mirror of
https://github.com/python-pillow/Pillow.git
synced 2024-11-17 07:07:10 +03:00
107 lines
3.4 KiB
C
107 lines
3.4 KiB
C
|
#include "Imaging.h"
|
|||
|
#include <math.h>
|
|||
|
|
|||
|
|
|||
|
#define ROUND_UP(f) ((int) ((f) >= 0.0 ? (f) + 0.5F : (f) - 0.5F))
|
|||
|
|
|||
|
|
|||
|
/* 8 bits for result. Table can overflow [0, 1.0] range,
|
|||
|
so we need extra bits for overflow and negative values. */
|
|||
|
#define PRECISION_BITS (16 - 8 - 2)
|
|||
|
|
|||
|
|
|||
|
static inline void
|
|||
|
interpolate3(INT16 out[3], const INT16 a[3], const INT16 b[3], float shift)
|
|||
|
{
|
|||
|
out[0] = a[0] * (1-shift) + b[0] * shift;
|
|||
|
out[1] = a[1] * (1-shift) + b[1] * shift;
|
|||
|
out[2] = a[2] * (1-shift) + b[2] * shift;
|
|||
|
}
|
|||
|
|
|||
|
static inline void
|
|||
|
interpolate4(INT16 out[3], const INT16 a[3], const INT16 b[3], float shift)
|
|||
|
{
|
|||
|
out[0] = a[0] * (1-shift) + b[0] * shift;
|
|||
|
out[1] = a[1] * (1-shift) + b[1] * shift;
|
|||
|
out[2] = a[2] * (1-shift) + b[2] * shift;
|
|||
|
out[3] = a[3] * (1-shift) + b[3] * shift;
|
|||
|
}
|
|||
|
|
|||
|
static inline int
|
|||
|
table3D_index3(int index1D, int index2D, int index3D,
|
|||
|
int size1D, int size1D_2D)
|
|||
|
{
|
|||
|
return (index1D + index2D * size1D + index3D * size1D_2D) * 3;
|
|||
|
}
|
|||
|
|
|||
|
static inline int
|
|||
|
table3D_index4(int index1D, int index2D, int index3D,
|
|||
|
int size1D, int size1D_2D)
|
|||
|
{
|
|||
|
return (index1D + index2D * size1D + index3D * size1D_2D) * 4;
|
|||
|
}
|
|||
|
|
|||
|
/*
|
|||
|
Transforms colors of imIn using provided 3D look-up table
|
|||
|
and puts the result in imOut. Returns imOut on sucess or 0 on error.
|
|||
|
|
|||
|
imOut, imIn — images, should be the same size and may be the same image.
|
|||
|
Should have 3 or 4 channels.
|
|||
|
table_channels — number of channels in the look-up table, 3 or 4.
|
|||
|
Should be less or equal than number of channels in imOut image;
|
|||
|
size1D, size_2D and size3D — dimensions of provided table;
|
|||
|
table — flatten table,
|
|||
|
array with table_channels × size1D × size2D × size3D elements,
|
|||
|
where channels are changed first, then 1D, then 2D, then 3D.
|
|||
|
Each element is signed 16-bit int where 0 is lowest output value
|
|||
|
and 255 << PRECISION_BITS (16320) is highest value.
|
|||
|
*/
|
|||
|
Imaging
|
|||
|
ImagingColorLUT3D_linear(Imaging imOut, Imaging imIn, int table_channels,
|
|||
|
int size1D, int size2D, int size3D,
|
|||
|
INT16* table)
|
|||
|
{
|
|||
|
int size1D_2D = size1D * size2D;
|
|||
|
float scale1D = (size1D - 1) / 255.0;
|
|||
|
float scale2D = (size2D - 1) / 255.0;
|
|||
|
float scale3D = (size3D - 1) / 255.0;
|
|||
|
int x, y;
|
|||
|
|
|||
|
if (table_channels < 3 || table_channels > 4) {
|
|||
|
PyErr_SetString(PyExc_ValueError, "table_channels could be 3 or 4");
|
|||
|
return NULL;
|
|||
|
}
|
|||
|
|
|||
|
if (imIn->type != IMAGING_TYPE_UINT8 ||
|
|||
|
imOut->type != IMAGING_TYPE_UINT8 ||
|
|||
|
imIn->bands < 3 ||
|
|||
|
imOut->bands < table_channels
|
|||
|
) {
|
|||
|
return (Imaging) ImagingError_ModeError();
|
|||
|
}
|
|||
|
|
|||
|
/* In case we have one extra band in imOut and don't have in imIn.*/
|
|||
|
if (imOut->bands > table_channels && imOut->bands > imIn->bands) {
|
|||
|
return (Imaging) ImagingError_ModeError();
|
|||
|
}
|
|||
|
|
|||
|
for (y = 0; y < imOut->ysize; y++) {
|
|||
|
UINT8 *rowIn = (UINT8 *)imIn->image[y];
|
|||
|
UINT8 *rowOut = (UINT8 *)imIn->image[y];
|
|||
|
for (x = 0; x < imOut->xsize; x++) {
|
|||
|
float scaled1D = rowIn[x*4 + 0] * scale1D;
|
|||
|
float scaled2D = rowIn[x*4 + 1] * scale2D;
|
|||
|
float scaled3D = rowIn[x*4 + 2] * scale3D;
|
|||
|
int index1D = (int) scaled1D;
|
|||
|
int index2D = (int) scaled2D;
|
|||
|
int index3D = (int) scaled3D;
|
|||
|
float shift1D = scaled1D - index1D;
|
|||
|
float shift2D = scaled2D - index2D;
|
|||
|
float shift3D = scaled3D - index3D;
|
|||
|
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
return imOut;
|
|||
|
}
|