Skip to content

Correction images tiff jpeg YCbCr #51

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Dolite merged 1 commit into from
Jul 23, 2025
Merged

Correction images tiff jpeg YCbCr #51

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
73 changes: 37 additions & 36 deletions src/image/file/LibtiffImage.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -190,15 +190,15 @@ LibtiffImage* LibtiffImage::create_to_read ( std::string filename, BoundingBox<
int width=0, height=0, channels=0, planarconfig=0, bitspersample=0, sf=0, ph=0, comp=0, rowsperstrip=0;
bool tiled = false, palette = false;
TIFF* tif = TIFFOpen ( filename.c_str(), "r" );


/************** RECUPERATION DES INFORMATIONS **************/

if ( tif == NULL ) {
BOOST_LOG_TRIVIAL(error) << "Unable to open TIFF (to read) " << filename ;
return NULL;
}

if ( TIFFGetField ( tif, TIFFTAG_IMAGEWIDTH, &width ) < 1 ) {
BOOST_LOG_TRIVIAL(error) << "Unable to read pixel width for file " << filename ;
TIFFClose ( tif );
Expand Down Expand Up @@ -256,8 +256,8 @@ LibtiffImage* LibtiffImage::create_to_read ( std::string filename, BoundingBox<
TIFFClose ( tif );
return NULL;
}
if (to_rok4_photometric ( ph ) == Photometric::PALETTE) {

if (to_rok4_photometric ( ph ) == Photometric::PALETTE) {
palette = true;
}

Expand Down Expand Up @@ -294,7 +294,7 @@ LibtiffImage* LibtiffImage::create_to_read ( std::string filename, BoundingBox<
TIFFClose ( tif );
return NULL;
}

/********************** CONTROLES **************************/

if ( to_rok4_sampleformat ( sf, bitspersample ) == SampleFormat::UNKNOWN ) {
Expand All @@ -315,7 +315,7 @@ LibtiffImage* LibtiffImage::create_to_read ( std::string filename, BoundingBox<
resx = 1.;
resy = 1.;
}

/******************** CRÉATION DE L'OBJET ******************/

return new LibtiffImage (
Expand Down Expand Up @@ -343,12 +343,12 @@ LibtiffImage* LibtiffImage::create_to_write (
BOOST_LOG_TRIVIAL(error) << "One dimension is not valid for the output image " << filename << " : " << width << ", " << height ;
return NULL;
}

if ( channels <= 0 ) {
BOOST_LOG_TRIVIAL(error) << "Number of samples per pixel is not valid for the output image " << filename << " : " << channels ;
return NULL;
}

if ( resx > 0 && resy > 0 ) {
// Vérification de la cohérence entre les résolutions et bbox fournies et les dimensions (en pixel) de l'image
// Arrondi a la valeur entiere la plus proche
Expand Down Expand Up @@ -470,9 +470,9 @@ LibtiffImage::LibtiffImage (
),

tif ( tif ), rowsperstrip ( rowsperstrip ), tiled (tiled), palette (palette) {

// Ce constructeur permet de déterminer si la conversion de 1 à 8 bits est nécessaire, et de savoir si 0 est blanc ou noir

if ( bps == 1 ) {
// On fera la conversion en entiers sur 8 bits à la volée.
// Cette image sera comme une image sur 8 bits.
Expand Down Expand Up @@ -501,7 +501,7 @@ LibtiffImage::LibtiffImage (
current_strip = -1;
int stripSize = width*rowsperstrip*pixel_size;
strip_buffer = new uint8_t[stripSize];

if (bit_to_byte) {
// On a besoin d'un buffer supplémentaire pour faire la conversion à la volée à la lecture
bit_to_byte_buffer = new uint8_t[stripSize];
Expand All @@ -516,7 +516,7 @@ LibtiffImage::LibtiffImage (
FileImage ( width, height, resx, resy, channels, bbox, name, sampleformat, photometric, compression, extra_sample_type ),

tif ( tif ), rowsperstrip ( rowsperstrip ) {

bit_to_byte = 0;

current_strip = -1;
Expand All @@ -530,9 +530,9 @@ LibtiffImage::LibtiffImage (
template<typename T>
int LibtiffImage::_getline ( T* buffer, int line ) {
// buffer doit déjà être alloué, et assez grand, en tenant compte de la conversion

if ( line / rowsperstrip != current_strip ) {

// Les données n'ont pas encore été lue depuis l'image (strip pas en mémoire).
current_strip = line / rowsperstrip;

Expand All @@ -559,7 +559,8 @@ int LibtiffImage::_getline ( T* buffer, int line ) {

for (int t = 0; t < tilenumber_widthwise; t++) {

if (palette) {
// Si l'image a une photometrie Palette ou bien YCBCR --> TIFFReadRGBATile
if (palette || photometric == Photometric::YCBCR) {
uint32* palette_buffer = new uint32[width * rowsperstrip];

size = TIFFReadRGBATile(tif, t * tile_width, current_strip * tile_height, palette_buffer );
Expand Down Expand Up @@ -604,7 +605,8 @@ int LibtiffImage::_getline ( T* buffer, int line ) {

_TIFFfree(tile_buf);
} else {
if (palette) {
// Si l'image a une photometrie Palette ou bien YCBCR --> TIFFReadRGBAStrip
if (palette || photometric == Photometric::YCBCR) {
uint32* palette_buffer = new uint32[width * rowsperstrip];
size = TIFFReadRGBAStrip ( tif, line, palette_buffer);
if ( size == 0 ) {
Expand All @@ -625,7 +627,7 @@ int LibtiffImage::_getline ( T* buffer, int line ) {
if (line / rowsperstrip == (height - 1) / rowsperstrip) {
rows_count = height % rowsperstrip;
}

for (int l = 0; l < rows_count; l++) {
for (int i = 0; i < width; i++) {
memcpy ( strip_buffer + (width * l + i) * pixel_size, palette_buffer + (rows_count - 1 - l) * width + i, 3 );
Expand All @@ -641,19 +643,19 @@ int LibtiffImage::_getline ( T* buffer, int line ) {
}
}
}

if (bit_to_byte == 1) {
OneBitConverter::minwhiteToGray(bit_to_byte_buffer, strip_buffer, size);
} else if (bit_to_byte == 2) {
OneBitConverter::minblackToGray(bit_to_byte_buffer, strip_buffer, size);
}
}


T buffertmp[width * channels];

/************* SI CONVERSION 1 bit -> 8 bits **************/

if (bit_to_byte) {
memcpy ( buffertmp, bit_to_byte_buffer + ( line%rowsperstrip ) * width * pixel_size, width * pixel_size );
} else {
Expand All @@ -672,7 +674,7 @@ int LibtiffImage::_getline ( T* buffer, int line ) {
memcpy(buffer, buffertmp, pixel_size * width);
}


return width * get_channels();
}

Expand All @@ -698,7 +700,7 @@ int LibtiffImage::get_line ( uint8_t* buffer, int line ) {
}

int LibtiffImage::get_line ( uint16_t* buffer, int line ) {

if ( sample_format == SampleFormat::UINT8 ) {
// On veut la ligne en entier 16 bits mais l'image lue est sur 8 bits : on convertit
uint8_t* buffer_t = new uint8_t[width * get_channels()];
Expand All @@ -707,7 +709,7 @@ int LibtiffImage::get_line ( uint16_t* buffer, int line ) {
delete [] buffer_t;
return width * get_channels();
} else if ( sample_format == SampleFormat::UINT16 ) { // uint16
return _getline ( buffer,line );
return _getline ( buffer,line );
} else if ( sample_format == SampleFormat::FLOAT32 ) { // float
/* On ne convertit pas les nombres flottants en entier sur 16 bits (aucun intérêt)
* On va copier le buffer flottant sur le buffer entier 16 bits, de même taille en octet (2 fois plus grand en "nombre de cases")*/
Expand All @@ -733,7 +735,7 @@ int LibtiffImage::get_line ( float* buffer, int line ) {
_getline ( buffer_t,line );
convert ( buffer, buffer_t, width * get_channels() );
delete [] buffer_t;
return width * get_channels();
return width * get_channels();
} else if ( sample_format == SampleFormat::FLOAT32 ) { // float
return _getline ( buffer, line );
}
Expand Down Expand Up @@ -798,7 +800,7 @@ int LibtiffImage::write_image ( Image* pIn ) {
}

int LibtiffImage::write_image ( uint8_t* buffer) {

// Si l'image à écrire n'a pas des canaux en entiers sur 8 bits, on sort en erreur

// Ecriture de l'image
Expand All @@ -813,14 +815,14 @@ int LibtiffImage::write_image ( uint8_t* buffer) {
} else {
BOOST_LOG_TRIVIAL(error) << "Image to write (from a buffer) has not 8-bit uint samples : " << filename;
print();
return -1;
return -1;
}

return 0;
}

int LibtiffImage::write_image ( uint16_t* buffer) {

// Si l'image à écrire n'a pas des canaux en entiers sur 16 bits, on sort en erreur

// Ecriture de l'image
Expand All @@ -835,14 +837,14 @@ int LibtiffImage::write_image ( uint16_t* buffer) {
} else {
BOOST_LOG_TRIVIAL(error) << "Image to write (from a buffer) has not 16-bit uint samples : " << filename;
print();
return -1;
return -1;
}

return 0;
}

int LibtiffImage::write_image ( float* buffer) {

// Si l'image à écrire n'a pas des canaux en flottant sur 32 bits, on sort en erreur

if ( sample_format == SampleFormat::FLOAT32 ) {
Expand Down Expand Up @@ -874,10 +876,10 @@ int LibtiffImage::write_line ( uint8_t* buffer, int line) {
} else {
BOOST_LOG_TRIVIAL(error) << "Image to write (line by line) has not 8-bit uint samples : " << filename;
print();
return -1;
return -1;
}


return 0;
}

Expand All @@ -894,15 +896,15 @@ int LibtiffImage::write_line ( uint16_t* buffer, int line) {
} else {
BOOST_LOG_TRIVIAL(error) << "Image to write (line by line) has not 16-bit uint samples : " << filename;
print();
return -1;
return -1;
}

return 0;
}

int LibtiffImage::write_line ( float* buffer, int line) {
// Si l'image à écrire n'a pas des canaux en flottant sur 32 bits, on sort en erreur

if ( sample_format == SampleFormat::FLOAT32 ) {
if ( TIFFWriteScanline ( tif, buffer, line, 0 ) < 0 ) {
BOOST_LOG_TRIVIAL(error) << "Cannot write file " << TIFFFileName ( tif ) << ", line " << line ;
Expand All @@ -916,4 +918,3 @@ int LibtiffImage::write_line ( float* buffer, int line) {

return 0;
}