H.266/VVC代碼學習:MIP技術相關代碼之initIntraMip函數

initIntraMip函數主要是對參考像素進行下采樣併爲MIP矩陣乘法準備輸入數據，函數結構如下：

MIP根據塊尺寸可以分爲以下三種情況：

	塊尺寸	下采樣後的邊界長度 m_reducedBdrySize	矩陣乘法輸出邊界長度 m_reducedPredSize
mipSizeId = 0	4x4	2	4
mipSizeId = 1	4xN、Nx4、8x8	4	4
mipSizeId = 2	其餘塊	4	8

initIntraMip函數主要是準備邊界參考像素，並調用prepareInputForPred函數爲MIP預測準備輸入數據

注意：MIP使用的參考像素是未經過濾波的參考像素

initIntraMip函數代碼如下所示：

void IntraPrediction::initIntraMip( const PredictionUnit &pu, const CompArea &area )
{
  CHECK( area.width > MIP_MAX_WIDTH || area.height > MIP_MAX_HEIGHT, "Error: block size not supported for MIP" );

  // prepare input (boundary) data for prediction
  // 準備輸入（邊界）數據進行預測
  // MIP使用未濾波的參考像素
  CHECK( m_ipaParam.refFilterFlag, "ERROR: unfiltered refs expected for MIP" );
#if JVET_R0350_MIP_CHROMA_444_SINGLETREE
  Pel       *ptrSrc     = getPredictorPtr(area.compID);//獲取參考像素
  const int  srcStride  = m_refBufferStride[area.compID];
  const int  srcHStride = 2;

  m_matrixIntraPred.prepareInputForPred(CPelBuf(ptrSrc, srcStride, srcHStride), area,
                                        pu.cu->slice->getSPS()->getBitDepth(toChannelType(area.compID)), area.compID);
#else
  Pel *ptrSrc = getPredictorPtr( COMPONENT_Y );
  const int srcStride  = m_refBufferStride[COMPONENT_Y];
  const int srcHStride = 2;

  m_matrixIntraPred.prepareInputForPred( CPelBuf( ptrSrc, srcStride, srcHStride ), area, pu.cu->slice->getSPS()->getBitDepth( CHANNEL_TYPE_LUMA ) );
#endif
}

prepareInputForPred函數主要分爲以下四個步驟：

Step 1: 保存塊大小並計算MIP相關參數，通過調用initPredBlockParams函數實現
Step 2: 獲取輸入數據（上一行參考像素和左一列參考像素）
Step 3: 通過Haar下采樣計算縮減邊界，通過boundaryDownsampling1D函數實現
Step 4: 推導矩陣乘法輸入向量

prepareInputForPred函數代碼如下所示

#if JVET_R0350_MIP_CHROMA_444_SINGLETREE
void MatrixIntraPrediction::prepareInputForPred(const CPelBuf &pSrc, const Area &block, const int bitDepth,
                                                const ComponentID compId)
{
  m_component = compId;
#else
void MatrixIntraPrediction::prepareInputForPred(const CPelBuf &pSrc, const Area& block, const int bitDepth)
{
#endif
  // Step 1: Save block size and calculate dependent values
  // Step 1: 保存塊大小並計算MIP相關參數
  initPredBlockParams(block);

  // Step 2: Get the input data (left and top reference samples)
  // Step 2: 獲取輸入數據（左上參考像素）
  // 獲取上一行參考像素
  m_refSamplesTop.resize(block.width);
  for (int x = 0; x < block.width; x++)
  {
    m_refSamplesTop[x] = pSrc.at(x + 1, 0);
  }
  // 獲取左一列參考像素
  m_refSamplesLeft.resize(block.height);
  for (int y = 0; y < block.height; y++)
  {
    m_refSamplesLeft[y] = pSrc.at(y + 1, 1);
  }

  // Step 3: Compute the reduced boundary via Haar-downsampling (input for the prediction)
  // Step 3: 通過Haar下采樣計算縮減邊界（預測輸入）
  // 下采樣後輸入向量的尺寸爲4或者8
  const int inputSize = 2 * m_reducedBdrySize;
  // 不需要轉置時，下采樣像素的順序：先上後左
  m_reducedBoundary          .resize( inputSize );
  // 轉置時，下采樣像素的順序：先左後上
  m_reducedBoundaryTransposed.resize( inputSize );

  int* const topReduced = m_reducedBoundary.data();
  boundaryDownsampling1D( topReduced, m_refSamplesTop.data(), block.width, m_reducedBdrySize );

  int* const leftReduced = m_reducedBoundary.data() + m_reducedBdrySize;
  boundaryDownsampling1D( leftReduced, m_refSamplesLeft.data(), block.height, m_reducedBdrySize );

  int* const leftReducedTransposed = m_reducedBoundaryTransposed.data();
  int* const topReducedTransposed  = m_reducedBoundaryTransposed.data() + m_reducedBdrySize;
  for( int x = 0; x < m_reducedBdrySize; x++ )
  {
    topReducedTransposed[x] = topReduced[x];
  }
  for( int y = 0; y < m_reducedBdrySize; y++ )
  {
    leftReducedTransposed[y] = leftReduced[y];
  }

  // Step 4: Rebase the reduced boundary
  // Step 4: 縮小邊界
  // 推導矩陣乘法輸入向量p，mipSizeId=0/1和mipSizeId=2的推導方法不一樣
  m_inputOffset       = m_reducedBoundary[0];
  m_inputOffsetTransp = m_reducedBoundaryTransposed[0];
  const bool hasFirstCol = (m_sizeId < 2);

  m_reducedBoundary          [0] = hasFirstCol ? ((1 << (bitDepth - 1)) - m_inputOffset      ) : 0; // first column of matrix not needed for large blocks
  m_reducedBoundaryTransposed[0] = hasFirstCol ? ((1 << (bitDepth - 1)) - m_inputOffsetTransp) : 0;
  for (int i = 1; i < inputSize; i++)
  {
    m_reducedBoundary          [i] -= m_inputOffset;
    m_reducedBoundaryTransposed[i] -= m_inputOffsetTransp;
  }
}

1、初始化MIP相關參數

initPredBlockParams函數是用來初始化MIP相關參數，主要是根據當前塊的尺寸來初始化mipSizeId，然後根據mipSizeId初始化下采樣後的邊界長度、矩陣乘法輸出邊界長度和上採樣因子

void MatrixIntraPrediction::initPredBlockParams(const Size& block)
{
  //獲得當前塊尺寸
  m_blockSize = block;
  // init size index
  // 根據當前塊尺寸初始化sizeId
  m_sizeId = getMipSizeId( m_blockSize );
  // init reduced boundary size
  // 初始縮減邊界尺寸
  // 對於4x4的塊寬度和高度分別縮減爲2個像素
  // 對於其餘尺寸的塊寬度和高度分別縮減爲4個像素
  m_reducedBdrySize = (m_sizeId == 0) ? 2 : 4;
  // init reduced prediction size
  // 初始化縮減預測後的尺寸
  // 對於mipSizeId = 0、1的塊，MIP預測後輸出4x4的塊
  // 對於mipSizeId = 2的塊，MIP預測後輸出8x8的塊
  m_reducedPredSize = ( m_sizeId < 2 ) ? 4 : 8;
  // init upsampling factors
  // 初始上採樣因子
  m_upsmpFactorHor = m_blockSize.width  / m_reducedPredSize;
  m_upsmpFactorVer = m_blockSize.height / m_reducedPredSize;

  CHECKD( (m_upsmpFactorHor < 1) || ((m_upsmpFactorHor & (m_upsmpFactorHor - 1)) != 0), "Need power of two horizontal upsampling factor." );
  CHECKD( (m_upsmpFactorVer < 1) || ((m_upsmpFactorVer & (m_upsmpFactorVer - 1)) != 0), "Need power of two vertical upsampling factor." );
}

2、下采樣

邊界參考像素的下采樣過程是由boundaryDownsampling1D函數實現的，下采樣過程其實就是對邊界參考像素求平均的過程，以8x8的塊爲例，如下圖所示，上一行存在8個參考像素，通過對兩兩相鄰的參考像素求平均後獲得4個下采樣後的參考像素，左一列參考像素地下采樣過程同理。

/*
一維下采樣
reducedDst表示下采樣後的邊界
fullSrc表示下采樣前的邊界
srcLen表示下采樣前的邊界長度
dstLen表示下采樣後的邊界長度
*/
void MatrixIntraPrediction::boundaryDownsampling1D(int* reducedDst, const int* const fullSrc, const SizeType srcLen, const SizeType dstLen)
{
  if (dstLen < srcLen)
  {
    //當下採樣後的邊界尺寸小於當前塊的邊界尺寸時，需要進行下采樣，下采樣操作即相當於求平均操作
    // Create reduced boundary by downsampling 通過下采樣創建縮小邊界
    const SizeType downsmpFactor = srcLen / dstLen;
    const int log2DownsmpFactor = floorLog2(downsmpFactor);
    const int roundingOffset = (1 << (log2DownsmpFactor - 1));

    SizeType srcIdx = 0;
    for( SizeType dstIdx = 0; dstIdx < dstLen; dstIdx++ )
    {
      int sum = 0;
      for( int k = 0; k < downsmpFactor; k++ )
      {
        sum += fullSrc[srcIdx++];
      }
      reducedDst[dstIdx] = (sum + roundingOffset) >> log2DownsmpFactor;
    }
  }
  else
  {
    // Copy boundary if no downsampling is needed 如果不需要下采樣，則複製邊界
    for (SizeType i = 0; i < dstLen; ++i)
    {
      reducedDst[i] = fullSrc[i];
    }
  }
}

下采樣過程結束後，根據mipTransposeFlag標誌將下采樣後的上參考像素和左參考像素排列成向量pTemp，排列方式如下：

mipTransposeFlag = 0時，pTemp = [redTop,redLeft]
mipTransposeFlag = 1時，pTemp = [redLeft,redTop]

3、推導矩陣乘法輸入向量

矩陣乘法輸入向量的推導方法和mipSizeId有關，輸入向量p的構造過程如下所示，其中inSize = 2*m_reducedBdrySize

H.266/VVC代碼學習:MIP技術相關代碼之initIntraMip函數

1、初始化MIP相關參數

2、下采樣

3、推導矩陣乘法輸入向量

H.266/VVC代碼學習:MIP技術相關代碼

H.266/VVC代碼學習:MIP技術相關代碼之initIntraMip函數

H.266/VVC代碼學習:MIP技術相關代碼之predIntraMip函數

H.266/VVC代碼學習：MPM列表建立

PTA 7-4 是否同一棵二叉搜索樹 (25分)

Mac下配置sublime實現LaTeX

https://yachay.unat.edu.pe/blog/index.php?comment_area=format_blog&comment_component=blog&comment_co

linux以太網驅動總結