Intra-frame prediction refers to prediction within a single frame, which is mainly used to reduce spatial redundancy within the same frame. Specifically, it uses the adjacent pixel blocks (macroblocks or coding blocks) that have been encoded in the current frame to predict the content of the blocks that have not yet been encoded. By predicting similar areas, the need to store or transmit each pixel can be reduced. Inter-frame prediction uses the temporal redundancy of the video, that is, it uses the similarity between the previous and next frames to make predictions. It predicts the pixel blocks of the current frame by referring to other frames (usually the previous or next frames), thereby reducing temporal redundancy. Inter-frame prediction can significantly improve compression efficiency because adjacent frames in a video are usually very similar in content.

[0055] The Sample Adaptive Offset (SAO) process is divided into two stages: classification and compensation. In the classification stage, SAO divides pixels into different categories and adjusts the offset value for each category. The purpose of classification is to find pixels with similar characteristics (such as edges, flat areas) and calculate appropriate compensation for them. SAO provides two classification methods: edge classification and bandwidth classification. The former classifies pixels into different edge directions (such as horizontal, vertical, diagonal, etc.) by checking the directional changes in the neighborhood around the pixel; the latter divides the pixel values ​​according to predefined intervals. For example, pixel values ​​can be divided into multiple bandwidths according to brightness values, and different offset values ​​can be used for pixels in each bandwidth. In the compensation stage, SAO calculates the offset value for each category or bandwidth based on the results of the classification stage and applies these offset values ​​to the pixels of the reconstructed image. The purpose of the offset value is to correct the quantization error of the reconstructed image to make it closer to the original image. The working process of the deblocking filter (DF) is mainly divided into three stages. The first stage is to detect the distortion of the block boundary and determine which boundaries need to be smoothed. The second stage applies adaptive filtering. After detecting the block boundary that needs to be processed, the filter will select different filtering strengths based on the characteristics of the block boundary (such as whether it belongs to the high-frequency area and the strength of the boundary). The strength of the filter usually depends on the following factors: quantization parameters, block motion vectors, and pixel differences within and between blocks. Finally, DF makes the transition between adjacent blocks smoother by adjusting the pixel values ​​near the block boundary. It usually processes separately on the vertical and horizontal boundaries of the block.

[0056] S40: After all large viewpoint frames are encoded, the code stream is output.

[0057] In a specific application example of the present invention, a coding optimization method for a multi-viewpoint spliced ​​video is provided, wherein calculating the average viewpoint offset Offset in S10 specifically includes:

[0058] Perform discrete Fourier transform DFT on all viewpoints of the first frame of the video sequence to convert the image from the spatial domain to the frequency domain. In the frequency domain, take the key viewpoint K and the auxiliary viewpoint A i , calculate the Fourier transform results F of these two images K (u,v) and Calculate the ratio of their phase spectra:

[0059]

[0060] Among them, F represents the frequency domain representation of the signal, and u and v in F(u,v) represent the image signal in the horizontal and vertical directions respectively. yes The complex conjugate of , the symbol |·| represents the magnitude of a complex number;

[0061] Then, the ratio is subjected to a two-dimensional inverse Fourier transform IDFT to obtain the phase correlation function, also known as the dot product graph r i (x,y):

[0062]

[0063] The peak position of the dot product image represents the relative displacement between the two images. In order to find this position, traverse the entire dot product image and find the position of its maximum value to obtain the relative offset of the two images:

[0064]

[0065] The relative offset of the i-th auxiliary viewpoint relative to the key viewpoint K is calculated, and the average relative offset Offset of the current frame is:

[0066]

[0067] This value reflects the average absolute offset between the key viewpoint and the auxiliary viewpoint, in pixels. Note that the scope of application of the embodiment of the present invention is only for scenes captured when the camera shooting angle does not change much or moves horizontally, and other scenes with large changes are not applicable to the embodiment of the present invention. Generally speaking, the shooting angle does not change much means that the shooting angles of adjacent cameras are less than or equal to 15°, and larger means greater than 15°.

[0068] In a specific application example of the present invention, a coding optimization method for a multi-view spliced ​​video is provided, in S20, Figures 2 to 6 The optimal stitching methods for 2, 3, 4, 5, and 7 viewpoints are shown respectively. They are all based on the shortest distance principle mentioned above, and the distance between the auxiliary viewpoints and the main viewpoint is minimized as much as possible. View 0 is the key viewpoint, and the rest are auxiliary viewpoints. The dotted part is filled with gray pixels to complete the image and does not participate in the actual encoding and decoding. The distance d is calculated as follows:

[0069]

[0070] Obviously, when the distance between any pixels of the two blocks is the minimum of the width and height, the pixel distance d between the two blocks is the minimum.

[0071] In a specific application example of the present invention, a coding optimization method for a multi-view spliced ​​video is implemented. In S30, the intra-block copy mode based on the viewpoint average offset Offset is specifically implemented as follows:

[0072] The coded content in the current coding frame is used as the reference area to perform inter-frame predictive coding. In the present invention, the intra-frame block copy mode only acts on the rate-distortion optimization process of the auxiliary viewpoint, and its reference area is the main viewpoint that has been coded. In order to reduce the complexity of this mode, the reference area will be limited based on the Offset calculated in step 1, such as Figure 7 As shown in the figure, taking dual viewpoints as an example, to limit the search area of ​​fast intra-block copy, based on the area information of the position (x, y), the horizontal component x is added to the Offset to obtain a widened area (x+Offset, y). It should be noted that if the main viewpoint and the auxiliary viewpoint are spliced ​​up and down, the vertical component y needs to be added to the Offset; if the main viewpoint and the auxiliary viewpoint are spliced ​​diagonally, the horizontal component x and the vertical component y must be added to the Offset.

[0073] Assume that the coded image frame is F, where each block to be coded can be marked as B i (x,y), then the goal of intra-block copying is to find B . (x′,y′), such that:

[0074] B i (x,y)=B . (x′,y′),(x′,y′)∈(F 已编码 )

[0075] Since the pixel differences of image blocks cannot be completely consistent, the above formula is equivalent to:

[0076]

[0077] That is, find the block that makes Dif as close to 0 as possible. If there are multiple such blocks, further consider the distance between the two blocks, and use the block with the shortest distance as the final selected copy block, or the one with the least bits required for the encoding block vector, which can be further expressed as the following optimization goal:

[0078]

[0079] Where BV represents the block vector (xx′, yy′), bit(BV) represents the bits required to encode the block vector; δ is a trade-off parameter used to determine the best block among multiple candidate blocks;

[0080] Considering that at the end of the coded frame, the coded area will accumulate to a relatively large area, which increases the complexity of subsequent search, a fast search method is adopted as follows: according to the position of the area processed by the current intra-block copy mode, the position corresponding to the main viewpoint is found; based on the position information, it is added to the offset to obtain a widened area, which is the reference area for fast search optimization of intra-block copy;

[0081] After the intra-frame block copy is completed, that is, after obtaining the optimal block to be copied and the corresponding BV as described above, the predicted coding residual R and the searched block vector BV are encoded into the bitstream; the coding residual refers to the pixel difference between the coding block to be coded and the copied block, that is:

[0082] R(x,y)=B i (x,y)-B . (x′,y′)

[0083] In terms of storage method, for each auxiliary viewpoint, only the first BV is fully encoded, and the remaining BVs are stored in raster order, and only the difference between its BV and the first stored BV is encoded to minimize the overall bit rate of the scheme.

[0084] In addition, in the application examples of the present invention, a parallel processing strategy is also used to reduce the complexity of the algorithm. Figure 8 As shown, when the main viewpoint has completed two lines of encoding, its auxiliary viewpoint can start encoding.

[0085] Table 1 Bit rate savings and encoding speed changes of dual-view and triple-view test videos after applying the optimization algorithm

[0086]

[0087] In order to verify the results of the method of the embodiment of the present invention, considering that most of the current multi-viewpoint video sources are dual-viewpoint or triple-viewpoint videos, Table 1 shows the bit rate savings and encoding speed changes of the dual-viewpoint and triple-viewpoint test videos after applying the optimization algorithm. The comparison object is the encoding results of these video sources in the standard codec. The BD-Rate indicator represents the bit rate savings, and the negative value represents the percentage of bit rate savings under the same objective quality. It can be clearly observed that the optimization scheme of the method of the embodiment of the present invention has a significant bit rate improvement effect regardless of the All-Intra or RandomAccess mode.

[0088] It should be understood that the exemplary embodiments described herein are illustrative rather than restrictive. Although one or more embodiments of the present invention have been described in conjunction with the accompanying drawings, it should be understood by those skilled in the art that various changes in form and detail may be made without departing from the spirit and scope of the present invention as defined by the appended claims.

Claims

1. A coding optimization method for multi-view splicing video, characterized in that: The following steps are involved: S10, input the multi-view video sequence to be encoded, determine the number of viewpoints N to be spliced, and the key viewpoint K, and the remaining viewpoints other than the key viewpoint K are auxiliary viewpoints A i , calculate the average viewpoint offset Offset; S20, according to the number and type of viewpoints, based on the shortest distance principle, all viewpoints of the previous frame in the time domain are spliced ​​into a large viewpoint frame. The number of key viewpoints is usually only one, and the remaining N-1 viewpoints are all auxiliary viewpoints. The shortest distance means that the distance d between the pixel in any auxiliary viewpoint and the pixel in the main viewpoint is the smallest; S30, inputting the video sequence composed of the spliced ​​large viewpoint frames into the HEVC / H.265 standard encoder for encoding, no matter how many viewpoints are spliced, encoding starts from the starting position of the key viewpoint frame first; the encoding order is based on raster scanning, and rate distortion optimization is performed in units of CTU in each viewpoint. For the area where the main viewpoint frame is located, the intra-frame and inter-frame predictions in the encoding process are consistent with the standard encoding process; for the area where the auxiliary viewpoint frame is located, the intra-frame and inter-frame predictions involved in the standard encoding process are performed, and the intra-frame block copy mode based on the average viewpoint offset Offset is performed to use the similarity between the main viewpoint and the auxiliary viewpoint for predictive encoding; for the boundary area where the main viewpoint and the auxiliary viewpoint are spliced, after encoding, sample adaptive compensation and deblocking filter technology are additionally performed to reduce edge discontinuity and block effects; S40: After all large viewpoint frames are encoded, the code stream is output.

2. The coding optimization method for multi-view spliced ​​video according to claim 1, characterized in that: The calculation of the average viewpoint offset Offset in S10 specifically includes: Perform discrete Fourier transform DFT on all viewpoints of the first frame of the video sequence to convert the image from the spatial domain to the frequency domain. In the frequency domain, take the key viewpoint K and the auxiliary viewpoint A i , calculate the Fourier transform results F of these two images K (u,v) and Calculate the ratio of their phase spectra: Among them, F represents the frequency domain representation of the signal, and u and v in F(u,v) represent the image signal in the horizontal and vertical directions respectively. yes The complex conjugate of , the symbol |·| represents the magnitude of a complex number; Then, the ratio is subjected to a two-dimensional inverse Fourier transform IDFT to obtain the phase correlation function, also known as the dot product graph r i (x,y): The peak position of the dot product image represents the relative displacement between the two images. In order to find this position, traverse the entire dot product image and find the position of its maximum value to obtain the relative offset of the two images: The relative offset of the i-th auxiliary viewpoint relative to the key viewpoint K is calculated, and the average relative offset Offset of the current frame is: This value reflects the average absolute offset between the key viewpoint and the auxiliary viewpoints, in pixels.

3. The coding optimization method for multi-view spliced ​​video according to claim 2, characterized in that: The intra-block copy mode based on the viewpoint average offset Offset in S30 specifically includes: The coded content in the current coding frame is used as the reference area to perform inter-frame predictive coding, which can be specifically described as follows: Assume that the coded image frame is F, where each block to be coded can be marked as B i (x,y), then the goal of intra-block copying is to find B . (x / ,y / ), such that: B i (x,y)=B . (x / ,y / ),(x′,y′)∈(F 已编码 ) Since the pixel differences of image blocks cannot be completely consistent, the above formula is equivalent to: That is, find the block that makes Dif as close to 0 as possible. If there are multiple such blocks, further consider the distance between the two blocks, and use the block with the shortest distance as the final selected copy block, or the one with the least bits required for the encoding block vector, which can be further expressed as the following optimization goal: Where BV represents the block vector (xx / ,yy′), bit(BV) represents the bits required to encode the block vector; δ is a trade-off parameter used to determine the best block among multiple candidate blocks; Considering that at the end of the coded frame, the coded area will accumulate to a relatively large area, which increases the complexity of subsequent search, a fast search method is adopted as follows: according to the position of the area processed by the current intra-block copy mode, the position corresponding to the main viewpoint is found; based on the position information, it is added to the offset to obtain a widened area, which is the reference area for fast search optimization of intra-block copy; After the intra-frame block copy is completed, that is, after obtaining the optimal block to be copied and the corresponding BV as described above, the predicted coding residual R and the searched block vector BV are encoded into the bitstream; the coding residual refers to the pixel difference between the coding block to be coded and the copied block, that is: R(x,y)=B i (x,y)-B . (x / ,y / ) In terms of storage method, for each auxiliary viewpoint, only the first BV is fully encoded, and the remaining BVs are stored in raster order, and only the difference between its BV and the first stored BV is encoded to minimize the overall bit rate of the scheme.

4. The coding optimization method for multi-view spliced ​​video according to claim 3, characterized in that: Based on the position information, the information is added to the Offset to obtain a widened area. Specifically, if the main viewpoint and the auxiliary viewpoint are spliced ​​in a vertical manner, the horizontal component x is added to the Offset.

5. The coding optimization method for multi-view spliced ​​video according to claim 3, characterized in that: Based on the position information, the offset is added to obtain a widened area. Specifically, if the main viewpoint and the auxiliary viewpoint are spliced ​​up and down, the vertical component y needs to be added to the offset.

6. The coding optimization method for multi-view spliced ​​video according to claim 3, characterized in that: Based on the position information, the Offset is added to obtain a widened area. Specifically, if the main viewpoint and the auxiliary viewpoint are diagonally spliced, the horizontal component x and the vertical component y must be added to the Offset.

Images