arm_math.h

 * @param[in]       blockSize length of the input vector
 * @param[out]      *pResult maximum value returned here
 * @param[out]      *pIndex index of maximum value returned here
 * @return none.
 */

  void arm_max_q15(
  q15_t * pSrc,
  uint32_t blockSize,
  q15_t * pResult,
  uint32_t * pIndex);

/**
 * @brief Maximum value of a Q31 vector.
 * @param[in]       *pSrc points to the input buffer
 * @param[in]       blockSize length of the input vector
 * @param[out]      *pResult maximum value returned here
 * @param[out]      *pIndex index of maximum value returned here
 * @return none.
 */

  void arm_max_q31(
  q31_t * pSrc,
  uint32_t blockSize,
  q31_t * pResult,
  uint32_t * pIndex);

/**
 * @brief Maximum value of a floating-point vector.
 * @param[in]       *pSrc points to the input buffer
 * @param[in]       blockSize length of the input vector
 * @param[out]      *pResult maximum value returned here
 * @param[out]      *pIndex index of maximum value returned here
 * @return none.
 */

  void arm_max_f32(
  float32_t * pSrc,
  uint32_t blockSize,
  float32_t * pResult,
  uint32_t * pIndex);

  /**
   * @brief  Q15 complex-by-complex multiplication
   * @param[in]  *pSrcA points to the first input vector
   * @param[in]  *pSrcB points to the second input vector
   * @param[out]  *pDst  points to the output vector
   * @param[in]  numSamples number of complex samples in each vector
   * @return none.
   */

  void arm_cmplx_mult_cmplx_q15(
  q15_t * pSrcA,
  q15_t * pSrcB,
  q15_t * pDst,
  uint32_t numSamples);

  /**
   * @brief  Q31 complex-by-complex multiplication
   * @param[in]  *pSrcA points to the first input vector
   * @param[in]  *pSrcB points to the second input vector
   * @param[out]  *pDst  points to the output vector
   * @param[in]  numSamples number of complex samples in each vector
   * @return none.
   */

  void arm_cmplx_mult_cmplx_q31(
  q31_t * pSrcA,
  q31_t * pSrcB,
  q31_t * pDst,
  uint32_t numSamples);

  /**
   * @brief  Floating-point complex-by-complex multiplication
   * @param[in]  *pSrcA points to the first input vector
   * @param[in]  *pSrcB points to the second input vector
   * @param[out]  *pDst  points to the output vector
   * @param[in]  numSamples number of complex samples in each vector
   * @return none.
   */

  void arm_cmplx_mult_cmplx_f32(
  float32_t * pSrcA,
  float32_t * pSrcB,
  float32_t * pDst,
  uint32_t numSamples);

  /**
   * @brief Converts the elements of the floating-point vector to Q31 vector. 
   * @param[in]       *pSrc points to the floating-point input vector 
   * @param[out]      *pDst points to the Q31 output vector
   * @param[in]       blockSize length of the input vector 
   * @return none. 
   */
  void arm_float_to_q31(
  float32_t * pSrc,
  q31_t * pDst,
  uint32_t blockSize);

  /**
   * @brief Converts the elements of the floating-point vector to Q15 vector. 
   * @param[in]       *pSrc points to the floating-point input vector 
   * @param[out]      *pDst points to the Q15 output vector
   * @param[in]       blockSize length of the input vector 
   * @return          none
   */
  void arm_float_to_q15(
  float32_t * pSrc,
  q15_t * pDst,
  uint32_t blockSize);

  /**
   * @brief Converts the elements of the floating-point vector to Q7 vector. 
   * @param[in]       *pSrc points to the floating-point input vector 
   * @param[out]      *pDst points to the Q7 output vector
   * @param[in]       blockSize length of the input vector 
   * @return          none
   */
  void arm_float_to_q7(
  float32_t * pSrc,
  q7_t * pDst,
  uint32_t blockSize);


  /**
   * @brief  Converts the elements of the Q31 vector to Q15 vector.
   * @param[in]  *pSrc is input pointer
   * @param[out]  *pDst is output pointer
   * @param[in]  blockSize is the number of samples to process
   * @return none.
   */
  void arm_q31_to_q15(
  q31_t * pSrc,
  q15_t * pDst,
  uint32_t blockSize);

  /**
   * @brief  Converts the elements of the Q31 vector to Q7 vector.
   * @param[in]  *pSrc is input pointer
   * @param[out]  *pDst is output pointer
   * @param[in]  blockSize is the number of samples to process
   * @return none.
   */
  void arm_q31_to_q7(
  q31_t * pSrc,
  q7_t * pDst,
  uint32_t blockSize);

  /**
   * @brief  Converts the elements of the Q15 vector to floating-point vector.
   * @param[in]  *pSrc is input pointer
   * @param[out]  *pDst is output pointer
   * @param[in]  blockSize is the number of samples to process
   * @return none.
   */
  void arm_q15_to_float(
  q15_t * pSrc,
  float32_t * pDst,
  uint32_t blockSize);


  /**
   * @brief  Converts the elements of the Q15 vector to Q31 vector.
   * @param[in]  *pSrc is input pointer
   * @param[out]  *pDst is output pointer
   * @param[in]  blockSize is the number of samples to process
   * @return none.
   */
  void arm_q15_to_q31(
  q15_t * pSrc,
  q31_t * pDst,
  uint32_t blockSize);


  /**
   * @brief  Converts the elements of the Q15 vector to Q7 vector.
   * @param[in]  *pSrc is input pointer
   * @param[out]  *pDst is output pointer
   * @param[in]  blockSize is the number of samples to process
   * @return none.
   */
  void arm_q15_to_q7(
  q15_t * pSrc,
  q7_t * pDst,
  uint32_t blockSize);


  /**
   * @ingroup groupInterpolation
   */

  /**
   * @defgroup BilinearInterpolate Bilinear Interpolation
   *
   * Bilinear interpolation is an extension of linear interpolation applied to a two dimensional grid.
   * The underlying function <code>f(x, y)</code> is sampled on a regular grid and the interpolation process
   * determines values between the grid points.
   * Bilinear interpolation is equivalent to two step linear interpolation, first in the x-dimension and then in the y-dimension.
   * Bilinear interpolation is often used in image processing to rescale images.
   * The CMSIS DSP library provides bilinear interpolation functions for Q7, Q15, Q31, and floating-point data types.
   *
   * <b>Algorithm</b>
   * \par
   * The instance structure used by the bilinear interpolation functions describes a two dimensional data table.
   * For floating-point, the instance structure is defined as:
   * <pre>
   *   typedef struct
   *   {
   *     uint16_t numRows;
   *     uint16_t numCols;
   *     float32_t *pData;
   * } arm_bilinear_interp_instance_f32;
   * </pre>
   *
   * \par
   * where <code>numRows</code> specifies the number of rows in the table;
   * <code>numCols</code> specifies the number of columns in the table;
   * and <code>pData</code> points to an array of size <code>numRows*numCols</code> values.
   * The data table <code>pTable</code> is organized in row order and the supplied data values fall on integer indexes.
   * That is, table element (x,y) is located at <code>pTable[x + y*numCols]</code> where x and y are integers.
   *
   * \par
   * Let <code>(x, y)</code> specify the desired interpolation point.  Then define:
   * <pre>
   *     XF = floor(x)
   *     YF = floor(y)
   * </pre>
   * \par
   * The interpolated output point is computed as:
   * <pre>
   *  f(x, y) = f(XF, YF) * (1-(x-XF)) * (1-(y-YF))
   *           + f(XF+1, YF) * (x-XF)*(1-(y-YF))
   *           + f(XF, YF+1) * (1-(x-XF))*(y-YF)
   *           + f(XF+1, YF+1) * (x-XF)*(y-YF)
   * </pre>
   * Note that the coordinates (x, y) contain integer and fractional components.  
   * The integer components specify which portion of the table to use while the
   * fractional components control the interpolation processor.
   *
   * \par
   * if (x,y) are outside of the table boundary, Bilinear interpolation returns zero output. 
   */

  /**
   * @addtogroup BilinearInterpolate
   * @{
   */

  /**
  *
  * @brief  Floating-point bilinear interpolation.
  * @param[in,out] *S points to an instance of the interpolation structure.
  * @param[in] X interpolation coordinate.
  * @param[in] Y interpolation coordinate.
  * @return out interpolated value.
  */


  __STATIC_INLINE float32_t arm_bilinear_interp_f32(
  const arm_bilinear_interp_instance_f32 * S,
  float32_t X,
  float32_t Y)
  {
    float32_t out;
    float32_t f00, f01, f10, f11;
    float32_t *pData = S->pData;
    int32_t xIndex, yIndex, index;
    float32_t xdiff, ydiff;
    float32_t b1, b2, b3, b4;

    xIndex = (int32_t) X;
    yIndex = (int32_t) Y;

    /* Care taken for table outside boundary */
    /* Returns zero output when values are outside table boundary */
    if(xIndex < 0 || xIndex > (S->numRows - 1) || yIndex < 0
       || yIndex > (S->numCols - 1))
    {
      return (0);
    }

    /* Calculation of index for two nearest points in X-direction */
    index = (xIndex - 1) + (yIndex - 1) * S->numCols;


    /* Read two nearest points in X-direction */
    f00 = pData[index];
    f01 = pData[index + 1];

    /* Calculation of index for two nearest points in Y-direction */
    index = (xIndex - 1) + (yIndex) * S->numCols;


    /* Read two nearest points in Y-direction */
    f10 = pData[index];
    f11 = pData[index + 1];

    /* Calculation of intermediate values */
    b1 = f00;
    b2 = f01 - f00;
    b3 = f10 - f00;
    b4 = f00 - f01 - f10 + f11;

    /* Calculation of fractional part in X */
    xdiff = X - xIndex;

    /* Calculation of fractional part in Y */
    ydiff = Y - yIndex;

    /* Calculation of bi-linear interpolated output */
    out = b1 + b2 * xdiff + b3 * ydiff + b4 * xdiff * ydiff;

    /* return to application */
    return (out);

  }

  /**
  *
  * @brief  Q31 bilinear interpolation.
  * @param[in,out] *S points to an instance of the interpolation structure.
  * @param[in] X interpolation coordinate in 12.20 format.
  * @param[in] Y interpolation coordinate in 12.20 format.
  * @return out interpolated value.
  */

  __STATIC_INLINE q31_t arm_bilinear_interp_q31(
  arm_bilinear_interp_instance_q31 * S,
  q31_t X,
  q31_t Y)
  {
    q31_t out;                                   /* Temporary output */
    q31_t acc = 0;                               /* output */
    q31_t xfract, yfract;                        /* X, Y fractional parts */
    q31_t x1, x2, y1, y2;                        /* Nearest output values */
    int32_t rI, cI;                              /* Row and column indices */
    q31_t *pYData = S->pData;                    /* pointer to output table values */
    uint32_t nCols = S->numCols;                 /* num of rows */


    /* Input is in 12.20 format */
    /* 12 bits for the table index */
    /* Index value calculation */
    rI = ((X & 0xFFF00000) >> 20u);

    /* Input is in 12.20 format */
    /* 12 bits for the table index */
    /* Index value calculation */
    cI = ((Y & 0xFFF00000) >> 20u);

    /* Care taken for table outside boundary */
    /* Returns zero output when values are outside table boundary */
    if(rI < 0 || rI > (S->numRows - 1) || cI < 0 || cI > (S->numCols - 1))
    {
      return (0);
    }

    /* 20 bits for the fractional part */
    /* shift left xfract by 11 to keep 1.31 format */
    xfract = (X & 0x000FFFFF) << 11u;

    /* Read two nearest output values from the index */
    x1 = pYData[(rI) + nCols * (cI)];
    x2 = pYData[(rI) + nCols * (cI) + 1u];

    /* 20 bits for the fractional part */
    /* shift left yfract by 11 to keep 1.31 format */
    yfract = (Y & 0x000FFFFF) << 11u;

    /* Read two nearest output values from the index */
    y1 = pYData[(rI) + nCols * (cI + 1)];
    y2 = pYData[(rI) + nCols * (cI + 1) + 1u];

    /* Calculation of x1 * (1-xfract ) * (1-yfract) and acc is in 3.29(q29) format */
    out = ((q31_t) (((q63_t) x1 * (0x7FFFFFFF - xfract)) >> 32));
    acc = ((q31_t) (((q63_t) out * (0x7FFFFFFF - yfract)) >> 32));

    /* x2 * (xfract) * (1-yfract)  in 3.29(q29) and adding to acc */
    out = ((q31_t) ((q63_t) x2 * (0x7FFFFFFF - yfract) >> 32));
    acc += ((q31_t) ((q63_t) out * (xfract) >> 32));

    /* y1 * (1 - xfract) * (yfract)  in 3.29(q29) and adding to acc */
    out = ((q31_t) ((q63_t) y1 * (0x7FFFFFFF - xfract) >> 32));
    acc += ((q31_t) ((q63_t) out * (yfract) >> 32));

    /* y2 * (xfract) * (yfract)  in 3.29(q29) and adding to acc */
    out = ((q31_t) ((q63_t) y2 * (xfract) >> 32));
    acc += ((q31_t) ((q63_t) out * (yfract) >> 32));

    /* Convert acc to 1.31(q31) format */
    return (acc << 2u);

  }

  /**
  * @brief  Q15 bilinear interpolation.
  * @param[in,out] *S points to an instance of the interpolation structure.
  * @param[in] X interpolation coordinate in 12.20 format.
  * @param[in] Y interpolation coordinate in 12.20 format.
  * @return out interpolated value.
  */

  __STATIC_INLINE q15_t arm_bilinear_interp_q15(
  arm_bilinear_interp_instance_q15 * S,
  q31_t X,
  q31_t Y)
  {
    q63_t acc = 0;                               /* output */
    q31_t out;                                   /* Temporary output */
    q15_t x1, x2, y1, y2;                        /* Nearest output values */
    q31_t xfract, yfract;                        /* X, Y fractional parts */
    int32_t rI, cI;                              /* Row and column indices */
    q15_t *pYData = S->pData;                    /* pointer to output table values */
    uint32_t nCols = S->numCols;                 /* num of rows */

    /* Input is in 12.20 format */
    /* 12 bits for the table index */
    /* Index value calculation */
    rI = ((X & 0xFFF00000) >> 20);

    /* Input is in 12.20 format */
    /* 12 bits for the table index */
    /* Index value calculation */
    cI = ((Y & 0xFFF00000) >> 20);

    /* Care taken for table outside boundary */
    /* Returns zero output when values are outside table boundary */
    if(rI < 0 || rI > (S->numRows - 1) || cI < 0 || cI > (S->numCols - 1))
    {
      return (0);
    }

    /* 20 bits for the fractional part */
    /* xfract should be in 12.20 format */
    xfract = (X & 0x000FFFFF);

    /* Read two nearest output values from the index */
    x1 = pYData[(rI) + nCols * (cI)];
    x2 = pYData[(rI) + nCols * (cI) + 1u];


    /* 20 bits for the fractional part */
    /* yfract should be in 12.20 format */
    yfract = (Y & 0x000FFFFF);

    /* Read two nearest output values from the index */
    y1 = pYData[(rI) + nCols * (cI + 1)];
    y2 = pYData[(rI) + nCols * (cI + 1) + 1u];

    /* Calculation of x1 * (1-xfract ) * (1-yfract) and acc is in 13.51 format */

    /* x1 is in 1.15(q15), xfract in 12.20 format and out is in 13.35 format */
    /* convert 13.35 to 13.31 by right shifting  and out is in 1.31 */
    out = (q31_t) (((q63_t) x1 * (0xFFFFF - xfract)) >> 4u);
    acc = ((q63_t) out * (0xFFFFF - yfract));

    /* x2 * (xfract) * (1-yfract)  in 1.51 and adding to acc */
    out = (q31_t) (((q63_t) x2 * (0xFFFFF - yfract)) >> 4u);
    acc += ((q63_t) out * (xfract));

    /* y1 * (1 - xfract) * (yfract)  in 1.51 and adding to acc */
    out = (q31_t) (((q63_t) y1 * (0xFFFFF - xfract)) >> 4u);
    acc += ((q63_t) out * (yfract));

    /* y2 * (xfract) * (yfract)  in 1.51 and adding to acc */
    out = (q31_t) (((q63_t) y2 * (xfract)) >> 4u);
    acc += ((q63_t) out * (yfract));

    /* acc is in 13.51 format and down shift acc by 36 times */
    /* Convert out to 1.15 format */
    return (acc >> 36);

  }

  /**
  * @brief  Q7 bilinear interpolation.
  * @param[in,out] *S points to an instance of the interpolation structure.
  * @param[in] X interpolation coordinate in 12.20 format.
  * @param[in] Y interpolation coordinate in 12.20 format.
  * @return out interpolated value.
  */

  __STATIC_INLINE q7_t arm_bilinear_interp_q7(
  arm_bilinear_interp_instance_q7 * S,
  q31_t X,
  q31_t Y)
  {
    q63_t acc = 0;                               /* output */
    q31_t out;                                   /* Temporary output */
    q31_t xfract, yfract;                        /* X, Y fractional parts */
    q7_t x1, x2, y1, y2;                         /* Nearest output values */
    int32_t rI, cI;                              /* Row and column indices */
    q7_t *pYData = S->pData;                     /* pointer to output table values */
    uint32_t nCols = S->numCols;                 /* num of rows */

    /* Input is in 12.20 format */
    /* 12 bits for the table index */
    /* Index value calculation */
    rI = ((X & 0xFFF00000) >> 20);

    /* Input is in 12.20 format */
    /* 12 bits for the table index */
    /* Index value calculation */
    cI = ((Y & 0xFFF00000) >> 20);

    /* Care taken for table outside boundary */
    /* Returns zero output when values are outside table boundary */
    if(rI < 0 || rI > (S->numRows - 1) || cI < 0 || cI > (S->numCols - 1))
    {
      return (0);
    }

    /* 20 bits for the fractional part */
    /* xfract should be in 12.20 format */
    xfract = (X & 0x000FFFFF);

    /* Read two nearest output values from the index */
    x1 = pYData[(rI) + nCols * (cI)];
    x2 = pYData[(rI) + nCols * (cI) + 1u];


    /* 20 bits for the fractional part */
    /* yfract should be in 12.20 format */
    yfract = (Y & 0x000FFFFF);

    /* Read two nearest output values from the index */
    y1 = pYData[(rI) + nCols * (cI + 1)];
    y2 = pYData[(rI) + nCols * (cI + 1) + 1u];

    /* Calculation of x1 * (1-xfract ) * (1-yfract) and acc is in 16.47 format */
    out = ((x1 * (0xFFFFF - xfract)));
    acc = (((q63_t) out * (0xFFFFF - yfract)));

    /* x2 * (xfract) * (1-yfract)  in 2.22 and adding to acc */
    out = ((x2 * (0xFFFFF - yfract)));
    acc += (((q63_t) out * (xfract)));

    /* y1 * (1 - xfract) * (yfract)  in 2.22 and adding to acc */
    out = ((y1 * (0xFFFFF - xfract)));
    acc += (((q63_t) out * (yfract)));

    /* y2 * (xfract) * (yfract)  in 2.22 and adding to acc */
    out = ((y2 * (yfract)));
    acc += (((q63_t) out * (xfract)));

    /* acc in 16.47 format and down shift by 40 to convert to 1.7 format */
    return (acc >> 40);

  }

  /**
   * @} end of BilinearInterpolate group
   */






#ifdef  __cplusplus
}
#endif


#endif /* _ARM_MATH_H */


/**
 *
 * End of file.
 */