WO2018072487A1 - 全景视频感兴趣区域的描述方法和编码方法 - Google Patents

全景视频感兴趣区域的描述方法和编码方法 Download PDF

Info

Publication number
WO2018072487A1
WO2018072487A1 PCT/CN2017/092552 CN2017092552W WO2018072487A1 WO 2018072487 A1 WO2018072487 A1 WO 2018072487A1 CN 2017092552 W CN2017092552 W CN 2017092552W WO 2018072487 A1 WO2018072487 A1 WO 2018072487A1
Authority
WO
WIPO (PCT)
Prior art keywords
interest
region
panoramic
layer
center
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2017/092552
Other languages
English (en)
French (fr)
Inventor
王振宇
王荣刚
王悦名
高文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Peking University Shenzhen Graduate School
Original Assignee
Peking University Shenzhen Graduate School
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Peking University Shenzhen Graduate School filed Critical Peking University Shenzhen Graduate School
Priority to US16/330,923 priority Critical patent/US10681355B2/en
Publication of WO2018072487A1 publication Critical patent/WO2018072487A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
    • H04N19/167—Position within a video image, e.g. region of interest [ROI]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/124—Quantisation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
    • H04N19/146—Data rate or code amount at the encoder output
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/172—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a picture, frame or field
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/174—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a slice, e.g. a line of blocks or a group of blocks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/597—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding specially adapted for multi-view video sequence encoding

Definitions

  • the present invention relates to the field of digital video codec technology, and in particular, to a method and a coding method for a panoramic video region of interest based on a spherical multi-layer circumference.
  • panoramic video is an important part. Since panoramic video records the entire picture of a 360-degree view with a very high amount of data, compression of panoramic video is a key technology in virtual reality applications.
  • the panoramic video Since the panoramic video records the entire picture of the 360-degree view, while a single user is watching, only a part of the panoramic video picture can be seen at the same time.
  • the traditional coding method is to perform the same quality or similar quality coding on different regions of the panoramic video picture without distinction, and generate a code stream and transmit it. Therefore, the existing conventional code transmission method has great redundancy.
  • one feasible method is to generate multiple code streams for panoramic video coding; different code streams focus on different areas of the panoramic video picture, ie, the region of interest; a single code stream, only for the region of interest High quality high bit rate coding, other areas use low quality low bit rate coding.
  • the code rate of a single code stream is much lower than that of the conventional coding method, and the same coding quality can be maintained in the region of interest, which greatly reduces the transmission cost of the panoramic video.
  • the existing description method is inefficient and cannot be adapted to the above-mentioned effective description of the region of interest for the panoramic video and meets the corresponding coding requirements. Therefore, there is currently no reasonable description and coding method for the region of interest of the panoramic video. .
  • the present invention provides a method and an encoding method for a panoramic video region of interest based on a spherical multi-layer circumference, which can reasonably describe a region of interest of a single panoramic video stream and an efficient code rate. distribution.
  • the core idea of the present invention is that the description information of the region of interest is written to the sequence header of a single code stream, the information described is the information of the current single code stream, and the different code streams have their respective regions of interest description.
  • the same region of interest with the same center of the multi-layer is defined as a multi-layered circumference with the same center and different radii on the panoramic spherical surface with the sphere radius of 1; the center and radius of the circumference, as defined by the spherical geometry , are located on the panoramic spherical surface; write the coordinates of the center, the number of circumferential layers N, and the radius of each layer into the sequence header of the code stream, thereby indicating the same region of interest in the multi-layer center.
  • the center of the region of interest is the center C of the circumference of the panoramic spherical surface, and the coordinates of the center can be expressed by latitude and longitude (C x , C y );
  • the size of the region of interest is the radius R of the circumference of the spherical surface on the panoramic spherical surface (the first The radius of interest of the n layer is denoted by R n );
  • the radius of the circumference on the panoramic spherical surface is equal to the arc of the angle between the two ends of the circumferential radius of the panoramic sphere and the connection of the spherical center, so it can also be used
  • the angle value corresponding to the arc is represented.
  • the nth layer QP adjustment value is recorded as ⁇ QP n , and ⁇ QP n ⁇ 0
  • the QP adjustment value is added to the initial value of the image block QP obtained by the conventional method to obtain the actual QP of the image block coding; according to the mapping mode of the current panorama and the coordinates of the image block center point A on the image, A is calculated.
  • the spherical distance D can be expressed by the angle value corresponding to the arc.
  • Another method for encoding a panoramic video region of interest specifically for each point, firstly calculating the distance D' between the current point B and the center point C of the region of interest on the panoramic spherical surface by using the above method, and determining the priority of point B according to the distance D' Level P, and according to D' and P select filter or filter strength, low-pass filtering the current point B, and finally encoding the filtered whole frame image in the traditional way.
  • a method for describing a panoramic video region of interest based on a spherical multi-layer circumference representing a panoramic video region of interest based on a multi-layered circumference on a panoramic spherical surface, setting a multi-layer region of interest with a circumferential layer number of N and a center
  • N-layer multi-layer
  • the coordinates of the center, the number of circumferential layers N, and the radius of each layer are written into the sequence header of the code stream. That is, the description of the multi-layer region of interest of the panoramic video is completed.
  • the above method is based on a multi-layered circumference on a panoramic sphere to represent a panoramic video multi-layer region of interest, including the following steps:
  • the sphere radius of the panoramic sphere is defined as 1 unit (for example, the radius is 1), and the panoramic video region of interest is defined as the spherical region surrounded by the circumference on the panoramic sphere;
  • A2) Describe the center of the region of interest using the latitude and longitude (C x , C y ) of the panoramic sphere at which the center of the circle is located; use the radius of the circle on the panoramic sphere to describe the size of the region of interest.
  • the circumference refers to the "circumference on the spherical surface of the panorama".
  • the center or radius mentioned here should be understood as defined in the spherical geometry.
  • the 60° north latitude weft coil on the earth in the spherical geometry, the center of the weft coil is the north pole point, and the radius is the line connecting the earth surface to the north pole point at any point on the north latitude 60°.
  • the center of the latitude 60° latitude coil is in the earth. Internal, the radius is not on the surface of the earth. Therefore, when referring to the center of the circle and the radius in the present invention, it should be the center or radius of the "on the spherical surface of the panorama".
  • the circumference on the sphere is a point on the circumference, along the sphere, reaching the shortest line in the center of the circumference of the sphere.
  • This connection is an arc for the ball, or in the three-dimensional space.
  • the radius of the circumference on the panoramic spherical surface corresponds to an arc
  • the radius length is the arc of the angle between the ends of the radius and the connecting line of the sphere, so The angle corresponding to the arc is used to indicate the size of the region of interest.
  • N-layer regions of interest of different priorities using N-layer circumferences of the same center and different radii; wherein the region inside the circle with the smallest radius is the highest priority; otherwise, the region inside the circle with the second smallest radius For the second priority, and so on.
  • the present invention represents a plurality of regions of interest by the above method.
  • the size of the nth layer of interest is recorded as R n .
  • the center of the region of interest, the number of layers, the size of each layer, and the like are written into the sequence header of the code stream.
  • the multi-layer region of interest of the panoramic video is described, firstly, the latitude and longitude (C x , C y ) of the panoramic spherical surface at the center of the circumference on the panoramic spherical surface is set as the center of the panoramic video region of interest;
  • the number of layers of the region of interest is N;
  • the size of the region of interest of the current layer R n is obtained by the above method (the radius of the circumference on the spherical surface of the sphere);
  • the size is obtained, the center of the region of interest, the number of layers, the size of each layer, and the like are written into the sequence header of the code stream, that is, the description of the multi-layer region of interest of the panoramic video is completed.
  • the present invention provides an encoding method for a panoramic video region of interest based on a spherical multi-layer circumference.
  • QP quantization step size
  • the spherical radius is defined as a panoramic spherical surface of unit 1.
  • the spherical distance D of A and C is the arc of the angle between the two points A and C and the line connecting the spherical center. Therefore, the angle corresponding to the arc may also be used to represent the distance. D; if D is less than R 1 , the QP adjustment value of the image block is ⁇ QP 1 ; otherwise, if D is greater than R N , the QP adjustment value of the image block is 0; otherwise, if D is smaller than R n and larger than R n -1 (n>1), the QP adjustment value of the image block is ⁇ QP n ;
  • the initial QP plus QP adjustment value of the current image block is the coded QP of the current image block
  • the current block is encoded by a conventional coding method flow
  • It can be encoded by the current common video coding standard, including mpeg2, h264, AVS, and the latest generation of H265, AVS2, and so on.
  • Another embodiment of the present invention further provides another encoding method for a multi-layer region of interest, including the following steps:
  • D' is less than R 1 , then point B is in the layer 1 region of interest, and its priority is defined as 1; otherwise, if D' is greater than R N , point B is outside the region of interest, and its priority is defined as N +1; otherwise, if D' is less than R n and greater than R n-1 (n>1), then point B is in the nth layer of interest, and its priority is defined as n;
  • the filtered image is encoded using the existing conventional encoding method, and can be used. Any existing encoder encodes the filtered image.
  • the existing video coding method is generally encoded by: segmenting an image and encoding each block; wherein, when encoding the current block, first predicting the current block to obtain a prediction block of the current block; The pixel value of the image around the current block predicts the pixel value of each pixel of the current block to obtain a prediction block, and may also use the image block on the encoded image to predict the current image block to obtain a prediction block; and then subtract the pixel value of the current block from the prediction block. Obtaining a residual block; then transforming the residual block to obtain a transform block, and performing quantization operations on the transform block to obtain a quantized block; finally, writing the value of each point in the quantized block to the code stream.
  • the invention provides a method and a coding method for a panoramic video region of interest based on a spherical multi-layer circumference.
  • the description method is based on a multi-layer circumference on a panoramic spherical surface to represent a panoramic video region of interest as a panoramic video multi-layer region of interest;
  • the method can perform flexible rate allocation on the multi-region region of interest of the panoramic video; while ensuring high image quality of the region of interest, the technical solution of the invention can greatly reduce the code rate required for encoding and transmission.
  • FIG. 1 is a flow chart of a method for describing a region of interest of a panoramic video provided by the present invention.
  • FIG. 2 is a flow chart of a first method for encoding a panoramic video region of interest according to the present invention.
  • FIG. 3 is a flow chart of a second method for encoding a panoramic video region of interest provided by the present invention.
  • FIG. 4 is a schematic diagram of partitioning a region of interest of a panoramic video in an embodiment of the present invention.
  • each region of interest defines two regions of interest with radii of 60 degrees and 90 degrees, respectively.
  • FIG. 5 is a schematic diagram showing a description of a region of interest of a panoramic video and a set of QP adjustment values in a first embodiment of the present invention
  • A is a point on the panoramic sphere, and the corresponding latitude and longitude of the point on the panoramic sphere is (A x , A y ); C is the center of the region of interest, C is latitude and longitude (0, 0); D is A, C The distance of the point on the panoramic sphere; ⁇ QP 1 is the QP adjustment value of the image block.
  • FIG. 6 is a schematic diagram showing a description of a region of interest of a panoramic video and a representation of a filter size set in Embodiment 2 of the present invention
  • B is a point on the panoramic sphere
  • the corresponding latitude and longitude of the point on the panoramic sphere is (B x , B y );
  • C is the center of the region of interest, C is latitude and longitude (0, 0);
  • Filter Size is the opposite point B
  • the template size used for Gaussian filtering is 3x3 or 5x5 respectively.
  • the invention provides a method and a coding method for a panoramic video region of interest based on a spherical multi-layer circumference.
  • the description method is based on a multi-layer circumference on a panoramic sphere to represent a panoramic video region of interest; the coding method can be interested in a panoramic video multilayer
  • the area performs rate allocation; the technical solution of the present invention can reasonably describe the region of interest of a single panoramic video stream and efficiently allocate rate.
  • the method for describing the region of interest of the panoramic video first describes the panoramic video region of interest using a multi-layered circumference based on the sphere; the panoramic spherical latitude and longitude at which the center of the circle is located to describe the center of the region of interest; and the use of the circle on the panoramic sphere
  • the number of angles of the radius relative to the center of the sphere is used to describe the size of the region of interest; circles of the same priority and different radii can be used to describe regions of interest of different priorities.
  • the method for describing the region of interest of the panoramic video includes the following steps:
  • the current layer is n layer, the size of the region of interest of the current layer is denoted as R n ; the number of layers counter is increased by 1;
  • step A4) performing step A3) cyclically until the size of all N-layer regions of interest are obtained;
  • A5) Write the coordinates of the center of the region of interest of the panoramic video, the number of layers N, and the radius of each layer into the sequence header of the code stream, that is, complete the description of the multi-layer region of interest of the panoramic video.
  • the multi-layer region of interest of the panoramic video is described by the above method.
  • the following two encoding methods can be used for encoding.
  • the encoding method of the first panoramic video region of interest is: calculating a distance between a center point of the image block to be encoded and a center point of the region of interest on the panoramic spherical surface, and selecting a region of interest region in which the image block is located according to the distance And using the corresponding QP adjustment value to adjust the initial QP of the image block to obtain the final coded QP;
  • the second coding method is: calculating the distance of each point in the image to the center point of the region of interest on the panoramic sphere, The level of the region of interest in which the image block is located is selected according to the distance, and the filtering template and filtering intensity of the point are determined according to the distance and level, and filtered, and finally the image is encoded using a conventional method.
  • Embodiment 1 The method and the encoding method for the panoramic video region of interest based on the spherical multi-layer circumference provided by the present invention are used to define the center of the region of interest for a typical panoramic video coding mapped in a latitude and longitude manner.
  • six representative region centers of interest are defined for a typical latitude and longitude map-mapped panoramic video coding.
  • (a)-(f) respectively indicate that the center latitude and longitude are (0,0), (90,0), (-90,0), (180,0), (0,90) and (0,-90) region of interest.
  • Each region of interest region defines two regions of interest with radii of 60 degrees and 90 degrees, respectively. Therefore, for the panoramic video, according to the division of FIGS. 4(a)-(f), six code streams can be encoded.
  • the latitude and longitude (0, 0) of the center of the region of interest, the number of layers of the region of interest (2 layers), and the size of the region of interest of the first layer are required ( 60 degrees), the size of the second layer of interest (90 degrees) and other information is written into the sequence header; encoding the code stream, using the rate control of the fixed image block QP, the initial QP of each image block is set to QP Base ; When encoding an image block, as shown in FIG.
  • the corresponding latitude and longitude (A x , A y ) of the point on the panoramic spherical surface is calculated, according to the latitude and longitude of the point A (A x , A y ) and the center of the region of interest C point latitude and longitude (0, 0) to calculate the distance D of the A and C points on the panoramic sphere; if D is smaller than the size of the first layer of interest (60 degrees), the image
  • the QP adjustment value ( ⁇ QP 1 ) of the block is -6; otherwise, if the value of D is between 60 degrees and 90 degrees, the QP adjustment value ( ⁇ QP 2 ) of the image block is -3, otherwise the image block
  • the QP adjustment value is 0; finally, the coded QP of the image block should be QP Base plus the QP adjustment value of the image block.
  • the latitude and longitude (90,0) of the center point of the region of interest, the number of layers of the region of interest (2 layers), the size of the region of interest of the first layer (60 degrees), and the second layer are required.
  • Information such as the size of the region of interest (90 degrees) is written to the sequence header; the remaining coding operations are the same as those for the code map 4(a).
  • Embodiment 2 is a diagrammatic representation of Embodiment 1:
  • each region of interest region defines two regions of interest with radii of 60 degrees and 90 degrees, respectively. Therefore, for the panoramic video, according to the division of FIGS. 4(a)-(f), six code streams can be encoded.
  • the latitude and longitude (90,0) of the center point of the region of interest, the number of layers of the region of interest (2 layers), the size of the region of interest of the first layer (60 degrees), and the second layer are required.
  • Information such as the size of the region of interest (90 degrees) is written to the sequence header; the rest of the operation is the same as the code stream corresponding to the code of FIG. 4(a).
  • the present invention is based on a multi-layered circumference on a panoramic spherical surface to represent a panoramic video region of interest as a panoramic video multi-layer region of interest; encoding a panoramic video multi-layer region of interest, the encoding method may be interested in panoramic video multilayer
  • the area performs flexible rate allocation; while ensuring that the region of interest has high image quality, the technical solution of the present invention can be large
  • the amplitude reduces the code rate required for encoding and transmission.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

本发明公布了一种基于球面多层圆周的全景视频感兴趣区域的描述方法和编码方法，对全景视频多层感兴趣区域进行描述，首先设定为全景视频感兴趣区域的中心；再设定感兴趣区域的层数为N；通过半径或夹角获得当前层感兴趣区域的大小Rn；全部获取N层感兴趣区域的大小，将感兴趣区域的中心、层数、每一层的大小等信息写入码流的序列头。编码方法通过使用QP调整值对初始QP进行调整或进行滤波，再对图像进行编码。本发明技术方案可对全景视频多层感兴趣区域进行灵活的码率分配；在保证感兴趣区域具有较高的图像质量的同时，本发明技术方案能够大幅度降低编码和传输所需的码率。

Description

全景视频感兴趣区域的描述方法和编码方法 技术领域
本发明涉及数字视频编解码技术领域，尤其涉及一种基于球面多层圆周的全景视频感兴趣区域描述方法和编码方法。
背景技术
目前，虚拟现实技术和相关应用正在快速发展。在虚拟现实技术中，全景视频是一个重要的组成部分。由于全景视频记录了360度视角的全部画面，具有极高的数据量，因此全景视频的压缩是虚拟现实应用中的一个关键技术。
由于全景视频记录了360度视角的全部画面，而单一用户在观看的时候，同一时刻只能看到全景视频画面的一部分。而传统的编码方法是不加区分地对全景视频画面的各个不同区域进行质量相同或质量相似的编码，编码生成一路码流，并进行传输。因此，现有传统的编码传输方法存在很大的冗余。要去除该冗余，一个可行的办法是：对全景视频编码生成多个码流；不同码流侧重编码全景视频画面的不同区域，即感兴趣区域；单个码流，只对其感兴趣区域进行高质量高码率编码，其他区域使用低质量低码率编码。这样，单个码流的码率远低于传统编码方法的码流，同时在其感兴趣区域又可以保持同样的编码质量，大大降低了全景视频的传输代价。但是，现有的描述方法效率较低，无法适应上述针对全景视频的感兴趣区域进行有效描述及达到相应的编码要求，因此，目前尚缺乏合理的针对全景视频的感兴趣区域的描述及编码方法。
发明内容
为了克服上述现有技术的不足，本发明提供一种基于球面多层圆周的全景视频感兴趣区域描述方法和编码方法，可对单个全景视频码流的感兴趣区域进行合理描述及高效的码率分配。
本发明的核心构思是：感兴趣区域的描述信息写入单个码流的序列头，所描述的信息是当前单个码流的信息，而不同码流有各自的感兴趣区域描述。基于球面多层圆周，将多层中心相同的感兴趣区域定义为球半径为单位1的全景球面上中心相同、半径不同的多层圆周； 所述圆周的中心和半径，按球面几何中的定义，均位于全景球面上；将中心的坐标、圆周层数N、每层圆周半径写入码流的序列头，由此表示多层中心相同的感兴趣区域。感兴趣区域中心即为全景球面上圆周的中心C，可以经纬度(Cx,Cy)表示所述中心的坐标；感兴趣区域大小即为全景球面上圆周在全景球面上的半径大小R(第n层感兴趣的半径大小记为Rn)；圆周在全景球面上的半径大小，其值等于该全景球面上的圆周半径的两端同球心连线的夹角的弧度，因此也可以用所述弧度对应的角度值表示。针对全景视频感兴趣区域的编码方法，一种是对位于不同层的感兴趣区域中的图像块使用不同的QP调整值(第n层QP调整值记为ΔQPn，且ΔQPn≤0)，并将QP调整值同通过传统方法得到的图像块QP初始值相加，得到图像块编码的实际QP；根据当前全景图的映射方式和图像块中心点A在图像上的坐标，计算得到A在所述全景球面上的经纬度(Ax,Ay),并根据经纬度(Ax,Ay)和感兴趣区域中心经纬度(Cx,Cy)计算得到A、C点的球面距离D；也可以用所述弧度对应的角度值表示球面距离D。另一种全景视频感兴趣区域编码方法，具体对每个点，首先采用上述方法计算在全景球面上当前点B与感兴趣区域中心点C的距离D’，根据距离D’确定B点的优先级P，并跟据D’和P选择滤波器或滤波强度，对当前点B进行低通滤波，最后对滤波后的整帧图像进行传统方式的编码。
本发明提供的技术方案是：
一种基于球面多层圆周的全景视频感兴趣区域的描述方法，基于全景球面上的多层圆周来表示全景视频感兴趣区域，设定圆周层数为N，将中心相同的多层感兴趣区域表示为球半径为1个单位的全景球面上中心相同、半径不同的多层(N层)圆周；将所述中心的坐标、圆周层数N、每层圆周半径写入码流的序列头，即完成对全景视频多层感兴趣区域的描述。
上述方法基于全景球面上的多层圆周来表示全景视频多层感兴趣区域，包括如下步骤：
A1)将全景球面的球半径定义为1个单位(如半径为1)，全景视频感兴趣区域定义为全景球面上的圆周包围的球面区域；
A2)使用所述圆周的圆心所处全景球面的经纬度(Cx,Cy)来描述感兴趣区域的中心；使用所述圆周在全景球面上的半径来描述感兴趣区域的大小。
本发明中，圆周指的是“全景球面上的圆周”，默认应该用球面几何中对圆的定义，这里提到的圆心或半径都应该按球面几何中定义的来理解。举个例子，地球上北纬60°的纬线圈，在球面几何中，这个纬线圈的中心就是北极点，半径就是北纬60°上的任何一个点沿着地球表面到北极点的连线。但是按普通的欧式几何，在三维空间里，北纬60°纬线圈的圆心是在地球 内部的，半径也不在地球表面。所以本发明中提到圆周中心和半径时，都应是“在全景球面上”的中心或半径。
在球面几何中，球面上的圆周是圆周上一点，沿着球面，到达球面上圆周中心最短的连线。这段连线对球来说，或者在三维空间中来看，就是一段弧线。
因此，对所述球半径为单位1的全景球面，所述圆周在全景球面上的半径对应为一段弧线，半径长度为半径的两端同球心连线的夹角的弧度，因此也可以用所述弧度对应的角度来表示感兴趣区域的大小。
可以使用中心相同，半径不同的N层圆周来描述不同优先级的N层感兴趣区域；其中，位于半径最小的圆周内部的区域为最高优先级；否则，位于半径第二小的圆周内部的区域为第二优先级，以此类推。本发明通过上述方法表示多层感兴趣区域。其中，第n层感兴趣区域的大小记为Rn。感兴趣区域的中心、层数、每一层的大小等信息写入码流的序列头。
具体地，对全景视频多层感兴趣区域进行描述，首先将全景球面上的圆周的中心所处全景球面的经纬度(Cx,Cy)设定为全景视频感兴趣区域的中心；再设定感兴趣区域的层数为N；设定层数计数器初始值n＝1；通过上述方法(圆周在全景球面上的半径)获得当前层感兴趣区域的大小Rn；直到N层感兴趣区域的大小都获取，感兴趣区域的中心、层数、每一层的大小等信息写入码流的序列头，即完成对全景视频多层感兴趣区域的描述。
基于上述描述方法对全景视频感兴趣区域进行的描述，可以采用不同的编码方法来降低非感兴趣区域的码率。本发明提供了基于球面多层圆周的全景视频感兴趣区域的编码方法，在本发明一实施例中，针对采用上述描述方法进行描述的多层感兴趣区域，通过调整每一层感兴趣区域的量化步长(QP)值进行编码，对全景视频多层感兴趣区域进行灵活的码率分配；包括如下步骤：
B1)首先设定感兴趣区域中心点C的经纬度(Cx,Cy)、层数N、每一层的半径(Rn)；设定每一层对应的量化步长(QP)调整值，将第n层QP调整值记为ΔQPn，且ΔQPn≤0；
B2)编码当前图像块时，根据当前图像块中心点A的坐标计算得到A点在全景球面的经纬度(Ax,Ay)，并计算全景球面上A、C两点的距离D；
对所述球半径定义为单位1的全景球面，A、C的球面距离D为A、C两点同球心连线的夹角的弧度，因此也可以用所述弧度对应的角度来表示距离D；如果D小于R1，则该图像块的QP调整值为ΔQP1；否则，如果D大于RN，则该图像块的QP调整值为0；否则，如果 D小于Rn且大于Rn-1(n>1)，则该图像块的QP调整值为ΔQPn；
B3)对当前图像块进行传统的码率控制操作得到当前图像块的初始QP，例如，可直接使用当前图像指定的QP作为当前图像块的初始QP；
B4)当前图像块的初始QP加QP调整值即为当前图像块的编码QP；
B5)根据得到的当前图像块的编码QP，对当前块采用传统的编码方法流程进行编码；
可采用目前通用的视频编码标准进行编码，编码方法包括：mpeg2、h264、AVS，和最新一代的H265、AVS2等。
B6)对图像中所有图像块进行B2-B5的循环操作，直至图像中所有图像块完成编码。
本发明另一实施例还提供了另一种针对多层感兴趣区域的编码方法，包括如下步骤：
C1)首先设定感兴趣区域中心点C的经纬度(Cx,Cy)、层数N、每一层的半径(Rn)；
C2)计算图像上每个点到感兴趣区域中心点C的距离；当前点B同点C的距离定义为D’；得到点B的优先级；
如果D’小于R1，则点B处于第1层感兴趣区域,其优先级定义为1；否则，如果D’大于RN，则点B处于感兴趣区域之外，其优先级定义为N+1；否则，如果D’小于Rn且大于Rn-1(n>1)，则点B处于第n层感兴趣区域，其优先级定义为n；
C3)根据点B同C的距离D’，以及点B的优先级确定滤波模版和滤波强度，并对点B进行低通滤波；滤波之后的图像使用现有的传统编码方法进行编码，可以使用任何现有编码器对所述滤波之后的图像进行编码。
现有的视频编码方法一般通过以下过程进行编码：对图像进行分块，对每一个块进行编码；其中，当编码当前块时，首先对当前块进行预测，得到当前块的预测块；可以使用当前块周围图像的像素值预测当前块每个像素的像素值得到预测块，也可以使用已编码图像上的图像块来预测当前图像块得到预测块；再将当前块的像素值减去预测块，得到残差块；之后对残差块进行变换得到变换块，并对变换块进行量化操作得到量化块；最后将量化块中每个点的值写入码流。
与现有技术相比，本发明的有益效果是：
本发明提供一种基于球面多层圆周的全景视频感兴趣区域描述方法和编码方法，描述方法基于全景球面上的多层圆周来表示全景视频感兴趣区域为全景视频多层感兴趣区域；编码 方法可对全景视频多层感兴趣区域进行灵活的码率分配；在保证感兴趣区域具有较高的图像质量的同时，本发明技术方案能够大幅度降低编码和传输所需的码率。
附图说明
图1是本发明提供的全景视频感兴趣区域描述方法的流程框图。
图2是本发明提供的第一种全景视频感兴趣区域编码方法的流程框图。
图3是本发明提供的第二种全景视频感兴趣区域编码方法的流程框图。
图4是本发明实施例中对全景视频感兴趣区域划分的示意图；
其中，(a)-(f)分别表示中心经纬度为(0,0)、(90,0)、(-90,0)、(180,0)、(0,90)和(0,-90)的感兴趣区域；每个感兴趣区域中心定义了两个半径分别为60度和90度大小的感兴趣区域。
图5是本发明实施例一中的全景视频感兴趣区域描述及设定的QP调整值表示的示意图；
其中，A为全景球面上一点，该点在全景球面上对应的经纬度为(Ax,Ay)；C为感兴趣区域中心，C点经纬度为(0,0)；D为A、C两点在全景球面上的距离；ΔQP1为图像块的QP调整值。
图6是本发明实施例二中的全景视频感兴趣区域描述及设定的滤波器大小表示的示意图；
其中，B为全景球面上一点，该点在全景球面上对应的经纬度为(Bx,By)；C为感兴趣区域中心，C点经纬度为(0,0)；Filter Size为对点B进行高斯滤波使用的模板大小，分别为3x3或5x5。
具体实施方式
下面结合附图，通过实施例进一步描述本发明，但不以任何方式限制本发明的范围。
本发明提供一种基于球面多层圆周的全景视频感兴趣区域描述方法和编码方法，描述方法基于全景球面上的多层圆周来表示全景视频感兴趣区域；编码方法可对全景视频多层感兴趣区域进行码率分配；本发明技术方案可对单个全景视频码流的感兴趣区域进行合理描述及高效的码率分配。
其中，全景视频感兴趣区域的描述方法首先用基于球面的多层圆周来描述全景视频感兴趣区域；使用圆周中心所处的全景球面经纬度来描述感兴趣区域的中心；使用圆周在全景球面上的半径相对球心的夹角度数来描述感兴趣区域的大小；可以使用中心相同，半径不同的圆周来描述不同优先级的感兴趣区域。感兴趣区域的中心、层数、每一层的大小等信息写入 码流的序列头。全景视频感兴趣区域描述方法包括如下步骤：
A1)将全景球面上的圆周的中心所处全景球面的经纬度(Cx,Cy)设定为全景视频感兴趣区域的中心；
A2)再设定感兴趣区域的层数为N；设定层数计数器初始值n＝1；
A3)使用圆周在全景球面上的半径来表示感兴趣区域的大小；当前层为n层，当前层感兴趣区域的大小记作Rn；层数计数器取值加1；
A4)循环执行步骤A3)，直到获取到所有N层感兴趣区域的大小；
A5)将所述全景视频感兴趣区域的中心的坐标、层数N和每层圆周半径都写入码流的序列头，即完成对全景视频多层感兴趣区域的描述。
采用上述方法对全景视频多层感兴趣区域进行描述，针对全景视频多层感兴趣区域，可采用以下两种编码方法进行编码。
第一种全景视频感兴趣区域的编码方法为：计算待编码图像块的中心点同感兴趣区域的中心点在全景球面上的距离，根据所述距离来选择该图像块所处的感兴趣区域层级，并使用相应的QP调整值对该图像块的初始QP进行调整，得到最终的编码QP；第二种编码方法为：计算图像中每个点到感兴趣区域中心点在全景球面上的距离，根据所述距离来选择该图像块所处的感兴趣区域层级，并跟据所述距离和层级确定对该点的滤波模版和滤波强度，并进行滤波，最后使用传统方法对图像进行编码。
实施例一采用本发明提供的基于球面多层圆周的全景视频感兴趣区域的描述方法和编码方法，针对典型的以经纬图方式映射的全景视频编码定义感兴趣区域中心。该实施例中，针对典型的以经纬图方式映射的全景视频编码，定义了6个感兴趣区域中心。如图4所示，(a)-(f)分别表示了中心经纬度为(0,0)，(90,0)，(-90,0)，(180,0)，(0,90)和(0,-90)的感兴趣区域。每个感兴趣区域中心定义了2个半径分别为60度和90度大小的感兴趣区域。因此，对该全景视频，根据图4(a)-(f)的划分，可以编码生成6个码流。
对应图4(a)的码流，如图5所示，需要将感兴趣区域中心C点经纬度(0,0)、感兴趣区域层数(2层)、第一层感兴趣区域的大小(60度)、第二层感兴趣区域的大小(90度)等信息写入序列头；编码该码流，使用固定图像块QP的码率控制，每一个图像块的初始QP设为QPBase；编码某个图像块时，如图5所示，根据该图像块中心点A在图像上的坐标，计算该点在全景球面上对应的经纬度(Ax,Ay),根据A点经纬度(Ax,Ay)和感兴趣区域中心C点经纬度(0,0)计算A、C点在全景球面上的距离D；如果D小于第一层感兴趣区域的大小(60度)，则该图像块 的QP调整值(ΔQP1)为-6；否则，如果D的值介于60度和90度之间，则该图像块的QP调整值(ΔQP2)为-3，否则该图像块的QP调整值为0；最后，该图像块的编码QP应为QPBase加上该图像块的QP调整值。
对应图4(b)的码流，需要将感兴趣区域中心点经纬度(90,0)、感兴趣区域层数(2层)、第一层感兴趣区域的大小(60度)、第二层感兴趣区域的大小(90度)等信息写入序列头；其余编码操作同编码图4(a)对应的码流相同。
对应图4(c)-(f)的码流，同理，将各自感兴趣区域中心点经纬度、感兴趣区域层数、每层感兴趣区域大小等信息写入序列头；并使用同编码图4(a)和(b)相同的方法编码即可。
实施例二：
该实施例中，针对典型的以经纬图方式映射的全景视频编码，定义了6个感兴趣区域中心，其表示也如图4所示，(a)-(f)分别表示了中心经纬度为(0,0)，(90,0)，(-90,0)，(180,0)，(0,90)和(0,-90)的感兴趣区域。每个感兴趣区域中心定义了2个半径分别为60度和90度大小的感兴趣区域。因此，对该全景视频，根据图4(a)-(f)的划分，可以编码生成6个码流。
对应图4(a)的码流，如图6所示，需要将感兴趣区域中心C点经纬度(0,0)、感兴趣区域层数(2层)、第一层感兴趣区域的大小(60度)、第二层感兴趣区域的大小(90度)等信息写入序列头；如图6所示，计算每个点到感兴趣区域中心点C的距离；根据当前点B在图像上的坐标，计算该点在全景球面上对应的经纬度(Bx,By),根据点B经纬度(Bx,By)和感兴趣区域中心C点经纬度(0,0)计算B、C点在全景球面上的距离D’；如果D’小于第一层感兴趣区域的大小(60度)，则不对B进行滤波；否则，如果D’的值介于60度和90度之间，则对点B使用模板大小(Filter Size)为3x3的高斯滤波，否则对该点使用模板大小为5x5的高斯滤波；最后，使用传统方法进行编码。
对应图4(b)的码流，需要将感兴趣区域中心点经纬度(90,0)、感兴趣区域层数(2层)、第一层感兴趣区域的大小(60度)、第二层感兴趣区域的大小(90度)等信息写入序列头；其余操作同编码图4(a)对应的码流相同。
对应图4(c)-(f)的码流，同理，将各自感兴趣区域中心点经纬度、感兴趣区域层数、每层感兴趣区域大小等信息写入序列头；并使用同编码图4(a)和(b)相同的方法操作即可。
综上，本发明基于全景球面上的多层圆周来表示全景视频感兴趣区域为全景视频多层感兴趣区域；对全景视频多层感兴趣区域进行编码，编码方法可对全景视频多层感兴趣区域进行灵活的码率分配；在保证感兴趣区域具有较高的图像质量的同时，本发明技术方案能够大 幅度降低编码和传输所需的码率。
需要注意的是，公布实施例的目的在于帮助进一步理解本发明，但是本领域的技术人员可以理解：在不脱离本发明及所附权利要求的精神和范围内，各种替换和修改都是可能的。因此，本发明不应局限于实施例所公开的内容，本发明要求保护的范围以权利要求书界定的范围为准。

Claims (10)

  1. 一种全景视频感兴趣区域描述方法，所述方法基于全景球面上的多层圆周来表示全景视频感兴趣区域，设定圆周层数为N，将中心相同的多层感兴趣区域表示为球半径为1个单位的全景球面上的中心相同、半径不同的多层圆周包围的球面区域；将所述中心的坐标、圆周层数N和每层圆周在全景球面上的半径均写入码流的序列头，即完成对全景视频多层感兴趣区域的描述。
  2. 如权利要求1所述全景视频感兴趣区域描述方法，其特征是，所述方法具体包括如下步骤：
    A1)将全景球面上的圆周的中心所处全景球面的经纬度(Cx,Cy)设定为全景视频感兴趣区域的中心；
    A2)再设定感兴趣区域的层数为N；设定层数计数器初始值n＝1；
    A3)使用圆周在全景球面上的半径来表示感兴趣区域的大小；当前层为n层，当前层感兴趣区域的大小记作Rn；层数计数器取值加1；
    A4)循环执行步骤A3)，直到获取到所有N层感兴趣区域的大小；
    A5)将所述全景视频感兴趣区域的中心的坐标、层数N和每层圆周在全景球面上的半径都写入码流的序列头，即完成对全景视频多层感兴趣区域的描述。
  3. 如权利要求2所述全景视频感兴趣区域描述方法，其特征是，步骤A3)将感兴趣区域的大小表示为全景球面上相应圆周的半径的两端同圆心连线的夹角的弧度。
  4. 一种全景视频感兴趣区域编码方法，针对采用全景视频感兴趣区域描述方法进行描述得到的全景视频多层感兴趣区域，所述编码方法对位于不同层的感兴趣区域中的图像块使用不同的QP调整值(第n层QP调整值记为ΔQPn，且ΔQPn≤0)，并将所述QP调整值同通过传统方法得到的图像块QP初始值相加，得到图像块编码的实际QP；再根据当前图像块所述实际QP，对当前图像块采用现有编码方法的流程进行编码；所述全景视频感兴趣区域描述方法基于全景球面上的多层圆周来描述全景视频感兴趣区域，设定圆周层数为N，将中心相同的多层感兴趣区域表示为半径为1个单位的全景球面上中心相同、半径不同的多层圆周包围的球面区域；将所述中心的坐标、圆周层数N、每层圆周半径写入码流的序列头，即完成对全景视频多层感兴趣区域的描述。
  5. 如权利要求4所述编码方法，其特征是，所述现有编码方法的流程是：对图像进行分块，再对每一个块进行编码；当编码当前块时，首先对当前块进行预测，得到当前块的预测 块；再将当前块的像素值减去预测块，得到残差块；之后对残差块进行变换得到变换块，并对变换块进行量化操作得到量化块；最后将量化块中每个点的值写入码流；对每一个块完成上述操作即完成图像编码。
  6. 如权利要求4所述编码方法，其特征是，所述QP调整值根据以下方式得到：如果当前图像块中心同感兴趣区域中心的距离D小于R1，则该图像块的QP调整值为ΔQP1；否则，如果D大于RN，则该图像块的QP调整值为0；否则，如果D小于Ri且大于Ri-1(i>1)，则该图像块的QP调整值为ΔQPi。
  7. 如权利要求6所述编码方法，其特征是，所述当前图像块中心点同感兴趣区域中心的距离D，根据以下方式得到：根据当前全景图的映射方式和图像块中心点A在图像上的坐标，计算得到A在所述全景球面上的经纬度(Ax,Ay)，并根据经纬度(Ax,Ay)和感兴趣区域中心经纬度(Cx,Cy)计算得到A、C点的球面距离，即为D。
  8. 如权利要求7所述编码方法，其特征是，所述的A、C点的球面距离D，其值等于所述全景球面上，A、C两点同球心连线的夹角的弧度，因此也可以用所述弧度对应的角度值表示球面距离D。
  9. 一种全景视频感兴趣区域编码方法，针对采用全景视频感兴趣区域描述方法进行描述得到的全景视频多层感兴趣区域，所述编码方法首先计算全景球面上的每个点在全景球面上与感兴趣区域中心点C的距离，设定当前点为B，计算得到当前点B和点C的距离D’，再根据D’确定点B的优先级P，并根据D’和P选择滤波器或滤波强度，对当前点B进行低通滤波，最后对滤波后的整帧图像进行编码；所述全景视频感兴趣区域描述方法基于全景球面上的多层圆周来描述全景视频感兴趣区域，设定圆周层数为N，将中心相同的多层感兴趣区域表示为半径为1个单位的全景球面上中心相同、半径不同的多层圆周包围的球面区域；将所述中心的坐标、圆周层数N、每层圆周半径写入码流的序列头，即完成对全景视频多层感兴趣区域的描述。
  10. 如权利要求9所述编码方法，其特征是，所述的优先级P根据以下方式得到：如果当前图像块中心同感兴趣区域中心的距离D小于R1,则该点的优先级为1；否则，如果D大于RN,则该点的优先级为N+1；否则，如果D小于Ri且大于Ri-1(i>1)，则该点的优先级为i。
PCT/CN2017/092552 2016-10-21 2017-07-12 全景视频感兴趣区域的描述方法和编码方法 Ceased WO2018072487A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/330,923 US10681355B2 (en) 2016-10-21 2017-07-12 Describing method and coding method of panoramic video ROIs

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201610921242.3 2016-10-21
CN201610921242 2016-10-21
CN201610945845.7A CN106412582B (zh) 2016-10-21 2016-10-26 全景视频感兴趣区域的描述方法和编码方法
CN201610945845.7 2016-10-26

Publications (1)

Publication Number Publication Date
WO2018072487A1 true WO2018072487A1 (zh) 2018-04-26

Family

ID=58013819

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/092552 Ceased WO2018072487A1 (zh) 2016-10-21 2017-07-12 全景视频感兴趣区域的描述方法和编码方法

Country Status (3)

Country Link
US (1) US10681355B2 (zh)
CN (1) CN106412582B (zh)
WO (1) WO2018072487A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109587495A (zh) * 2018-11-05 2019-04-05 深圳威尔视觉传媒有限公司 视频编码方法、装置、设备及存储介质
CN110611815A (zh) * 2019-09-26 2019-12-24 深圳市半冬科技有限公司 一种视频输出方法、装置以及存储装置

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106412582B (zh) 2016-10-21 2019-01-29 北京大学深圳研究生院 全景视频感兴趣区域的描述方法和编码方法
FR3064145A1 (fr) * 2017-03-16 2018-09-21 Orange Procede de codage et decodage d'images, dispositif de codage et decodage et programmes d'ordinateur correspondants
CN108810427B (zh) * 2017-05-02 2020-12-04 北京大学 基于视点的全景视频内容表示的方法及装置
CN107146274B (zh) 2017-05-05 2021-06-22 上海兆芯集成电路有限公司 图像数据处理系统、纹理贴图压缩和产生全景视频的方法
CN109862356B (zh) * 2019-01-17 2020-11-10 中国科学院计算技术研究所 一种基于感兴趣区域的视频编码方法及系统
CN111818336B (zh) * 2019-04-12 2022-08-26 华为技术有限公司 视频处理方法、装置、存储介质以及通信装置
US10939126B1 (en) * 2019-12-09 2021-03-02 Guangzhou Zhijing Technology Co., Ltd Method of adding encoded range-of-interest location, type and adjustable quantization parameters per macroblock to video stream
CN113630622B (zh) * 2021-06-18 2024-04-26 中图云创智能科技(北京)有限公司 全景视频图像处理方法、服务端、目标设备、装置和系统
CN114529483B (zh) * 2022-02-10 2024-12-06 Oppo广东移动通信有限公司 数据处理方法、装置、终端和可读存储介质
CN119301944A (zh) * 2022-04-01 2025-01-10 Op方案有限责任公司 用于基于区域封装的压缩的系统和方法
CN116489364A (zh) * 2023-04-28 2023-07-25 哲库科技(上海)有限公司 一种编码方法、编码器、存储介质及芯片

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120092348A1 (en) * 2010-10-14 2012-04-19 Immersive Media Company Semi-automatic navigation with an immersive image
CN104735464A (zh) * 2015-03-31 2015-06-24 华为技术有限公司 一种全景视频交互传输方法、服务器和客户端
US20160165309A1 (en) * 2013-07-29 2016-06-09 Koninklijke Kpn N.V. Providing tile video streams to a client
CN105915907A (zh) * 2016-06-07 2016-08-31 北京圣威特科技有限公司 全景图的压缩和解压方法、装置及系统
CN106412582A (zh) * 2016-10-21 2017-02-15 北京大学深圳研究生院 全景视频感兴趣区域的描述方法和编码方法

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7676022B2 (en) * 2005-05-02 2010-03-09 Oy Ajat Ltd. Extra-oral digital panoramic dental x-ray imaging system
CN101119482B (zh) * 2007-09-28 2011-07-20 北京智安邦科技有限公司 一种全景监控方法及设备
CN100542298C (zh) * 2007-09-29 2009-09-16 中国科学院计算技术研究所 一种柱面全景视频编码的块尺寸计算方法和传输方法
CN102175221B (zh) * 2011-01-20 2014-04-09 北京百度网讯科技有限公司 基于鱼眼镜头的车载移动摄影测量系统
CN102999918B (zh) * 2012-04-19 2015-04-22 浙江工业大学 全景视频序列图像的多目标对象跟踪系统
CN102833525A (zh) * 2012-07-19 2012-12-19 中国人民解放军国防科学技术大学 一种360度全景视频的浏览操作方法
CN115118952B (zh) * 2016-05-13 2024-07-19 索尼公司 信息处理装置和信息处理方法以及再现装置和再现方法
CN109478347A (zh) * 2016-07-29 2019-03-15 索尼公司 图像处理装置和图像处理方法
US11172005B2 (en) * 2016-09-09 2021-11-09 Nokia Technologies Oy Method and apparatus for controlled observation point and orientation selection audiovisual content
US20180176468A1 (en) * 2016-12-19 2018-06-21 Qualcomm Incorporated Preferred rendering of signalled regions-of-interest or viewports in virtual reality video

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120092348A1 (en) * 2010-10-14 2012-04-19 Immersive Media Company Semi-automatic navigation with an immersive image
US20160165309A1 (en) * 2013-07-29 2016-06-09 Koninklijke Kpn N.V. Providing tile video streams to a client
CN104735464A (zh) * 2015-03-31 2015-06-24 华为技术有限公司 一种全景视频交互传输方法、服务器和客户端
CN105915907A (zh) * 2016-06-07 2016-08-31 北京圣威特科技有限公司 全景图的压缩和解压方法、装置及系统
CN106412582A (zh) * 2016-10-21 2017-02-15 北京大学深圳研究生院 全景视频感兴趣区域的描述方法和编码方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
SANCHEZ, Y. ET AL.: "Compressed Domain Video Processing for Tile Based Panora- mic Streaming Using HEVC", 2015 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING(ICIP, vol. 31, 31 December 2015 (2015-12-31), XP058074928, DOI: 10.1145/2814347.2814353 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109587495A (zh) * 2018-11-05 2019-04-05 深圳威尔视觉传媒有限公司 视频编码方法、装置、设备及存储介质
CN110611815A (zh) * 2019-09-26 2019-12-24 深圳市半冬科技有限公司 一种视频输出方法、装置以及存储装置

Also Published As

Publication number Publication date
US10681355B2 (en) 2020-06-09
US20190253719A1 (en) 2019-08-15
CN106412582B (zh) 2019-01-29
CN106412582A (zh) 2017-02-15

Similar Documents

Publication Publication Date Title
WO2018072487A1 (zh) 全景视频感兴趣区域的描述方法和编码方法
CN112889282B (zh) 用于对图像执行人工智能编码和人工智能解码的方法和设备
CN110115037A (zh) 球面投影运动估计/补偿和模式决策
US11240512B2 (en) Intra-prediction for video coding using perspective information
CN107211131B (zh) 对数字图像块进行基于掩码的处理的系统和方法
CN111417985A (zh) 用于编码/解码表示3d对象的点云的几何形状的方法和设备
CN111131831B (zh) 数据传输方法及装置
CN105812759A (zh) 一种360度全景视频的平面投射方法及编码方法
CN107959856A (zh) 通过帧内预测来对图像进行编码和解码的方法和设备
Li et al. End-to-end optimized 360° image compression
CN112715029B (zh) Ai编码设备及其操作方法和ai解码设备及其操作方法
CN116250234A (zh) 用于对基于mpi的体积视频进行编码的方法和装置
CN111433816A (zh) 用于对表示3d对象的点云的几何形状进行编码/解码的方法和装置
CN115086658B (zh) 点云数据的处理方法、装置、存储介质及编解码设备
CN117136543A (zh) 对自旋传感器头捕获的点云几何数据进行编码/解码的方法和装置
JP6983463B2 (ja) ３６０画像のｑｐコーディングおよびビデオコーディングのための技術
CN117157973A (zh) 编解码方法及相关设备、存储介质
CN103716622B (zh) 用于图像处理的方法和装置
WO2024060161A1 (zh) 编解码方法、编码器、解码器以及存储介质
WO2021253721A1 (zh) 一种帧内预测方法、装置、终端及存储介质
RU2857375C2 (ru) Способ и устройство для кодирования/декодирования позиций точек облака точек, содержащихся в кубоидных объемах
CN113179408B (zh) 一种图像编解码方法及装置
CN102685485A (zh) 编码方法以及装置、解码方法以及装置
CN109510986B (zh) 带宽压缩中自适应纹理渐变预测方法
CN120283266A (zh) 用于位置编码中初始顶点选择的最优子网格编码顺序

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17861322

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17861322

Country of ref document: EP

Kind code of ref document: A1