/* ----------------------------------------------------------------------------
Function : PyOptMed9()
In : pointer to an array of 9 pixel values
Out : a pixel value
Job : optimized search of the median of 9 pixel values
Notice : in theory, cannot go faster without assumptions on the
signal.
Formula from:
XILINX XCELL magazine, vol. 23 by John L. Smith
The input array is modified in the process
The result array is guaranteed to contain the median
value in middle position, but other elements are NOT sorted.
Code adapted from Nicolas Devillard.
--------------------------------------------------------------------------- */
float
PyOptMed9(float* p)
{
PyDoc_STRVAR(PyOptMed9__doc__, "PyOptMed9(a) -> float\n\n"
"Get the median of array a of length 9 using a search tree.");
PIX_SORT(p[1], p[2]);
PIX_SORT(p[4], p[5]);
PIX_SORT(p[7], p[8]);
PIX_SORT(p[0], p[1]);
PIX_SORT(p[3], p[4]);
PIX_SORT(p[6], p[7]);
PIX_SORT(p[1], p[2]);
PIX_SORT(p[4], p[5]);
PIX_SORT(p[7], p[8]);
PIX_SORT(p[0], p[3]);
PIX_SORT(p[5], p[8]);
PIX_SORT(p[4], p[7]);
PIX_SORT(p[3], p[6]);
PIX_SORT(p[1], p[4]);
PIX_SORT(p[2], p[5]);
PIX_SORT(p[4], p[7]);
PIX_SORT(p[4], p[2]);
PIX_SORT(p[6], p[4]);
PIX_SORT(p[4], p[2]);
return p[4];
}
/* ----------------------------------------------------------------------------
Function : PyOptMed3()
In : pointer to array of 3 pixel values
Out : a pixel value
Job : optimized search of the median of 3 pixel values
Notice : found on sci.image.processing
cannot go faster unless assumptions are made on the nature of the input
signal.
Code adapted from Nicolas Devillard.
--------------------------------------------------------------------------- */
float
PyOptMed3(float* p)
{
PyDoc_STRVAR(PyOptMed3__doc__, "PyOptMed3(a) -> float\n\n"
"Get the median of array a of length 3 using a search tree.");
PIX_SORT(p[0], p[1]);
PIX_SORT(p[1], p[2]);
PIX_SORT(p[0], p[1]);
return p[1];
}
/* Calculate the 7x7 separable median filter of an array data that has
* dimensions nx x ny. The results are saved in the output array. The output
* array should already be allocated as we work on it in place. The median
* filter is not calculated for a 3 pixel border around the image. These pixel
* values are copied from the input data. The data should be striped along the
* x direction, such that pixel i,j in the 2D image should have memory location
* data[i + nx *j]. Note that the rows are median filtered first, followed by
* the columns.
*/
void
PySepMedFilt7(float* data, float* output, int nx, int ny)
{
PyDoc_STRVAR(PySepMedFilt7__doc__,
"PySepMedFilt7(data, output, nx, ny) -> void\n\n"
"Calculate the 7x7 separable median filter on an array data with "
"dimensions nx x ny. The results are saved in the output array "
"which should already be allocated as we work on it in place. The "
"median filter is not calculated for a 3 pixel border which is "
"copied from the input data. The data array should be striped in "
"the x direction such that pixel i,j has memory location "
"data[i + nx * j]. Note that the rows are median filtered first, "
"followed by the columns.");
/* Total number of pixels */
int nxny = nx * ny;
/* Output array for the median filter of the rows. We later median filter
* the columns of this array. */
float* rowmed = (float *) malloc(nxny * sizeof(float));
/* Loop indices */
int i, j, nxj;
/* 7 element array to calculate the median and a counter index. Note that
* this array needs to be unique for each thread so it needs to be
* private and we wait to allocate memory until the pragma below. */
float* medarr;
/* Median filter the rows first */
/* Each thread needs to access the data and rowmed so we make them
* firstprivate. We make sure that our algorithm doesn't have multiple
* threads read or write the same piece of memory. */
#pragma omp parallel firstprivate(data, rowmed, nx, ny) \
private(i, j, nxj, medarr)
{
/*Each thread allocates its own array. */
medarr = (float *) malloc(7 * sizeof(float));
/* For each pixel excluding the border */
#pragma omp for nowait
for (j = 0; j < ny; j++) {
nxj = nx * j;
for (i = 3; i < nx - 3; i++) {
medarr[0] = data[nxj + i];
medarr[1] = data[nxj + i - 1];
medarr[2] = data[nxj + i + 1];
medarr[3] = data[nxj + i - 2];
medarr[4] = data[nxj + i + 2];
medarr[5] = data[nxj + i - 3];
medarr[6] = data[nxj + i + 3];
/* Calculate the median in the fastest way possible */
rowmed[nxj + i] = PyOptMed7(medarr);
}
}
/* Each thread needs to free its own medarr */
free(medarr);
}
/* Fill in the borders of rowmed with the original data values */
#pragma omp parallel for firstprivate(rowmed, data, nx, ny) private(j, nxj)
for (j = 0; j < ny; j++) {
nxj = nx * j;
rowmed[nxj] = data[nxj];
rowmed[nxj + 1] = data[nxj + 1];
rowmed[nxj + 2] = data[nxj + 2];
rowmed[nxj + nx - 1] = data[nxj + nx - 1];
rowmed[nxj + nx - 2] = data[nxj + nx - 2];
rowmed[nxj + nx - 3] = data[nxj + nx - 3];
}
/* Median filter the columns */
#pragma omp parallel firstprivate(rowmed, output, nx, ny) \
private(i, j, nxj, medarr)
{
/* Each thread needs to reallocate a new medarr */
medarr = (float *) malloc(7 * sizeof(float));
/* For each pixel excluding the border */
#pragma omp for nowait
for (j = 3; j < ny - 3; j++) {
nxj = nx * j;
for (i = 3; i < nx - 3; i++) {
medarr[0] = rowmed[i + nxj - nx];
medarr[1] = rowmed[i + nxj + nx];
medarr[2] = rowmed[i + nxj + nx + nx];
medarr[3] = rowmed[i + nxj - nx - nx];
medarr[4] = rowmed[i + nxj];
medarr[5] = rowmed[i + nxj + nx + nx + nx];
medarr[6] = rowmed[i + nxj - nx - nx - nx];
/* Calculate the median in the fastest way possible */
output[nxj + i] = PyOptMed7(medarr);
}
}
/* Each thread needs to free its own medarr */
free(medarr);
}
/* Clean up rowmed */
free(rowmed);
/* Copy the border pixels from the original data into the output array */
#pragma omp parallel for firstprivate(output, data, nx, nxny) private(i)
for (i = 0; i < nx; i++) {
output[i] = data[i];
output[i + nx] = data[i + nx];
output[i + nx + nx] = data[i + nx + nx];
output[nxny - nx + i] = data[nxny - nx + i];
output[nxny - nx - nx + i] = data[nxny - nx - nx + i];
output[nxny - nx - nx - nx + i] = data[nxny - nx - nx - nx + i];
}
#pragma omp parallel for firstprivate(output, data, nx, ny) private(j, nxj)
for (j = 0; j < ny; j++) {
nxj = nx * j;
output[nxj] = data[nxj];
output[nxj + 1] = data[nxj + 1];
output[nxj + 2] = data[nxj + 2];
output[nxj + nx - 1] = data[nxj + nx - 1];
output[nxj + nx - 2] = data[nxj + nx - 2];
output[nxj + nx - 3] = data[nxj + nx - 3];
}
return;
}
/* Calculate the 7x7 median filter of an array data that has dimensions
* nx x ny. The results are saved in the output array. The output array should
* already be allocated as we work on it in place. The median filter is not
* calculated for a 3 pixel border around the image. These pixel values are
* copied from the input data. The data should be striped along the
* x direction, such that pixel i,j in the 2D image should have memory
* location data[i + nx *j].
*/
void
PyMedFilt7(float* data, float* output, int nx, int ny)
{
PyDoc_STRVAR(PyMedFilt7__doc__,
"PyMedFilt7(data, output, nx, ny) -> void\n\n"
"Calculate the 7x7 median filter on an array data with dimensions "
"nx x ny. The results are saved in the output array. The output "
"array should already be allocated as we work on it in place. The "
"median filter is not calculated for a 3 pixel border around the "
"image. These pixel values are copied from the input data. Note "
"that the data array needs to be striped in the x direction such "
"that pixel i,j has memory location data[i + nx * j]");
/*Total size of the array */
int nxny = nx * ny;
/* Loop indices */
int i, j, nxj;
int k, l, nxk;
/* 49 element array to calculate the median and a counter index. Note that
* these both need to be unique for each thread so they both need to be
* private and we wait to allocate memory until the pragma below. */
float* medarr;
int medcounter;
/* Each thread needs to access the data and the output so we make them
* firstprivate. We make sure that our algorithm doesn't have multiple
* threads read or write the same piece of memory. */
#pragma omp parallel firstprivate(output, data, nx, ny) \
private(i, j, k, l, medarr, nxj, nxk, medcounter)
{
/*Each thread allocates its own array. */
medarr = (float *) malloc(49 * sizeof(float));
/* Go through each pixel excluding the border.*/
#pragma omp for nowait
for (j = 3; j < ny - 3; j++) {
/* Precalculate the multiplication nx * j, minor optimization */
nxj = nx * j;
for (i = 3; i < nx - 3; i++) {
medcounter = 0;
/* The compiler should optimize away these loops */
for (k = -3; k < 4; k++) {
nxk = nx * k;
for (l = -3; l < 4; l++) {
medarr[medcounter] = data[nxj + i + nxk + l];
medcounter++;
}
}
/* Calculate the median in the fastest way possible */
output[nxj + i] = PyMedian(medarr, 49);
}
}
/* Each thread needs to free its own copy of medarr */
free(medarr);
}
#pragma omp parallel firstprivate(output, data, nx, nxny) private(i)
/* Copy the border pixels from the original data into the output array */
for (i = 0; i < nx; i++) {
output[i] = data[i];
output[i + nx] = data[i + nx];
output[i + nx + nx] = data[i + nx + nx];
output[nxny - nx + i] = data[nxny - nx + i];
output[nxny - nx - nx + i] = data[nxny - nx - nx + i];
output[nxny - nx - nx - nx + i] = data[nxny - nx - nx - nx + i];
}
#pragma omp parallel firstprivate(output, data, nx, ny) private(j, nxj)
for (j = 0; j < ny; j++) {
nxj = nx * j;
output[nxj] = data[nxj];
output[nxj + 1] = data[nxj + 1];
output[nxj + 2] = data[nxj + 2];
output[nxj + nx - 1] = data[nxj + nx - 1];
output[nxj + nx - 2] = data[nxj + nx - 2];
output[nxj + nx - 3] = data[nxj + nx - 3];
}
return;
}
float
PyMedian(float* a, int n)
{
/* Get the median of an array "a" with length "n"
* using the Quickselect algorithm. Returns a float.
* This Quickselect routine is based on the algorithm described in
* "Numerical recipes in C", Second Edition, Cambridge University Press,
* 1992, Section 8.5, ISBN 0-521-43108-5
* This code by Nicolas Devillard - 1998. Public domain.
*/
PyDoc_STRVAR(PyMedian__doc__, "PyMedian(a, n) -> float\n\n"
"Get the median of array a of length n using the Quickselect "
"algorithm.");
/* Make a copy of the array so that we don't alter the input array */
float* arr = (float *) malloc(n * sizeof(float));
/* Indices of median, low, and high values we are considering */
int low = 0;
int high = n - 1;
int median = (low + high) / 2;
/* Running indices for the quick select algorithm */
int middle, ll, hh;
/* The median to return */
float med;
/* running index i */
int i;
/* Copy the input data into the array we work with */
for (i = 0; i < n; i++) {
arr[i] = a[i];
}
/* Start an infinite loop */
while (true) {
/* Only One or two elements left */
if (high <= low + 1) {
/* Check if we need to swap the two elements */
if ((high == low + 1) && (arr[low] > arr[high]))
ELEM_SWAP(arr[low], arr[high]);
med = arr[median];
free(arr);
return med;
}
/* Find median of low, middle and high items;
* swap into position low */
middle = (low + high) / 2;
if (arr[middle] > arr[high])
ELEM_SWAP(arr[middle], arr[high]);
if (arr[low] > arr[high])
ELEM_SWAP(arr[low], arr[high]);
if (arr[middle] > arr[low])
ELEM_SWAP(arr[middle], arr[low]);
/* Swap low item (now in position middle) into position (low+1) */
ELEM_SWAP(arr[middle], arr[low + 1]);
/* Nibble from each end towards middle,
* swap items when stuck */
ll = low + 1;
hh = high;
while (true) {
do
ll++;
while (arr[low] > arr[ll]);
do
hh--;
while (arr[hh] > arr[low]);
if (hh < ll)
break;
ELEM_SWAP(arr[ll], arr[hh]);
}
/* Swap middle item (in position low) back into
* the correct position */
ELEM_SWAP(arr[low], arr[hh]);
/* Re-set active partition */
if (hh <= median)
low = ll;
if (hh >= median)
high = hh - 1;
}
}
/* ----------------------------------------------------------------------------
Function : PyOptMed25()
In : pointer to an array of 25 pixel values
Out : a pixel value
Job : optimized search of the median of 25 pixel values
Notice : in theory, cannot go faster without assumptions on the
signal.
Code taken from Graphic Gems.
Code adapted from Nicolas Devillard.
--------------------------------------------------------------------------- */
float
PyOptMed25(float* p)
{
PyDoc_STRVAR(PyOptMed25__doc__, "PyOptMed25(a) -> float\n\n"
"Get the median of array a of length 25 using a search tree.");
PIX_SORT(p[0], p[1]);
PIX_SORT(p[3], p[4]);
PIX_SORT(p[2], p[4]);
PIX_SORT(p[2], p[3]);
PIX_SORT(p[6], p[7]);
PIX_SORT(p[5], p[7]);
PIX_SORT(p[5], p[6]);
PIX_SORT(p[9], p[10]);
PIX_SORT(p[8], p[10]);
PIX_SORT(p[8], p[9]);
PIX_SORT(p[12], p[13]);
PIX_SORT(p[11], p[13]);
PIX_SORT(p[11], p[12]);
PIX_SORT(p[15], p[16]);
PIX_SORT(p[14], p[16]);
PIX_SORT(p[14], p[15]);
PIX_SORT(p[18], p[19]);
PIX_SORT(p[17], p[19]);
PIX_SORT(p[17], p[18]);
PIX_SORT(p[21], p[22]);
PIX_SORT(p[20], p[22]);
PIX_SORT(p[20], p[21]);
PIX_SORT(p[23], p[24]);
PIX_SORT(p[2], p[5]);
PIX_SORT(p[3], p[6]);
PIX_SORT(p[0], p[6]);
PIX_SORT(p[0], p[3]);
PIX_SORT(p[4], p[7]);
PIX_SORT(p[1], p[7]);
PIX_SORT(p[1], p[4]);
PIX_SORT(p[11], p[14]);
PIX_SORT(p[8], p[14]);
PIX_SORT(p[8], p[11]);
PIX_SORT(p[12], p[15]);
PIX_SORT(p[9], p[15]);
PIX_SORT(p[9], p[12]);
PIX_SORT(p[13], p[16]);
PIX_SORT(p[10], p[16]);
PIX_SORT(p[10], p[13]);
PIX_SORT(p[20], p[23]);
PIX_SORT(p[17], p[23]);
PIX_SORT(p[17], p[20]);
PIX_SORT(p[21], p[24]);
PIX_SORT(p[18], p[24]);
PIX_SORT(p[18], p[21]);
PIX_SORT(p[19], p[22]);
PIX_SORT(p[8], p[17]);
PIX_SORT(p[9], p[18]);
PIX_SORT(p[0], p[18]);
PIX_SORT(p[0], p[9]);
PIX_SORT(p[10], p[19]);
PIX_SORT(p[1], p[19]);
PIX_SORT(p[1], p[10]);
PIX_SORT(p[11], p[20]);
PIX_SORT(p[2], p[20]);
PIX_SORT(p[2], p[11]);
PIX_SORT(p[12], p[21]);
PIX_SORT(p[3], p[21]);
PIX_SORT(p[3], p[12]);
PIX_SORT(p[13], p[22]);
PIX_SORT(p[4], p[22]);
PIX_SORT(p[4], p[13]);
PIX_SORT(p[14], p[23]);
PIX_SORT(p[5], p[23]);
PIX_SORT(p[5], p[14]);
PIX_SORT(p[15], p[24]);
PIX_SORT(p[6], p[24]);
PIX_SORT(p[6], p[15]);
PIX_SORT(p[7], p[16]);
PIX_SORT(p[7], p[19]);
PIX_SORT(p[13], p[21]);
PIX_SORT(p[15], p[23]);
PIX_SORT(p[7], p[13]);
PIX_SORT(p[7], p[15]);
PIX_SORT(p[1], p[9]);
PIX_SORT(p[3], p[11]);
PIX_SORT(p[5], p[17]);
PIX_SORT(p[11], p[17]);
PIX_SORT(p[9], p[17]);
PIX_SORT(p[4], p[10]);
PIX_SORT(p[6], p[12]);
PIX_SORT(p[7], p[14]);
PIX_SORT(p[4], p[6]);
PIX_SORT(p[4], p[7]);
PIX_SORT(p[12], p[14]);
PIX_SORT(p[10], p[14]);
PIX_SORT(p[6], p[7]);
PIX_SORT(p[10], p[12]);
PIX_SORT(p[6], p[10]);
PIX_SORT(p[6], p[17]);
PIX_SORT(p[12], p[17]);
PIX_SORT(p[7], p[17]);
PIX_SORT(p[7], p[10]);
PIX_SORT(p[12], p[18]);
PIX_SORT(p[7], p[12]);
PIX_SORT(p[10], p[18]);
PIX_SORT(p[12], p[20]);
PIX_SORT(p[10], p[20]);
PIX_SORT(p[10], p[12]);
return p[12];
}
void initpp_packing(void)
{
/* The module doc string */
PyDoc_STRVAR(pp_packing__doc__,
"This extension module provides access to the underlying libmo_unpack library functionality.\n"
""
);
PyDoc_STRVAR(wgdos_unpack__doc__,
"Unpack PP field data that has been packed using WGDOS archive method.\n"
"\n"
"Provides access to the libmo_unpack library function Wgdos_Unpack.\n"
"\n"
"Args:\n\n"
"* data (numpy.ndarray):\n"
" The raw field byte array to be unpacked.\n"
"* lbrow (int):\n"
" The number of rows in the grid.\n"
"* lbnpt (int):\n"
" The number of points (columns) per row in the grid.\n"
"* bmdi (float):\n"
" The value used in the field to indicate missing data points.\n"
"\n"
"Returns:\n"
" numpy.ndarray, 2d array containing normal unpacked field data.\n"
""
);
PyDoc_STRVAR(rle_decode__doc__,
"Uncompress PP field data that has been compressed using Run Length Encoding.\n"
"\n"
"Provides access to the libmo_unpack library function runlenDecode.\n"
"Decodes the field by expanding out the missing data points represented\n"
"by a single missing data value followed by a value indicating the length\n"
"of the run of missing data values.\n"
"\n"
"Args:\n\n"
"* data (numpy.ndarray):\n"
" The raw field byte array to be uncompressed.\n"
"* lbrow (int):\n"
" The number of rows in the grid.\n"
"* lbnpt (int):\n"
" The number of points (columns) per row in the grid.\n"
"* bmdi (float):\n"
" The value used in the field to indicate missing data points.\n"
"\n"
"Returns:\n"
" numpy.ndarray, 2d array containing normal uncompressed field data.\n"
""
);
/* ==== Set up the module's methods table ====================== */
static PyMethodDef pp_packingMethods[] = {
{"wgdos_unpack", wgdos_unpack_py, METH_VARARGS, wgdos_unpack__doc__},
{"rle_decode", rle_decode_py, METH_VARARGS, rle_decode__doc__},
{NULL, NULL, 0, NULL} /* marks the end of this structure */
};
Py_InitModule3("pp_packing", pp_packingMethods, pp_packing__doc__);
import_array(); // Must be present for NumPy.
}