WO2009033390A1 - Method and device for acquiring block boundary strength and reducing block distortion - Google Patents

Method and device for acquiring block boundary strength and reducing block distortion Download PDF

Info

Publication number
WO2009033390A1
WO2009033390A1 PCT/CN2008/072082 CN2008072082W WO2009033390A1 WO 2009033390 A1 WO2009033390 A1 WO 2009033390A1 CN 2008072082 W CN2008072082 W CN 2008072082W WO 2009033390 A1 WO2009033390 A1 WO 2009033390A1
Authority
WO
WIPO (PCT)
Prior art keywords
block boundary
block
sides
filter
blocks
Prior art date
Application number
PCT/CN2008/072082
Other languages
English (en)
French (fr)
Inventor
Jiali Fu
Jiantong Zhou
Changqi Hu
Sixin Lin
Shan Gao
Yingjia Liu
Qingpeng Xie
Xiaosen Zhang
Lianhuan Xiong
Original Assignee
Huawei Technologies Co., Ltd.
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 Huawei Technologies Co., Ltd. filed Critical Huawei Technologies Co., Ltd.
Publication of WO2009033390A1 publication Critical patent/WO2009033390A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/85Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression
    • H04N19/86Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression involving reduction of coding artifacts, e.g. of blockiness
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102Methods 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
    • H04N19/117Filters, e.g. for pre-processing or post-processing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/134Methods 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
    • H04N19/136Incoming video signal characteristics or properties
    • H04N19/14Coding unit complexity, e.g. amount of activity or edge presence estimation
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169Methods 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
    • H04N19/17Methods 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
    • H04N19/176Methods 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
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/46Embedding additional information in the video signal during the compression process
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/60Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
    • H04N19/61Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/80Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation

Definitions

  • This invention relates to the field of image processing, and more particularly to a method and apparatus for obtaining block boundary strength and removing blockiness.
  • the image needs to be processed in blocks.
  • the basic processing unit of the H.264 standard is 16x16 macroblocks.
  • block effects occur to varying degrees, that is, discontinuous visual mutations occur at block boundaries between blocks.
  • the block effect is caused by the quantization error caused by the block processing during quantization.
  • the image content in the block range there are mainly two different expressions, one is trapezoidal noise and the other is lattice noise. The above two kinds of noise will make the edges of the image appear obvious or even blocky areas. Specifically, trapezoidal noise appears at the strong edges of the image.
  • Step 1 Calculate the flatness of the image region where the current block boundary is located by using the actual pixel interpolation on both sides of the current block boundary.
  • the flatness of the region is represented by the number of similar point pairs around the block boundary. If there are many similar points, It is considered that the flatness of the image area where the current block boundary is located is large.
  • the definition of similar points can be referred to FIG. 1.
  • FIG. 1 is a schematic diagram showing an example of similarity discrimination samples in the prior art.
  • a thick solid line refers to the main block boundary of two blocks, and a broken line is a secondary block boundary within two blocks.
  • Abs (y) is the absolute value of y and can be expressed as formula (1):
  • the flatness values of 73 ⁇ 4 and 73 ⁇ 4 are obtained by looking up the table.
  • the average value QP av can be calculated according to the quantization parameters (QP, Quantization Parameter) of the two blocks, where QP is used in the quantization process.
  • the quantization step size is then determined by QPav to look up the table index / «ife, and finally look up the table to find the flatness thresholds Tin and Th 2 .
  • the values of Thi and 73 ⁇ 4 2 are obtained from Table 1.
  • Table 1 shows the relationship between flatness thresholds Tin and 73 ⁇ 42 and InckxA:
  • QP av The QP of each block is the data known when the block is quantized, and QP av can be expressed as the formula (2) with the following pseudo code:
  • Step 2 According to the specific condition of the flatness, the filtering strength of the current block boundary is determined.
  • the specific method is as follows: According to the macroblock type and the regional flatness, the value of the block boundary filtering strength (BS, Boundary Strength) is determined, and the steps are as follows:
  • BS is obtained, which is specifically as follows:
  • BS is equal to 1; otherwise, BS is equal to 2.
  • the BS is adjusted according to the macroblock type, specifically:
  • the macroblock type here can be obtained in the header information of each macroblock, and can be classified into various types such as an intra prediction macroblock and an inter prediction macroblock.
  • Step 3 After determining the BS value, filter the pixels on both sides of the block boundary according to the flatness for different non-zero BS values.
  • Three levels of BS values can be obtained from step two, and this step is to determine three different filtering processes for different levels of BS values, so as to filter each row or column in the block accordingly.
  • the filtering process of q 2 is explained, and by P. And Q. Represents o and q, respectively. Filtered value. If in the filtering process? If ⁇ is not assigned, then p ⁇ qO is not filtered; otherwise, it is either or filtered.
  • P 2 or Q 2 is a filtered value of p 2 or q 2 .
  • the three different filtering processes determined by the BS value are:
  • P2 ap ? (pO + pi + 4 x p2 + 2 x p3 + 4)) » 3 : p2 ;
  • two sets of filters are used respectively.
  • a strong filter that is, (2 x pl + p2 + 2 X (pO + qO) + ql + 4)»3
  • a weak set of filters ie (2 XO + pi + q0 + 2 ) 2 .
  • ap can be obtained by the following formula (3)
  • aq can be obtained by the following formula (4):
  • IndexB Clip3(0, 63, QPav + BetaOffset)
  • the values of ⁇ and ⁇ are obtained from Table 2, and a is obtained according to the IndexA lookup table, and is obtained according to the IndexB lookup table.
  • Clip3(a, b, c) is expressed as:
  • P0 and Q0 can be divided into three cases:
  • the first case When the block boundary is a chroma block boundary, pi and ql are not filtered.
  • the current method of obtaining the block boundary filtering strength the obtained BS is inaccurate, and thus the current method of removing the block effect cannot effectively reduce the block effect.
  • Embodiments of the present invention provide a method for obtaining block boundary filtering strength, which can improve the accuracy of the obtained block boundary filtering strength.
  • Embodiments of the present invention provide a method for obtaining block boundary filtering strength, which can improve the accuracy of the obtained block boundary filtering strength.
  • Embodiments of the present invention provide a method for removing a block effect, which can improve the accuracy of the obtained block boundary filtering strength, thereby effectively reducing the block effect.
  • Embodiments of the present invention also provide an apparatus for obtaining block boundary filtering strength, which can improve the accuracy of the obtained block boundary filtering strength.
  • Embodiments of the present invention also provide an apparatus for removing blockiness, which can improve the accuracy of the obtained block boundary filtering strength, thereby effectively reducing the blockiness.
  • a method for obtaining a block boundary filtering strength comprising: determining a similar threshold, and calculating a block boundary in a current pane by using the similar threshold
  • the area flatness is determined according to the area flatness interval in which the area flatness is located, and the initial value of the filter intensity of each block boundary in each pane is determined, and each of the area flatness intervals corresponds to an initial value of the filter strength;
  • the header information of the blocks on both sides of the block boundary is adjusted, and the block boundary filter strength initial value is adjusted to obtain the block boundary filter strength.
  • a method for obtaining a block boundary filter strength comprising: determining, according to header information of blocks on both sides of a block boundary, whether to filter pixels of blocks on both sides of the block boundary, and if not, determining a similar threshold Calculating the area flatness by using the similarity threshold, the header information of the blocks on both sides of the block boundary includes: type information or coding block mode information; determining each window according to the area flatness interval in which the area flatness is located An initial value of the filter strength of the intra-block boundary, each of the regional flatness intervals corresponds to an initial value of the filter strength; and the initial value of the filter strength is adjusted by using the header information of the blocks on both sides of the block boundary obtained in advance, The block boundary filtering strength is obtained.
  • a method for removing a block effect comprising: determining a similarity threshold, calculating a region flatness using the similarity threshold; obtaining a block boundary filter strength from the region flatness, using each of the blocks on both sides of the block boundary
  • the first-order difference and the second-order difference between adjacent pixels determine whether the block boundary is a pseudo-boundary.
  • the corresponding filter combination is determined, and the pixels of the blocks on both sides of the block boundary are subjected to corresponding filtering processing.
  • a device for obtaining block boundary filtering strength comprising:
  • a flatness calculation module configured to determine a similarity threshold, and calculate a flatness of a block boundary area in the current pane by using the similarity threshold;
  • the initial value determining module is configured to determine an initial value of the filtering intensity of the block boundary in each pane according to the regional flatness interval in which the regional flatness calculated by the flatness calculating module is located, and each of the regional flatness intervals corresponds to An initial value of the filter strength;
  • an adjustment module configured to use the header information of the blocks on both sides of the block boundary that is obtained in advance, and adjust the initial value of the block boundary filter strength determined by the initial value determining module to obtain the block boundary filtering strength.
  • a device for removing blockiness comprising:
  • a block boundary filter strength obtaining module configured to determine a similarity threshold value, calculate a region flatness by using the similarity threshold value, and obtain a block boundary filter strength from the region flatness;
  • a boundary determining module configured to determine whether the block boundary is a pseudo boundary
  • a filter determining module configured to determine, at the boundary determination module, the block boundary When it is a pseudo boundary, according to the correspondence between the pre-established block boundary filter strength and the filter combination, the block boundary filter strength obtained by the block boundary filter strength acquisition module determines the corresponding filter combination;
  • a filtering module configured to filter the pixels determined by the filter determining module, and perform corresponding filtering processing on pixels of the blocks on both sides of the block boundary.
  • the method and apparatus for obtaining block boundary filtering strength and removing block effect determine a similar threshold according to a block boundary threshold obtained by pre-checking a table, or by a quantization parameter, a macroblock mode,
  • the color space or the frame type determines the similarity threshold, and uses the header information of the block to adjust the initial value of the block boundary filter strength, which can be independent of the coding block unit, and fully utilizes the relationship between the block and the block and the header information of the block. Therefore, the accuracy of the obtained block boundary filter strength is improved, thereby effectively reducing the block effect.
  • FIG. 1 is a schematic diagram showing an example of a similarity discrimination sample in the prior art
  • Figure 2 is a schematic diagram of the pixel distribution of one row/column on both sides of the block boundary
  • FIG. 3 is a schematic flowchart of a method for obtaining block boundary filtering strength according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram showing an example of a tile size N of 4 samples according to an embodiment of the present invention
  • FIG. 5 is a pane size N in an embodiment of the present invention.
  • FIG. 6 is a schematic flow chart of a method for removing a block effect by using the obtained block boundary filtering strength according to another embodiment of the present invention.
  • FIG. 7 is a schematic diagram of a pixel distribution of a block boundary on both sides of a block boundary according to another embodiment of the present invention
  • FIG. 8 is a schematic structural diagram of an apparatus for obtaining a block boundary filter strength according to still another embodiment of the present invention
  • FIG. 9 is a view of another embodiment of the present invention. Schematic diagram of the block structure device structure.
  • An embodiment of the present invention provides a method for obtaining a block boundary filtering strength by using pixel and header information, specifically: determining a similar threshold, and calculating a region flatness by using the similar threshold; An area flatness interval in which the area flatness is located, determining an initial value of the filter strength of the block boundary, each of the area flatness intervals corresponding to an initial value of the filter strength; using the pre-obtained header information of the blocks on both sides of the block boundary And adjusting the initial value of the filtering strength to obtain the block boundary filtering strength.
  • the header information of the blocks on both sides of the block boundary determining whether the pixels of the blocks on both sides of the block boundary need to be filtered, and the header information of the block includes type information or coding block mode information, and if yes, ending the process; Otherwise, a similar threshold is determined, and the regional flatness is calculated by using the similarity threshold; and an initial value of the filtering strength of the block boundary in each pane is determined according to the regional flatness interval in which the region is flat, each of the regions The flatness interval corresponds to an initial value of the filter strength; the header information of the blocks on both sides of the block boundary acquired in advance is utilized.
  • An embodiment of the present invention further provides a method for removing a block effect, specifically: determining a similarity threshold, calculating a region flatness by using the similarity threshold; obtaining a block boundary filter strength from the region flatness, using the block
  • the first-order difference and the second-order difference between adjacent pixels in the block on both sides of the boundary determine whether the block boundary is a pseudo-boundary.
  • determining a similarity threshold calculating a region flatness by using the similarity threshold
  • the first-order difference and the second-order difference between adjacent pixels in the block on both sides of the boundary determine whether the block boundary is a pseudo-boundary.
  • FIG. 3 is a schematic flow chart of a method for obtaining block boundary filtering strength according to an embodiment of the present invention. As shown in Figure 3, the following steps are specifically included:
  • Steps 301 to 302 are processes for calculating the flatness of the block boundary area in the current pane.
  • Step 301 Divide the current filter block boundary into several panes.
  • the method of dividing the pane is: pre-setting the pane size, and dividing the blocks on both sides of the block boundary into more than one pane according to the pane size according to the direction of the vertical block boundary.
  • Step 302 Calculate the area flatness of the block boundary in the current pane.
  • the process of calculating the flatness of the area includes: calculating the block separately in the current pane Pixel flatness of the blocks on both sides of the boundary; Accumulate the pixel flatness of the blocks on both sides to obtain the flatness of the area within the pane. That is to say, the area flatness of the block boundary in the current pane is obtained by summing the pixel flatness on both sides of the block boundary in the current pane, and the flatness of the block boundary area in the current pane is similar to the next block boundary in the current pane.
  • the Offset_ ⁇ value is a pre-set offset that can adjust the filter strength.
  • the Offset_ ⁇ value can be obtained from the transmitted video stream.
  • Steps 303 to 304 are processes for obtaining a BS according to the flatness and header information.
  • the BS can be divided into multiple levels.
  • the BS is divided into four levels, that is, the BS values can be 0, 1, 2, and 3.
  • Step 303 Calculate the initial value of the current block boundary filter strength.
  • the method for calculating the initial value of the current block boundary filter strength bs is specifically as follows:
  • the CO and CI in the formula are preset flatness width values according to statistical laws.
  • the CO and CI in the formula are preset flatness width values according to statistical laws.
  • C0 is set to 2 and C1 is set to 7.
  • Step 304 Correct the block boundary strength according to the header information to obtain the block boundary filtering strength.
  • the current block is not filtered; if it is indicated that one block is an intra block, the current block boundary filtering strength The initial value bs is increased by 1; if the information in the header information indicates the motion vector of the block on both sides of the block boundary The difference is greater than the preset threshold Tmv.
  • Tmv takes 3 integer pixels, the current bs is increased by 1, or if the information in the header information indicates that the two regions of the block boundary are from different reference frames, respectively, bs is increased by 1. If the coded block mode of the block on both sides of the block boundary is 0 at the same time, BS is 0.
  • the method for obtaining the BS obtains the current block boundary filter strength initial value bs according to the similarity point number num of the blocks on both sides of the block boundary in the pane, and thus only needs to determine the similarity threshold T of the flatness in one pane. It is not necessary to perform the method of looking up Table 1 in the prior art to obtain num, so the method for obtaining the BS in this embodiment is more convenient and flexible.
  • FIG. 6 is a schematic flow chart of a method for removing a block effect by using the obtained block boundary filter strength according to another embodiment of the present invention. As shown in FIG. 6, the following steps are specifically included:
  • x, y, z will be based on quantization parameters, macroblock mode, ie intra macroblock (Intra) and inter macroblock Inter), color space, ie luma and chroma (chroma) and frame type, ie intraframe
  • the difference between the prediction frame (I), the bidirectional prediction frame (B), and the unidirectional prediction frame (P) is adaptively adjusted.
  • T (xa + yp) / z
  • Step 605 Determine whether the block boundary is a real block boundary. If yes, no filtering is required, and the process ends; otherwise, step 606 is performed.
  • the method for determining whether the block boundary is a true block boundary is as follows: Whether the difference between each adjacent pixel is greater than a certain threshold width, and if so, the true edge of the image, no filtering is required; otherwise, it is a pseudo edge, which needs to be filtered.
  • the block boundary is a vertical block boundary
  • the pixel distribution of one line of the two sides of the block boundary is as shown in FIG. 7.
  • it may be a block boundary or a horizontal block boundary, and both sides of the block boundary are needed.
  • a column of image pixels determines whether the block boundary is a true block boundary. For a pair of pixels R0 and L0 on both sides of the block boundary of a given block in FIG. 7, the two pairs of pixels L0 and LI, R0 and R1 in the small block respectively obtain a difference.
  • Ml is less than o and greater than 1
  • M2 is less than ⁇
  • M3 is less than ⁇ .
  • any one of ⁇ 4 and ⁇ 5 is greater than 0, it is considered that the characteristic edge of the image is not at the edge of the block, that is, BS is not equal to 0, and filtering can be performed; otherwise, no filtering is performed.
  • Step 606 Determine the corresponding filtering process according to different levels of the non-zero BS for the pseudo block boundary.
  • R0 aq? (2*Rl+R0+L0+2)»2: ( Rl+2*R0+L0+2) »2;
  • L0 ap? (2*Ll+L0+R0+2)»2: ( Ll+2*L0+R0+2) »2;
  • R0 aq? ( R2+2 *R1 +2 *R0+2 *L0+L 1 +4 ) »3: (R2+2*Rl+3*R0+2*L0+4)
  • L0 ap? ( L2+2 *L 1 +2 *L0+2 *R0+R1 +4 ) »3: (L2+2*Ll+3*L0+2*R0+4) » 3;
  • Rl aq? ( 2*R2 + 4*R1 + RO + LO + 4 ) » 3 : Rl;
  • R0 ( 3*R2+3*Rl+4*R0+3*L0+3*Ll+8) » 4;
  • Rl aq ? (3*R2+4*Rl+4*R0+3*L0+2*Ll+8) » 4 : (2*R2+4*Rl+R0+L0+4) » 3;
  • LI ap ? (3*L2+4*Ll+4*L0+3*R0+2*Rl+8) » 4 : ( 2*L2+4*Ll+L0+R0+4) » 3;
  • R2 aq? (2*R3+3*R2+R1 + RO +L0+ 4 )»3: R2;
  • L2 ap ? (2*L3+3*L2+L1 + LO +R0+ 4 )»3: L2;
  • the same filtering method is used only for R Q and L Q pixels on both sides of the block boundary, and R2, Rl, L1, and L2 are not filtered.
  • the filtering process of the vertical block boundary is taken as an example.
  • the filtering window is in a vertical state, but the filtering process on the corresponding horizontal block boundary is also the same, so the details are not detailed. Said.
  • the value of the BS may be adjusted according to the skip mode or the cbp. For example, when two adjacent blocks are skip mode or cbp is all 0, the BS is decremented by one.
  • the header information of the block may be used to determine whether filtering is needed, for example, according to skip mode or cbp, whether the filtering needs to be performed, when the blocks on both sides of the adjacent block boundary are all skip mode or cbp. When it is 0, it is judged that filtering is not required, and the flow is ended.
  • the weight of each row of pixels may be different according to the filtering. For example, according to the macroblock type, the weight of each row of pixels of the block boundary is determined, and according to the weight of each row of pixels, filtering is performed by using different intensity filters. A relatively strong filtering can be performed on the block boundary of the block, and a relatively weak filtering weighting strategy is applied to the middle row of the block for filtering.
  • the pixel to be filtered is to be a block boundary line of the current block, such as 0, 7, 8, or 15 lines of the macro block, the most powerful one of the filters to which the current BS value belongs is used for filtering.
  • the filter in this embodiment may be shared, and the specific process may use a weak filter with a larger BS as a strong filter with a smaller BS, or a strong filter with a smaller BS as a weak filter for a larger BS, so that In the same case of BS, there are more filters of different strengths to choose, which increases the accuracy of filtering without increasing the complexity of the operation. For example, a weak filter when BS is 3 is a strong filter when BS is 2, and a weak filter when BS is 2 can be used as a strong filter when BS is 1.
  • the block boundaries in the current pane can be divided into different levels of BS values, and the filters of each level can also be selected differently, but the overall principle is to use the mean filtering method, and for the BS. The higher the level, the stronger the filter is used.
  • the row/column pixels belonging to the pseudo boundary in the pane are filtered, and the block effect caused by transform quantization and motion compensation can be effectively removed.
  • FIG. 8 is a schematic structural diagram of an apparatus for obtaining block boundary filtering strength according to still another embodiment of the present invention.
  • the apparatus includes: a flatness calculation module 810, an initial value determination module 820, and an adjustment module 830.
  • the flatness calculation module 810 determines a similarity threshold, and calculates the regional flatness using the similarity threshold.
  • the initial value determining module 820 determines an initial value of the filtering intensity of the block boundary in each pane according to the area flatness interval in which the area flatness is located, and each of the area flatness intervals corresponds to an initial value of the filtering intensity.
  • the adjusting module 830 adjusts the initial value of the filtering strength by using the header information of the blocks on both sides of the block boundary obtained in advance, and obtains the block boundary filtering strength.
  • FIG. 9 is a schematic structural diagram of an apparatus for removing a block effect according to still another embodiment of the present invention.
  • the apparatus includes: a block boundary filter strength acquiring module 910, a boundary determining module 920, and a filter determining module 930. And filtering module 940.
  • the block boundary filter strength obtaining module 910 is configured to determine a similarity threshold, calculate a region flatness by using the similar threshold, and obtain a block boundary filter strength from the region flatness; and a boundary determining module 920, determine the block boundary It is a pseudo boundary, and the determination result is input to the filter determination module 930.
  • the filter determining module 930 when the determination result of the boundary determining module indicates that the block boundary is a pseudo boundary, the block boundary filtering obtained by the block boundary filtering strength acquiring module 910 according to the correspondence between the pre-established block boundary filtering strength and the filter combination. Intensity, determine the corresponding filter combination.
  • the filtering module 940 determines the filter combination determined by the filter determining module 930, and performs corresponding filtering processing on the pixels of the blocks on both sides of the block boundary in the pane.
  • each functional unit in each embodiment of the present invention may be integrated into one processing module, or each unit may exist physically separately, or two or more units may be integrated into one module.
  • the above integrated modules can be implemented in the form of hardware or in the form of software functional modules.
  • the integrated modules, if implemented in the form of software functional modules and sold or used as separate products, may also be stored in a computer readable storage medium.
  • the above-mentioned storage medium may be a read only memory, a magnetic disk or an optical disk or the like.

Landscapes

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

Description

获得块边界强度和去除块效应的方法和装置
本申请要求于 2007 年 9 月 13 日提交中国专利局、 申请号为 200710149478.0、 发明名称为"获得块边界强度和去除块效应的方法和装置" 的中国专利申请的优先权, 其全部内容通过引用结合在本申请中。
技术领域
本发明涉及图像处理领域, 特别涉及获得块边界强度和去除块效应的方 法和装置。
背景技术
在 H.264和数字音视频编解码技术标准等视频编码标准中, 对图像均需 要作分块处理, 比如, H.264标准的基本处理单位为 16x 16宏块。 而每个块 在经过变换及量化之后, 都会不同程度的出现块效应, 即在块与块之间的块 边界上出现不连续的视觉突变。 块效应是由于量化时分块处理造成的量化误 差而引起的,按照分块范围内图像内容的不同,主要有两种不同的表现形式, 一种为梯形噪声, 另一种为格形噪声。 以上两种噪声会使图像边缘明显甚至 出现块状区域。 具体来讲, 梯形噪声出现在图像的强边缘处。 比如, 在对图 像进行离散余弦变换得到各频点的变换系数时, 许多高频系数被量化为零, 使得图像的强边缘处在变换域内不能被完全数字化,又由于图像的分块处理, 使得穿过块边界的强边缘处的连续性不能得到保证, 从而在图像的强边缘处 出现锯齿状噪声, 使得视觉上有不自然的数据块边缘, 称之为梯形噪声。 而 格形噪声则出现在图像的平坦区域。 在图像的平坦区域, 亮度有递增或者递 减的情况下,由于量化取整时进行了四舍五入,可能导致变换域的直流(DC ) 系数越过了相邻量化级的判决门限, 造成在重建图像中相邻两个块出现亮度 突变, 使得视觉上出现片状轮廓, 称之为格形噪声。
为了避免或降低块效应, 在对图像进行量化编码完成后, 在存储起来作 为后一帧的参考帧之前, 需要对图像做去除块效应的处理。 下面具体介绍一 种去除块效应的方法, 以一条 8x8块边界为例, 具体说明去除块效应方法具 体步骤: 步骤一: 利用当前块边界两边的实际像素插值情况, 来计算当前块边界 所在图像区域的平坦度, 区域的平坦度用块边界周围的相似点对的个数来表 示, 如果相似点对多, 则认为当前块边界所在图像区域的平坦度较大。 这里 相似点的定义可参照图 1, 图 1为现有技术中相似度判别样本示例示意图。 粗实线指的是两个块的主块边界, 虚线则是分别处于两个块内的次块边界。 对于处于主块边界两侧的像素集 和 Lx (0< = x< = 7 ), 如果其差的绝对值 小于门限 1 ( 7¾ ), 即认为 Rx和 Lx是一对相似点。 对于处于次块边界两侧的 像素集 Lx和 Mx ( 0< = x< = 7 ) 以及 Nx和 Rx ( 0< = x< = 7 ), 如果其差的绝对 值小于门限 2 (Th2), 即认为LX和MX或NX和RX是一对相似点。
下面以一条 8x8块边界为例, 具体说明计算图像区域的平坦度的过程: 首先, 设区域平坦度( RegionFlatness ) 的初始值为 0;
然后, 当主块边界两侧的 8对像素差值的绝对值小于平坦度阔值 Tt , 即 Abs ( Rx-Lx ) < 7¾时, 区域平坦度加 1, 或者在次块边界两侧的 16对 像素的差值的绝对值小于平坦度阔值 Th2, 即 Abs ( Lx-Mx ) < 7¾或 Abs (Nx-Rx) < 7¾时, 区域平坦度加 1, 累加得到区域平坦度。 这里, Abs (y) 是 y的绝对值, 可表示为公式(1 ):
Figure imgf000004_0001
而 7¾和 7¾这两个平坦度阔值, 是通过查表得到的, 可以先根据两个 块的量化参数 ( QP, Quantization Parameter )计算出平均值 QPav, 这里的 QP 为量化过程中釆用的量化步长, 然后由 QPav确定出查表索引 /«ife , 最后 查表得出平坦度阔值 Tin和 Th2。根据索引 IndexA与阔值 Tin和 Th 间的对应 关系, 由表 1得到 Thi、 7¾2的取值。 表 1为平坦度阔值 Tin和 7¾2与 InckxA 的关系:
Figure imgf000004_0002
0 0 0
1 0 0
2 0 1
3 1 1
4 1 2
5 2 2
6 2 2
7 2 2
8 2 2
9 2 2
10 2 2
11 2 3
12 2 3 表 1
每个块的 QP是在对块做量化时已知的数据, QPav可以用以下的伪代码 表示为公式(2):
QPav={QPP + QPq+ )» (2)
其中》 1表示右移一位, 在二进制计算机系统中, 其意义为除以 2, 同 理, 如果是》2表示右移两位, 其意义为除以 4, 如果是》3表示右移三位, 其意义为除以 8, 以此类推。 而索引 IndexA为 QPav的五分之一, 即 IndexA= QPav / 5。 由表 1即可确 定出平坦度阔值 77^和 7¾2
步骤二: 根据平坦度的具体情况, 判断得到当前块边界的滤波强度。 具 体方法为: 根据宏块类型和区域平坦度, 判断得到块边界滤波强度 (BS , Boundary Strength)的值, 其步骤为:
首先, 根据区域平坦度的判断得到 BS , 具体为:
当 RegionFlatness小于 4时, BS等于 0;
当 RegionFlatness大于 4时, 判断水平像素尺度( horizontal— size )是否 大于 500:
如果 horizontal— size大于 500且当 RegionFlatness小于 12时, BS等于 1 , 否则, BS等于 2;
如果 horizontal— size小于 500且当 RegionFlatness小于 16时, BS等于 1 ; 否则, BS等于 2。
接着, 根据宏块类型对 BS进行调整, 具体为:
判断块边界两边的两个 8x8块有一个或两个是否属于帧内预测宏块, 如 果是, 则 BS加 1 , 否则不改变 BS。 这里的宏块类型可以在每个宏块的头信 息中获得, 可分为帧内预测宏块、 帧间预测宏块等多种类型。
步骤三: 通过判断得到 BS值后, 对不同的非零 BS值, 根据平坦度, 对 块边界两边像素进行滤波。
由步骤二可得到三种级别的 BS 值, 而本步骤则是针对不同级别的 BS 值, 确定出三个不同的滤波过程, 以此对块中的每一行或每一列进行相应的 滤波。 下面, 以图 2所示的块边界两边一行 /列图像素分布示意图, 对图 2中 的 Po、 i , p2、 q。、 和 q2的滤波过程进行说明, 并且由 P。和 Q。分别表示 o和 q。滤波后的值。 如果在滤波过程中?^^不被赋值, 则不对 p^qO滤波; 否则, 或 为 或 滤波后的值。 P2或 Q2为 p2或 q2滤波后的值。 由 BS 值确定出的三个不同的滤波过程为:
过程一: 当 BS=3时, 对亮度块边界两边的样本均进行滤波, 对图 2中 的 Po、 i , p2、 qo、 91和¾的滤波过程, 釆用伪代码可表示为:
P0 = ap? (2 X pi + p2 + 2 X (p0 + q0) + ql + 4)»3: (2 x pO + pi + qO + 2 )
»2;
PI = ap ? (pO + 2 x pi + qO + 2) »2: pi;
P2 = ap ? (pO + pi + 4 x p2 + 2 x p3 + 4)) » 3 : p2 ;
QO = aq ? (2 x ql + q2 + 2 x (qO + pO) + pi + 4) » 3 : (2 x qO + ql + pO + 2 ) » 2;
Qi = aq ? (q0 + 2 x qi + po + 2) » 2 : qi;
Q2 = (qo + qi + 4 x q2 + 2 x q3 + 4)) » 3 : q2 ;
在以上对 p0、 pl、 p2、 q0、 ql和 q2的滤波过程中, 分别釆用两组滤波 器, 比如, 在对 pO滤波时, 当 ap大于 0时, 釆用一组强滤波器, 即 (2 x pl + p2 + 2 X (pO + qO) + ql + 4)»3 , 而 ap小于等于 0时, 釆用一组弱滤波器, 即 (2 X O + pi + q0 + 2 )》2。
并且, 在以上对 p0、 pl、 p2、 q0、 ql和 q2的滤波过程中, ap可由以下 的公式(3 )得到, aq可由以下公式(4 )得到:
ap =Abs(p2-pO) < β && Abs(p0 - q0)< ((α» 2) + 2) ( 3 )
aq=Abs(q2-q0) < β && Abs(p。 - q。)< (( » 2) + 2) ( 4 )
公式(3 )和公式(4 ) 中符号 "&&,,表示求 "与"运算, 而公式(3 )和公 式(4 ) 中的 α和 β是为块边界阔值, 可以根据两个块的 QP平均值 QPav, 以 及 AlphaCOffset和 BetaOffset计算查表索弓 I IndexA和 IndexB。 两个块的 QP 平均值 QPav为: QPav=( QPp + QPq + 1 ) » 1 , 而索引 IndexA和 IndexB为: IndexA = Clip3(0, 63, QPav + AlphaCOffset)
IndexB = Clip3(0, 63, QPav + BetaOffset)
根据索引 IndexA和 IndexB 与块边界阔值 α和块边界阔值 β间的对应关 系, 由表 2得到 α、 β的取值, 根据 IndexA查表得到 a, 根据 IndexB查表得 到 。
索 a β 索 a β 索引 a β 索 a β 引 引 引
0 0 0 16 4 2 32 22 6 48 46 15
1 0 0 17 4 2 33 24 7 49 48 16
2 0 0 18 5 3 34 26 7 50 50 17
3 0 0 19 5 3 35 28 7 51 52 18
4 0 0 20 6 3 36 30 8 52 53 19
5 0 0 21 7 3 37 33 8 53 54 20
6 1 1 22 8 4 38 33 8 54 55 21
7 1 1 23 9 4 39 35 9 55 56 22
8 1 1 24 10 4 40 35 9 56 57 23
9 1 1 25 11 4 41 36 10 57 58 23
10 1 1 26 12 5 42 37 10 58 59 24
11 2 1 27 13 5 43 37 11 59 60 24
12 2 1 28 15 5 44 39 11 60 61 25
13 2 2 29 16 5 45 39 12 61 62 25
14 3 2 30 18 6 46 42 13 62 63 26
15 3 2 31 20 6 47 44 14 63 64 27 表 2 在过程一中, 对色度块边界两边的样本 po和 qo釆用了同样的方法滤波。 伪代码中的" d=a? b:c"的含义为: 当 a大于 0时, d=b; 否则 d = c。 以上伪代 码所对应的具体实现程序如下所示:
if ( ap > 0 )
P0 = (2 X pi + p2 + 2 X (ρθ + qO) + ql + 4) » 3
Pi = (po + 2 x i + q0 + 2) » 2
P2 = (pO + pi + 4 x p2 + 2 x p3 + 4)) » 3 else
PO = (2 x O + pi + qO + 2 ) » 2 P2 = p2 if(aq>0)
QO = (2 x ql + q2 + 2 x (qO + pO) + pi + 4) » 3
Ql = (q0 + 2xqi+p0 + 2) »2
Q2 = (qo + qi + 4 x q2 + 2 x q3 + 4)) » 3 else
QO = (2 x qO + ql + pO + 2 ) » 2
Ql=ql
Q2 = q2 过程二: 当 BS=2, 对亮度块边界两边的样本均进行滤波, 只对图 2 中 的 p0、 pl、 qO和 ql进行滤波, 釆用伪代码表示为:
PO = ap? (2 X pi + p2 + 2 X (pO + qO) + ql + 4)»3: (2 x pO + pi + qO + 2 )
»2;
PI = ap ? (pO + 2 x pi + qO + 2) »2: pi;
QO = aq ? (2 x ql + q2 + 2 x (qO + pO) + pi + 4) » 3 : (2 x qO + ql + pO + 2 ) » 2;
Q2 = aq ? (qO + 2 x ql + pO + 2) » 2 : ql; 在对 p0、 pi、 q0和 ql进行滤波中, 由公式( 3 )得到 ap, 由公式( 4 ) 得到 aq。 并且公式(3 )和公式(4 ) 中的 α和 β同过程一中的相同。
过程三: 当 BS=1 , 对亮度块边界两边的样本均进行滤波, 具体为: 首先,对 ρθ和 q0进行滤波(P0和 Q0分别为 ρθ和 q0滤波后的值)可表 示为:
delta = Clip3(-C, C, (((q0 -ρθ) χ 3 + (pi - ql) + 4) » 3))
P0 = Cli l(p0 + delta)
Q0 = Clipl(q0 - delta),
其中, Clip3(a, b, c)表示为 :
Clip3(a, b, c)=
Figure imgf000010_0001
Clipl(a)表示为 : Clipl(x) = Clip3(0, 255, x)
然后, 判断是否需要对 pi和 ql滤波, 在需要对 pi和 ql滤波时, 计算 P0和 Q0 可分为三种情况:
第一种情况: 当块边界为色度块边界时, 不对 pi和 ql滤波。
第二种情况: 当在亮度块边界处有 ap大于 0, 则对 Pl滤波, 滤波后的值 为:
Pi = Clipl(Pl + Clip3(-C, C, (((Ρ。 - Ρι) χ 3 + (p2 - Q。) + 4) »3))) 第三种情况: 当在亮度块边界处有 aq大于 0, 则对 qi滤波, 滤波后的值 为:
Qi = Clipl(qi - Clip3(-C, C, (((¾1 - Q。) x 3 + (P。 - q2) + 4) »3)))
在过程三的滤波过程中, ap和 aq的定义同过程一和过程二, 由公式(3 ) 得到 ap, 由公式(4 )得到 aq。 并且公式(3 )和公式(4 ) 中的 α和 β也与过 程一和二中的相同。而 C称为滤波裁减参数, C与 IndexA之间的关系见表 3。 表 3为滤波裁减参数 C与 IndexA的关系: /:/:/ O soi/-o800si>l£ 06εεεο6§ίAV
Figure imgf000011_0001
15 0 31 2 47 4 63 9 表 3 在实现本发明的过程中, 发明人经过研究发现: 上述现有去除块效应的 方法中, 为确定滤波过程而获得 BS时, 是通过两次查表后, 才能够得到相 似阔值, 由于没有充分利用块边界两侧块与块之间的关系和头信息, 因而判 断得到的 BS不准确, 进而使得目前去除块效应的方法, 不能釆用有效的滤 波过程去除块效应, 因而去除块效应的性能较差。
综上所述, 目前获得块边界滤波强度的方法, 得到的 BS不准确, 进而 使得目前去除块效应的方法, 无法有效的降低块效应。
发明内容
本发明实施例提供一种获得块边界滤波强度的方法, 该方法能够提高获 得的块边界滤波强度的准确性。
本发明实施例提供一种获得块边界滤波强度的方法, 该方法能够提高获 得的块边界滤波强度的准确性。
本发明实施例提供一种去除块效应的方法, 该方法能够提高获得的块边 界滤波强度的准确性, 从而有效降低块效应。
本发明实施例还提供一种获得块边界滤波强度的装置, 该装置能够提高 获得的块边界滤波强度的准确性。
本发明实施例还提供一种去除块效应的装置, 该装置能够提高获得的块 边界滤波强度的准确性, 从而有效降低块效应。
为达到上述目的, 本发明实施例的技术方案具体是这样实现的: 一种获得块边界滤波强度的方法, 该方法包括: 确定相似阔值, 利用所 述相似阔值计算当前窗格内块边界区域平坦度; 根据所述区域平坦度所在的 区域平坦度区间, 确定每个窗格内块边界的滤波强度初值, 每一个所述区域 平坦度区间对应一个滤波强度初值; 利用预先获取到的所述块边界两侧的块 的头信息, 调整所述块边界滤波强度初值, 得到所述块边界滤波强度。 一种获得块边界滤波强度的方法, 该方法包括: 根据块边界两侧的块的 头信息, 判断是否需要对所述块边界两侧的块的像素进行滤波, 如果否, 则 确定相似阔值, 利用所述相似阔值计算区域平坦度, 所述块边界两侧的块的 头信息包括: 类型信息或编码块模式信息; 根据所述区域平坦度所在的区域 平坦度区间, 确定每个窗格内块边界的滤波强度初值, 每一个所述区域平坦 度区间对应一个滤波强度初值; 利用预先获取到的所述块边界两侧的块的头 信息, 调整所述滤波强度初值, 得到所述块边界滤波强度。
一种去除块效应的方法, 该方法包括: 确定相似阔值, 利用所述相似阔 值计算区域平坦度; 由所述区域平坦度获得块边界滤波强度, 利用所述块边 界两侧块内各个相邻像素间的一阶差分和二阶差分判断所述块边界是否是伪 边界, 在判断出所述块边界是伪边界时, 根据预先建立的块边界滤波强度与 滤波器组合的对应关系, 由获得的块边界滤波强度, 确定出相应的滤波器组 合, 对所述块边界两侧的块的像素进行相应的滤波处理。
一种获得块边界滤波强度的装置, 该装置包括:
平坦度计算模块, 用于确定相似阈值, 利用所述相似阈值计算当前窗格 内块边界区域平坦度;
初值确定模块, 用于根据所述平坦度计算模块计算出的区域平坦度所在 的区域平坦度区间, 确定每个窗格内块边界的滤波强度初值, 每一个所述区 域平坦度区间对应一个滤波强度初值;
调整模块, 用于利用预先获取到的所述块边界两侧的块的头信息, 调整 所述初值确定模块确定的所述块边界滤波强度初值, 得到所述块边界滤波强 度。
一种去除块效应的装置, 该装置包括:
块边界滤波强度获取模块, 用于确定相似阔值, 利用所述相似阔值计算 区域平坦度, 由所述区域平坦度获得块边界滤波强度;
边界判断模块, 用于判断所述块边界是否为伪边界;
滤波器确定模块, 用于在所述边界判断模块的判断结果表明所述块边界 是伪边界时, 根据预先建立的块边界滤波强度与滤波器组合的对应关系, 由 所述块边界滤波强度获取模块获得的块边界滤波强度, 确定出相应的滤波器 组合;
滤波模块, 用于釆用所述滤波器确定模块确定出的滤波器组合, 对所述 块边界两侧的块的像素进行相应的滤波处理。
由上述技术方案可见, 本发明实施例提供的获得块边界滤波强度和去除 块效应的方法和装置, 根据预先查表得到的块边界阔值确定相似阔值, 或者 由量化参数、 宏块模式、 色彩空间或帧类型确定相似阔值, 并利用块的头信 息调整该块边界滤波强度初值, 能够不受限于编码块单位, 充分利用两侧块 与块之间的关系和块的头信息, 因而提高获得的块边界滤波强度的准确性, 从而有效降低块效应。
附图说明
图 1为现有技术中相似度判别样本示例示意图;
图 2为块边界两边一行 /列图像素分布示意图;
图 3为本发明一个实施例中获得块边界滤波强度的方法流程示意图; 图 4为本发明一个实施例中窗格大小 N为 4样本示例示意图; 图 5本发明一个实施例中窗格大小 N为 2样本示例示意图;
图 6为本发明另一个实施例中利用获得的块边界滤波强度实现去除块效 应的方法流程示意图;
图 7为本发明另一个实施例中块边界两边一行图像素分布示意图; 图 8为本发明再一个实施例中获得块边界滤波强度的装置结构示意图; 图 9为本发明又一个实施例中去除块效应的装置结构示意图。
具体实施方式
为使本发明的目的、 技术方案及优点更加清楚明白, 以下参照附图并举 实施例, 对本发明作进一步详细说明。
本发明实施例提供了一种利用像素和头信息获得块边界滤波强度的方 法, 具体为: 确定相似阔值, 利用所述相似阔值计算区域平坦度; 根据所述 区域平坦度所在的区域平坦度区间, 确定块边界的滤波强度初值, 每一个所 述区域平坦度区间对应一个滤波强度初值; 利用预先获取到的所述块边界两 侧的块的头信息, 调整所述滤波强度初值, 得到所述块边界滤波强度。
或者, 先根据块边界两侧的块的头信息, 判断是否需要对所述块边界两 侧的块的像素进行滤波, 块的头信息包括类型信息或编码块模式信息, 如果 是, 结束流程; 否则, 确定相似阔值, 利用所述相似阔值计算区域平坦度; 根据所述区域平坦度所在的区域平坦度区间, 确定每个窗格内块边界的滤波 强度初值, 每一个所述区域平坦度区间对应一个滤波强度初值; 利用预先获 取到的所述块边界两侧的块的头信息。
本发明实施例还提供了一种去除块效应的方法,具体为: 确定相似阔值, 利用所述相似阔值计算区域平坦度;由所述区域平坦度获得块边界滤波强度, 利用所述块边界两侧块内各个相邻像素间的一阶差分和二阶差分判断所述块 边界是否是伪边界, 在判断出所述块边界是伪边界时, 根据预先建立的块边 界滤波强度与滤波器组合的对应关系, 由获得的块边界滤波强度, 确定出相 应的滤波器组合, 对所述块边界两侧的块的像素进行相应的滤波处理。
图 3为本发明一个实施例中获得块边界滤波强度的方法流程示意图。 如 图 3所示, 具体包括以下步骤:
步骤 301〜步骤 302为计算当前窗格内块边界区域平坦度的过程。
步骤 301 : 将当前滤波块边界分成若干个窗格。
本实施例中, 划分窗格的方法为: 预先设置窗格大小, 沿着垂直块边界 的方向, 按照所述窗格大小, 将块边界两侧的块划分成一个以上窗格。 窗格 大小为 N, N可调, 在本实施例中, 假设对于 8x8变换块, 其窗格大小 N = 4; 如图 4所示,图 4为本发明一个实施例中窗格大小 N为 4样本示例示意图。 当然也可以 N = 2, 如图 5所示或其它值。 当然也可不划分窗格, 以块为单位 进行以下的步骤。
步骤 302: 计算当前窗格内块边界的区域平坦度。
本实施例中, 计算区域平坦度的过程包括: 在当前窗格内, 分别计算块 边界两侧块的像素平坦度; 累加两侧块的像素平坦度, 得到窗格内的区域平 坦度。 也就是说, 当前窗格内块边界的区域平坦度是通过当前窗格内块边界 两边的像素平坦度累加和得到, 当前窗格内块边界区域平坦度由当前窗格内 次块边界相似点的个数确定, 按照如下公式(5)计算: num =∑ ((abs (Ln - H <Τ) Ι:0) + ^ ((abs (Rn-Kn)<T)l\ :0)
( 5 ) 其中, 像素平坦度是通过一阶差分和阔值的比较来进行的, 相似阔值 T 可调, 对于本实施例 Τ= (β + 2) / 3。 β值见表 1, 用于索引值 IndexB按如 下公式(6)计算:
IndexB = ^Ρ Δ
ι + Offset— β (6)
其中, QPm是当前窗格内块边界两边块 m的 QP值, Offset— β值一个预 先设置的可以调节滤波强度的偏移量, 该 Offset— β值可从传递的视频码流中 获取到。
步骤 303〜步骤 304为根据平坦度和头信息获得 BS的过程, BS可以分为 多级, 在本实施例中 BS被分为四级, 也就是 BS值可以为 0、 1、 2和 3。
步骤 303: 计算当前块边界滤波强度初值。
本实施例中, 计算当前块边界滤波强度初值 bs方法具体为:
若 num<C0, bs = 0;
若 num>=C0且 num<Cl , bs = 1;
若 num>=Cl, bs = 2;
式中的 CO和 CI为根据统计规律,预先设置的平坦度阔值,本实施例中,
C0设置为 2, C1设置为 7。
步骤 304: 根据头信息对块边界强度修正得到块边界滤波强度。
本实施例中, 如果头信息中的信息表明相邻两个块都是跳过模式(skip mode), 则当前块不进行滤波; 如果表明有一个块是帧内块, 则当前块边界 滤波强度初值 bs加 1; 如果头信息中的信息表明块边界两边块的运动矢量之 差大于预先设置的阔值 Tmv, 在本实施例中, Tmv取 3个整像素, 则当前 bs增加 1, 或者如果头信息中的信息表明块边界两边区域分别来自不同参考 帧, 则 bs增加 1; 如果块边界两侧块的编码块模式同时为 0, 则 BS为 0。
本实施例中, 获得 BS的方法中只根据窗格内块边界两侧的块的相似点 数 num得到当前块边界滤波强度初值 bs,因而只需确定一个窗格内平坦度的 相似阈值 T, 不必进行如现有技术中查找表 1的方法就可以得出 num, 因而 本实施例获得 BS的方法更方便灵活。
图 6为本发明另一个实施例中利用获得的块边界滤波强度实现去除块效 应的方法流程示意图。 如图 6所示, 具体包括以下步骤:
步骤 601〜步骤 604与步骤 301〜步骤 304相似, 只是在步骤 602中, 窗 格内平坦度的相似阔值 T是由头信息来确定的, 定义 T = (xa+ yp) /z。 这里, x、 y、 z将根据量化参数、 宏块模式即帧内宏块(Intra)和帧间宏块 Inter )、 色彩空间即亮度( luma )和色度( chroma )和帧类型即帧内预测帧( I )、 双向预测帧 (B)和单向预测帧 (P) 的不同进行自适应的调整。 如:
I 0≤QP<20
3
T = 20≤QP< 40
3
β + 2
40≤QP< 64
3
Inter
Intra
Figure imgf000017_0001
α+β + 4
Luma
8
T =
α+β + l
Chroma
8
步骤 605: 判断块边界是否为真实块边界, 如果是, 不需要进行滤波, 结束流程; 否则, 执行步骤 606。
本实施例中, 判断块边界是否为真块边界的方法为: 判断块边界两侧块 内各个相邻像素间的差值是否大于某一设定的边界阔值, 如果是, 则是图像 真实边缘, 不需要进行滤波; 否则是伪边缘, 需要进行滤波。
在本实施例中, 假设块边界为垂直块边界, 则块边界两边的一行图象像 素分布如图 7所示, 当然既可以是块边界, 也可以是水平块边界, 则需要使 用块边界两边一列图象像素判断块边界是否为真块边界, 对图 7中给定块的 块边界两边一对像素 R0和 L0 , 小块内部两对像素 L0和 LI , R0和 R1分别 求差值。 即用 R0和 L0相减, 得到一差值 Ml ; 用 R0减 R1 , 得到一差值 M2; 用 L0减 L1 , 得到一差值 M3; 用 Ml加 M2, 得到 M4; 用 Ml加 M3 , 得到 M5。
在最终的判别公式中,如果 Ml小于 o且大于 1 , M2小于 β且 M3小于 β,
Μ4和 Μ5有任意一个大于 0, 则认为图像的特征边缘不处于块的边缘, 也就 是 BS不等于 0, 可以进行滤波; 反之则不滤。 即:
Bs != 0 && Abs(R0 - L0) < α && Abs(R0 - L0)>1 && Abs(R0 - l) < β && Abs(L0 - LI) < β && (Abs(2 R0 - Rl- L0)>0||Abs(2 L0 - LI- R0)>0)
式中, 符号"! = "表示不等于, 符号"『表示求"或"运算。
步骤 606: 对伪块边界, 根据非零 BS的不同级别, 确定出相应的滤波过 程。
本实施例中, 对于 BS分为四级, 其中, 当 BS=0时, 不进行滤波, 在 如图 7所示的块边界两边一行 /列图象像素分布图中,进行滤波三种过程具体 为:
过程一: 当 BS = 1时, 釆用均值滤波方式进行滤波:
R0 = aq? (2*Rl+R0+L0+2)»2: ( Rl+2*R0+L0+2) »2;
L0 = ap? (2*Ll+L0+R0+2)»2: ( Ll+2*L0+R0+2) »2;
过程二: 当 BS=2时, 采用均值滤波方式进行滤波:
R0 = aq? ( R2+2 *R1 +2 *R0+2 *L0+L 1 +4 ) »3: (R2+2*Rl+3*R0+2*L0+4)
»3;
L0 = ap? ( L2+2 *L 1 +2 *L0+2 *R0+R1 +4 ) »3: (L2+2*Ll+3*L0+2*R0+4) » 3;
Rl = aq? ( 2*R2 + 4*R1 + RO + LO + 4 ) » 3 : Rl;
Ll= ap? ( 2*L2 + 4*L1 + LO + RO + 4 ) » 3 : LI;
过程三: 当 BS=3时, 仍釆用均值滤波方式进行滤波:
R0 = ( 3*R2+3*Rl+4*R0+3*L0+3*Ll+8) » 4;
LO = ( 3*L2+3*Ll+4*L0+3*R0+3*Rl+8) » 4;
Rl = aq ? (3*R2+4*Rl+4*R0+3*L0+2*Ll+8) » 4 : (2*R2+4*Rl+R0+L0+4) » 3;
LI = ap ? (3*L2+4*Ll+4*L0+3*R0+2*Rl+8) » 4 : ( 2*L2+4*Ll+L0+R0+4) » 3;
R2= aq? (2*R3+3*R2+R1 + RO +L0+ 4 )»3: R2;
L2 = ap ? (2*L3+3*L2+L1 + LO +R0+ 4 )»3: L2;
其中, 过程二、 三和四中的 ap和 aq分别为:
ap=Abs(R2-R0) < β && Abs(R。 - L。)< ((α» 2) + 2),
aq=Abs(L2-L0) < β && Abs(R0 - L0)< (( » 2) + 2)。
对于色度块, 只对块边界两边 RQ和 LQ像素釆用同样的滤波方法, 不对 R2、 Rl、 L1和 L2滤波。
本实施例中是以对垂直块边界的滤波处理过程为例进行的说明, 对于水 平块边界, 其滤波窗口处于竖直状态, 但相应水平块边界上的滤波处理过程 也相同, 故不再详述。
本实施例的步骤 604中,还可以根据 skip mode或者 cbp来对 BS的值进 行相应的调整, 例如, 当相邻两个块都是 skip mode或者 cbp全为 0时, BS 减一。
另外, 还可以在步骤 601之前, 先利用块的头信息判断是否需要进行滤 波, 比如根据 skip mode或者 cbp来判断是否需要进行滤波, 当相邻块边界 两侧的块都是 skip mode或者 cbp全为 0时, 则判断出不需要进行滤波, 结 束流程。 本实施例根据每行像素在滤波时的权重也可以不一样, 比如根据宏块类 型确定块边界的每行像素的权重, 则根据每行像素的权重, 釆用不同强度滤 波器进行滤波。 可以对块的块边界行釆用相对较强滤波, 对块的中间行采用 相对较弱滤波的加权策略进行滤波处理。尤其用于 Intra块的情况。如可以将 待滤波像素行为当前块的块边界行, 比如宏块的 0、 7、 8、 15行时, 釆用当 前 BS值所属滤波器中最强的一组进行滤波。
本实施例中的滤波器可共享, 具体过程可以将 BS较大的弱滤波器作为 BS较小的强滤波器, 或者将 BS较小的强滤波器作为 BS较大的弱滤波器, 使得在 BS相同的情况下有更多不同强度的滤波器可选择, 增加滤波的准确 性, 而且不增加运算的复杂度。 例如, 将 BS为 3时的弱滤波器作为 BS为 2 时的强滤波器, BS为 2时的弱滤波器可以作为 BS为 1时的强滤波器。 再根 据平坦度值可以将当前窗格内块边界划分为不同级别的 BS值, 而对每级的 滤波器也可以选取不同的组合, 但总体的原则是全部釆用均值滤波方式, 并 且对于 BS级别越高的, 用越强的滤波器进行滤波。
本实施例利用获得的 BS,对窗格内的属于伪边界的行 /列像素进行滤波, 可以有效的去除因变换量化和运动补偿引起的块效应。
图 8为本发明再一个实施例中获得块边界滤波强度的装置结构示意图, 该装置包括: 平坦度计算模块 810、 初值确定模块 820和调整模块 830。
具体地, 平坦度计算模块 810 , 确定相似阔值, 利用所述相似阔值计算 区域平坦度。
初值确定模块 820, 根据所述区域平坦度所在的区域平坦度区间, 确定 每个窗格内块边界的滤波强度初值, 每一个所述区域平坦度区间对应一个滤 波强度初值。
调整模块 830, 利用预先获取到的所述块边界两侧的块的头信息, 调整 滤波强度初值, 得到块边界滤波强度。
图 9为本发明又一个实施例中去除块效应的装置结构示意图, 该装置包 括: 块边界滤波强度获取模块 910、 边界判断模块 920、 滤波器确定模块 930 和滤波模块 940。
具体地, 块边界滤波强度获取模块 910 , 用于确定相似阔值, 利用所述 相似阈值计算区域平坦度, 由所述区域平坦度获得块边界滤波强度; 边界判 断模块 920, 判断所述块边界是伪边界, 向滤波器确定模块 930输入判断结 果。
滤波器确定模块 930 , 在边界判断模块的判断结果表明块边界是伪边界 时, 根据预先建立的块边界滤波强度与滤波器组合的对应关系, 由块边界滤 波强度获取模块 910获得的块边界滤波强度, 确定出相应的滤波器组合。
滤波模块 940 , 釆用滤波器确定模块 930确定出的滤波器组合, 对窗格 内的所述块边界两侧的块的像素进行相应的滤波处理。
本领域普通技术人员可以理解实现上述实施例方法携带的全部或部分步 骤是可以通过程序来指令相关的硬件完成 ,所述的程序可以存储于一种计算机 可读存储介质中, 该程序在执行时, 包括方法实施例的步骤之一或其组合。
另外, 在本发明各个实施例中的各功能单元可以集成在一个处理模块中 , 也可以是各个单元单独物理存在,也可以两个或两个以上单元集成在一个模块 中。上述集成的模块既可以采用硬件的形式实现, 也可以釆用软件功能模块的 形式实现。所述集成的模块如果以软件功能模块的形式实现并作为独立的产品 销售或使用时, 也可以存储在一个计算机可读取存储介质中。
上述提到的存储介质可以是只读存储器, 磁盘或光盘等。
以上所述仅为本发明的较佳实施例而已, 并非用于限定本发明的保护范 围。 凡在本发明的精神和原则之内所作的任何修改、 等同替换、 改进等, 均 应包含在本发明的保护范围之内。

Claims

权 利 要 求
1、 一种获得块边界滤波强度的方法, 其特征在于, 包括:
确定相似阔值, 利用所述相似阔值计算当前窗格内块边界区域平坦度; 根据所述区域平坦度所在的区域平坦度区间, 确定每个窗格内块边界的 滤波强度初值, 每一个所述区域平坦度区间对应一个滤波强度初值;
利用预先获取到的所述块边界两侧的块的头信息, 调整所述块边界滤波 强度初值, 得到所述块边界滤波强度。
2、根据权利要求 1所述的方法, 其特征在于, 所述利用所述相似阔值计 算区域平坦度的过程包括:
分别计算所述块边界两侧的像素与同一个块中相邻像素之间的差值, 所 述相邻像素为沿着所述块边界垂直方向的相邻像素;
当任意一个所述差值的绝对值小于预先设置的相似阔值时, 所述块边界 两侧的像素平坦度加 1 ;
将所述像素平坦度累加得到区域平坦度。
3、根据权利要求 1所述的方法, 其特征在于, 所述确定相似阔值的过程 包括: 根据预先查表得到的块边界阔值确定所述相似阔值。
4、根据权利要求 3所述的方法, 其特征在于, 所述根据预先查表得到的 块边界阔值确定所述相似阔值的过程进一步包括:
预先设置块边界阔值的加权系数和增量, 利用所述加权系数对所述块边 界阔值加权, 再与所述增量求和, 确定出所述相似阔值。
5、根据权利要求 3所述的方法, 其特征在于, 所述根据预先查表得到的 块边界阔值确定所述相似阔值的过程进一步包括:
由量化参数、 宏块模式、 色彩空间或帧类型确定每一个所述块边界阈值 的加权系数, 利用所述加权系数对每一个所述块边界阔值加权, 对加权后的 每一个所述块边界阔值求和, 确定出所述相似阔值。
6、根据权利要求 1所述的方法, 其特征在于, 所述块边界两侧的块的头 信息包括: 所述块边界两侧的块的类型信息; 所述调整所述块边界滤波强度初值的过程进一步包括:
在所述块边界两侧的块均为跳过模式类型时, 所述滤波强度初值减去预 先设置的任意整数步长, 作为调整后得到的块边界滤波强度;
在所述块边界两侧的块中的任意一个为帧内块类型时, 所述滤波强度初 值加上预先设置的任意整数步长, 作为调整后得到的块边界滤波强度。
7、根据权利要求 1所述的方法, 其特征在于, 所述块边界两侧的块的头 信息包括: 所述块边界两侧的块的参考帧信息;
所述调整所述块边界滤波强度初值的过程进一步包括:
在所述块边界两侧的块来自不同的参考帧时, 所述滤波强度初值加上预 先设置的任意整数步长, 作为调整后得到的块边界滤波强度。
8、根据权利要求 1所述的方法, 其特征在于, 所述块边界两侧的块的头 信息包括: 所述块边界两侧的块的运动矢量信息;
所述调整所述块边界滤波强度初值的过程进一步包括:
在所述块边界两侧的块的运动矢量大于预先设置的矢量预置时, 所述滤 波强度初值加上预先设置的任意整数步长, 作为调整后得到的块边界滤波强 度。
9、根据权利要求 1所述的方法, 其特征在于, 所述块边界两侧的块的头 信息包括: 所述块边界两侧的块的编码块模式信息;
所述调整所述块边界滤波强度初值的过程进一步包括:
在所述块边界两侧的块均为编码块模式为 0时, 所述滤波强度初值减去 任意整数步长, 作为调整后得到的块边界滤波强度。
10、 一种获得块边界滤波强度的方法, 其特征在于, 包括:
根据块边界两侧的块的头信息, 判断是否需要对所述块边界两侧的块的 像素进行滤波, 如果否, 则确定相似阔值, 利用所述相似阔值计算区域平坦 度, 所述块边界两侧的块的头信息包括: 类型信息或编码块模式信息; 根据所述区域平坦度所在的区域平坦度区间, 确定每个窗格内块边界的 滤波强度初值, 每一个所述区域平坦度区间对应一个滤波强度初值; 利用所述块边界两侧的块的头信息, 调整所述滤波强度初值, 得到所述 块边界滤波强度。
11、根据权利要求 10所述的方法, 其特征在于, 所述块边界两侧的块的 头信息还包括: 所述块边界两侧的块的类型信息;
所述判断是否需要对所述块边界两侧的块的像素进行滤波包括: 在所述块边界两侧的块均为跳过模式类型时, 判断出不需要进行滤波; 否则, 需要进行滤波。
12、根据权利要求 10所述的方法, 其特征在于, 所述块边界两侧的块的 头信息还包括: 所述块边界两侧的块的编码块模式信息;
所述判断是否需要所述块边界两侧的块的像素进行滤波包括:
在所述块边界两侧的块均为编码块模式为 0时,判断出不需要进行滤波; 否则, 需要进行滤波。
13、 一种去除块效应的方法, 其特征在于, 包括:
确定相似阔值, 利用所述相似阔值计算区域平坦度;
由所述区域平坦度获得块边界滤波强度, 利用所述块边界两侧块内各个 相邻像素间的一阶差分和二阶差分, 判断所述块边界是否是伪边界;
在判断出所述块边界是伪边界时, 根据预先建立的块边界滤波强度与滤 波器组合的对应关系, 由获得的块边界滤波强度,确定出相应的滤波器组合, 对所述块边界两侧的块的像素进行滤波处理。
14、根据权利要求 13所述的方法, 其特征在于, 所述判断块边界是否是 伪边界的过程包括:
计算所述块边界两侧块内各个相邻像素间的一阶差分和二阶差分, 当每 一个一阶差分和二阶差分均处于由预先设置的边界阔值确定的区间中时, 判 断出所述块边界是伪边界, 否则是真边界。
15、 根据权利要求 13所述的方法, 其特征在于, 所述滤波器组合包括: 强滤波器和弱滤波器;
建立所述块边界滤波强度与滤波器组合的对应关系的过程包括: 不同的 块边界滤波强度对应相同的滤波器, 块边界滤波强度较小者将所述滤波器作 为对应的强滤波器, 块边界滤波强度较大者将所述滤波器作为对应的弱滤波 器。
16、根据权利要求 13所述的方法, 其特征在于, 所述块边界两侧像素为 块边界左侧四个和右侧四个像素, 设从块边界向右的四个像素依次为 R0、
Rl、 R2和 R3 , 从块边界向左的四个像素依次为 L0、 Ll、 L2和 L3 , 则所述 滤波器组合为以下一个或任意多个均值滤波器的组合:
(2*Rl+R0+L0+2)/4 、 ( Rl+2*R0+L0+2)/4 、 (2*Ll+L0+R0+2)/4 、 ( Ll+2*L0+R0+2)/4 、 ( R2+2*Rl+2*R0+2*L0+Ll+4 )/8 、 (R2+2*Rl+3*R0+2*L0+4)/8 、 ( L2+2*Ll+2*L0+2*R0+Rl+4 ) /8 、 (L2+2 *L 1 +3 *L0+2 *R0+4) /8、 ( 2*R2 + 4*R1 + R0 + L0 + 4 ) /8 , ( 2*L2 + 4*L1 + L0 + R0 + 4 ) /8 、 ( 3*R2+3*Rl+4*R0+3*L0+3*Ll+8) /16 、 ( 3*L2+3*Ll+4*L0+3*R0+3*Rl+8) /16、 (3 *R2+4 *R1 +4 *R0+3 *L0+2 *L 1 +8) /16、 (2*R2+4*Rl+R0+L0+4) /8、 (3*L2+4*Ll+4*L0+3*R0+2*Rl+8) /16、 ( 2*L2+4*Ll+L0+R0+4) /8、 (2*R3+3*R2+R1 + R0 +L0+ 4 ) /8、 或 (2*L3+3*L2+L1 + L0 +R0+ 4 ) /8。
17、根据权利要求 13所述的方法, 其特征在于, 所述对块边界两侧的块 的像素进行滤波处理的过程包括:
根据预先设置的所述块边界两侧的块中每行像素的权重, 从确定出的所 述滤波器组合中选择出一个滤波器, 对所述块边界两侧的块中像素进行滤波 处理。
18、 一种获得块边界滤波强度的装置, 其特征在于, 包括:
平坦度计算模块, 用于确定相似阔值, 利用所述相似阔值计算当前窗格 内块边界区域平坦度;
初值确定模块, 用于根据所述平坦度计算模块计算出的区域平坦度所在 的区域平坦度区间, 确定每个窗格内块边界的滤波强度初值, 每一个所述区 域平坦度区间对应一个滤波强度初值; 调整模块, 用于利用预先获取到的所述块边界两侧的块的头信息, 调整 所述初值确定模块确定的所述块边界滤波强度初值, 得到所述块边界滤波强 度。
19、 一种去除块效应的装置, 其特征在于, 包括:
块边界滤波强度获取模块, 用于确定相似阔值, 利用所述相似阔值计算 区域平坦度, 由所述区域平坦度获得块边界滤波强度;
边界判断模块, 用于判断所述块边界是否为伪边界;
滤波器确定模块, 用于在所述边界判断模块的判断结果表明所述块边界 是伪边界时, 根据预先建立的块边界滤波强度与滤波器组合的对应关系, 由 所述块边界滤波强度获取模块获得的块边界滤波强度, 确定出相应的滤波器 组合;
滤波模块, 用于釆用所述滤波器确定模块确定出的滤波器组合, 对所述 块边界两侧的块的像素进行相应的滤波处理。
PCT/CN2008/072082 2007-09-13 2008-08-21 Method and device for acquiring block boundary strength and reducing block distortion WO2009033390A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN 200710149478 CN101389016B (zh) 2007-09-13 2007-09-13 获得块边界强度和去除块效应的方法和装置
CN200710149478.0 2007-09-13

Publications (1)

Publication Number Publication Date
WO2009033390A1 true WO2009033390A1 (en) 2009-03-19

Family

ID=40451576

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2008/072082 WO2009033390A1 (en) 2007-09-13 2008-08-21 Method and device for acquiring block boundary strength and reducing block distortion

Country Status (2)

Country Link
CN (1) CN101389016B (zh)
WO (1) WO2009033390A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2454877A2 (en) * 2009-08-04 2012-05-23 Samsung Electronics Co., Ltd. Apparatus and method for deblocking filtering image data and video decoding apparatus and method using the same
US20180054613A1 (en) * 2016-08-22 2018-02-22 Mediatek Inc. Video encoding method and apparatus with in-loop filtering process not applied to reconstructed blocks located at image content discontinuity edge and associated video decoding method and apparatus

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102098501B (zh) * 2009-12-09 2013-05-08 中兴通讯股份有限公司 一种视频图像去块效应的方法和装置
KR101813189B1 (ko) * 2010-04-16 2018-01-31 에스케이 텔레콤주식회사 영상 부호화/복호화 장치 및 방법
KR20110123651A (ko) * 2010-05-07 2011-11-15 한국전자통신연구원 생략 부호화를 이용한 영상 부호화 및 복호화 장치 및 그 방법
WO2012044074A2 (ko) * 2010-09-28 2012-04-05 삼성전자 주식회사 적응적 필터링 방법 및 장치
KR101910502B1 (ko) * 2011-01-10 2018-10-22 삼성전자주식회사 변환계수를 이용한 영상 블록현상 제거 방법 및 그 장치
KR102030977B1 (ko) 2011-02-22 2019-10-10 타지반 투 엘엘씨 필터 방법, 동화상 부호화 장치, 동화상 복호 장치 및 동화상 부호화 복호 장치
WO2012114725A1 (ja) 2011-02-22 2012-08-30 パナソニック株式会社 画像符号化方法、画像復号化方法、画像符号化装置、画像復号化装置および画像符号化復号化装置
CN102123285B (zh) * 2011-03-23 2015-10-28 深圳市云宙多媒体技术有限公司 一种块效应去噪方法、装置
CN103563374B (zh) * 2011-05-27 2017-02-08 索尼公司 图像处理设备及方法
EP4336841A3 (en) 2011-07-19 2024-03-20 Tagivan Ii Llc Coding method
KR102094939B1 (ko) * 2011-11-25 2020-03-31 선 페이턴트 트러스트 화상 처리 방법 및 화상 처리 장치
AR092786A1 (es) * 2012-01-09 2015-05-06 Jang Min Metodos para eliminar artefactos de bloque
MY195620A (en) 2012-01-17 2023-02-02 Infobridge Pte Ltd Method Of Applying Edge Offset
CN103957425B (zh) * 2014-04-23 2018-01-23 华为技术有限公司 图像处理方法及系统
US20180192074A1 (en) * 2017-01-03 2018-07-05 Mediatek Inc. Video processing method for processing projection-based frame with 360-degree content represented by projection faces packed in 360-degree virtual reality projection layout
CN107222743B (zh) * 2017-07-04 2020-02-04 西安万像电子科技有限公司 一种图像处理方法、装置和系统
CN108337509A (zh) * 2018-04-06 2018-07-27 北京慧摩尔科技有限公司 块状噪声的降噪方法及装置
WO2020073904A1 (en) * 2018-10-08 2020-04-16 Huawei Technologies Co., Ltd. An image processing device and method for performing deblocking

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030026337A1 (en) * 2001-06-15 2003-02-06 Lg Electronics Inc. Loop filtering method in video coder
CN1552161A (zh) * 2002-07-11 2004-12-01 松下电器产业株式会社 滤波强度的决定方法、动态图像编码方法及解码方法
US20060126962A1 (en) * 2001-03-26 2006-06-15 Sharp Laboratories Of America, Inc. Methods and systems for reducing blocking artifacts with reduced complexity for spatially-scalable video coding
CN1867075A (zh) * 2005-09-19 2006-11-22 华为技术有限公司 图像编码处理中的环路滤波方法

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1285214C (zh) * 2004-01-14 2006-11-15 华中科技大学 一种环路滤波方法和环路滤波器

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060126962A1 (en) * 2001-03-26 2006-06-15 Sharp Laboratories Of America, Inc. Methods and systems for reducing blocking artifacts with reduced complexity for spatially-scalable video coding
US20030026337A1 (en) * 2001-06-15 2003-02-06 Lg Electronics Inc. Loop filtering method in video coder
CN1552161A (zh) * 2002-07-11 2004-12-01 松下电器产业株式会社 滤波强度的决定方法、动态图像编码方法及解码方法
CN1867075A (zh) * 2005-09-19 2006-11-22 华为技术有限公司 图像编码处理中的环路滤波方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2454877A2 (en) * 2009-08-04 2012-05-23 Samsung Electronics Co., Ltd. Apparatus and method for deblocking filtering image data and video decoding apparatus and method using the same
EP2454877A4 (en) * 2009-08-04 2012-12-26 Samsung Electronics Co Ltd DEVICE AND METHOD FOR DISABLEING THE FILING OF IMAGE DATA AND VIDEO ENCRYPTING DEVICE AND METHOD THEREFOR
US20180054613A1 (en) * 2016-08-22 2018-02-22 Mediatek Inc. Video encoding method and apparatus with in-loop filtering process not applied to reconstructed blocks located at image content discontinuity edge and associated video decoding method and apparatus
CN109219958A (zh) * 2016-08-22 2019-01-15 联发科技股份有限公司 对位于图像内容不连续性边缘处的重构块不应用环路滤波处理的视频编码方法和设备和相关的视频解码方法和设备

Also Published As

Publication number Publication date
CN101389016A (zh) 2009-03-18
CN101389016B (zh) 2010-10-20

Similar Documents

Publication Publication Date Title
WO2009033390A1 (en) Method and device for acquiring block boundary strength and reducing block distortion
WO2007033551A1 (en) A loop filtering method for use in image encoding process
TWI538521B (zh) 用於減少區塊邊界處之區塊效應假影之方法、解塊濾波器單元、編碼器、解碼器、使用者設備、網路器件、電腦程式及電腦程式產品
EP1555832A2 (en) Adaptive loop filtering for reducing blocking artifacts
TWI418218B (zh) 用於基於區塊之數位視訊之解塊及邊緣增強之方法、裝置、電子元件、電腦可讀媒體及處理器
US7911538B2 (en) Estimation of block artifact strength based on edge statistics
EP2664149B1 (en) Deblocking filtering
JP2006513633A (ja) エラー隠蔽中に生成されるアーチファクトをスムージングするデコーダ装置及び方法
EP1723796A1 (en) Method, medium, and filter removing a blocking effect
WO2004064397A1 (en) Spatial error concealment based on the intra-prediction modes transmitted in a coded stream
WO2006041305A1 (en) Deblocking filter
KR20050074906A (ko) 루프 필터링 방법 및 루프 필터
JP2011523300A (ja) ビデオのコーダとデコーダとの同時最適化のための方法およびシステム
CN101321277A (zh) 去块效应的方法及装置
CN114710669A (zh) 视频编解码方法/装置和相应非易失性计算机可读介质
CN101267560A (zh) 去块滤波方法及装置
JP2008503177A (ja) 色差デブロッキングのための方法
CN101321276B (zh) 去块效应的方法及装置
US20180103274A1 (en) Deblocking filtering method and deblocking filter
JP2004215216A (ja) 画像処理装置、画像処理方法および画像処理プログラム
EP1721468A1 (en) Ringing artifact reduction for compressed video applications
WO2018177395A1 (zh) 一种对视频帧进行滤波的方法及装置
Kim et al. Adaptive deblocking technique for mobile video
US7844124B2 (en) Method of estimating a quantization parameter
JPH07307942A (ja) 画像雑音除去装置

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: 08784073

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: 08784073

Country of ref document: EP

Kind code of ref document: A1