[GIL] Black filled or not ? - Boost-users - lists.preview.boost.org

newer
[asio] asio-client only connects...

[GIL] Black filled or not ?

older
Re: [Boost-users]...

Eloi Du Bois

13 Jan 2010 13 Jan '10

12:37 p.m.

Hi all ! I would like to know, when creating an image like that: rgb32f_image_t img(w, h); is the image filled with 0 by default or should I call fill_pixels ? Thank you ! Eloi.

Attachments:

attachment.html (text/html — 209 bytes)

Reply

Sign in to reply online Use email software

Show replies by date

Christian Henning

13 Jan 13 Jan

3:36 p.m.

Hi Eloi, for performance reasons gil is not initializing the array with a default value. Also, I wouldn't trust the compiler to do anything special since it might be different from platform to platform. Just call fill_pixels which in your case ( rgb32f_image_t ) should be reduced to a memset call. Regards, Christian On Wed, Jan 13, 2010 at 7:37 AM, Eloi Du Bois wrote:

Hi all !

I would like to know, when creating an image like that: rgb32f_image_t img(w, h);

is the image filled with 0 by default or should I call fill_pixels ?

Thank you ! Eloi.

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

Reply

Sign in to reply online Use email software

Eloi Du Bois

4:40 p.m.

I tried to do a fill_pixels, but the thing is I get an error (because I am doing special things) on something like this: typedef pixel

comp_pixel_t; typedef image comp_image_t; typedef typename view_type_from_pixel::type comp_view_t; typedef typename comp_view_t::x_iterator tmp_iterator;

comp_image_t itmp( dst.dimensions( ) ); comp_view_t tmpv( view( itmp ) ); fill_pixels(tmpv, 0); But this is not working... Anyway, thank you Eloi. 2010/1/13 Christian Henning

Hi Eloi, for performance reasons gil is not initializing the array with a default value. Also, I wouldn't trust the compiler to do anything special since it might be different from platform to platform. Just call fill_pixels which in your case ( rgb32f_image_t ) should be reduced to a memset call.

Regards, Christian

On Wed, Jan 13, 2010 at 7:37 AM, Eloi Du Bois wrote:

...
Hi all !

I would like to know, when creating an image like that: rgb32f_image_t img(w, h);

is the image filled with 0 by default or should I call fill_pixels ?

Thank you ! Eloi.

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

Reply

Sign in to reply online Use email software

Christian Henning

6:59 p.m.

Please use: fill_pixels( tmpv, comp_pixel_t( 0.f, 0.f, 0.f )); This call wont be a memset unfortunately. But you could always just get the char* and memset yourself. Regards, Christian On Wed, Jan 13, 2010 at 11:40 AM, Eloi Du Bois wrote:

I tried to do a fill_pixels, but the thing is I get an error (because I am doing special things) on something like this:

        typedef pixel

...
comp_pixel_t;         typedef image comp_image_t;         typedef typename view_type_from_pixel::type comp_view_t;         typedef typename comp_view_t::x_iterator tmp_iterator;

        comp_image_t itmp( dst.dimensions( ) );         comp_view_t tmpv( view( itmp ) );         fill_pixels(tmpv, 0);

But this is not working...

Anyway, thank you Eloi.

2010/1/13 Christian Henning

...
Hi Eloi, for performance reasons gil is not initializing the array with a default value. Also, I wouldn't trust the compiler to do anything special since it might be different from platform to platform. Just call fill_pixels which in your case ( rgb32f_image_t ) should be reduced to a memset call.

Regards, Christian

On Wed, Jan 13, 2010 at 7:37 AM, Eloi Du Bois wrote:

...
Hi all !

I would like to know, when creating an image like that: rgb32f_image_t img(w, h);

is the image filled with 0 by default or should I call fill_pixels ?

Thank you ! Eloi.

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

Reply

Sign in to reply online Use email software

Eloi Du Bois

8:45 p.m.

Okay, I see. But the problem is that I don't know the pixel number of channel. It have to be full generic. That's why I am trying this (unsuccessfully): comp_pixel_t black; color_convert( bgil::gray8_pixel_t( 0 ), black ); fill_pixels( tmpv, black); 2010/1/13 Christian Henning

Please use:

fill_pixels( tmpv, comp_pixel_t( 0.f, 0.f, 0.f ));

This call wont be a memset unfortunately. But you could always just get the char* and memset yourself.

Regards, Christian

On Wed, Jan 13, 2010 at 11:40 AM, Eloi Du Bois wrote:

...
I tried to do a fill_pixels, but the thing is I get an error (because I am doing special things) on something like this:

typedef pixel

...
comp_pixel_t; typedef image
comp_image_t; typedef typename view_type_from_pixel::type comp_view_t; typedef typename comp_view_t::x_iterator tmp_iterator;

comp_image_t itmp( dst.dimensions( ) ); comp_view_t tmpv( view( itmp ) ); fill_pixels(tmpv, 0);

But this is not working...

Anyway, thank you Eloi.

2010/1/13 Christian Henning

...
Hi Eloi, for performance reasons gil is not initializing the array with a default value. Also, I wouldn't trust the compiler to do anything special since it might be different from platform to platform. Just call fill_pixels which in your case ( rgb32f_image_t ) should be reduced to a memset call.

Regards, Christian

On Wed, Jan 13, 2010 at 7:37 AM, Eloi Du Bois wrote:

...
Hi all !

I would like to know, when creating an image like that: rgb32f_image_t img(w, h);

is the image filled with 0 by default or should I call fill_pixels ?

Thank you ! Eloi.

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

Reply

Sign in to reply online Use email software

Christian Henning

10:04 p.m.

Eloi, this would be my solution. Wording might be a little off but I'm sure you'll get the idea: #include using namespace std; using namespace boost; using namespace gil; template< typename Data, int N > struct black_pixel_factory { BOOST_STATIC_ASSERT(( false )); }; template< typename Data > struct black_pixel_factory< Data, 1 > { typedef pixel< Data, devicen_layout_t< 1 > > pixel_t; static const pixel_t black() { return pixel_t( 0 ); } }; template< typename Data > struct black_pixel_factory< Data, 2 > { typedef pixel< Data, devicen_layout_t< 2 > > pixel_t; static const pixel_t black() { return pixel_t( 0, 0 ); } }; template< typename Data > struct black_pixel_factory< Data, 3 > { typedef pixel< Data, devicen_layout_t< 3 > > pixel_t; static const pixel_t black() { return pixel_t( 0, 0, 0 ); } }; template< typename Data > struct black_pixel_factory< Data, 4 > { typedef pixel< Data, devicen_layout_t< 4 > > pixel_t; static const pixel_t black() { return pixel_t( 0, 0, 0, 0 ); } }; template< typename Data > struct black_pixel_factory< Data, 5 > { typedef pixel< Data, devicen_layout_t< 5 > > pixel_t; static const pixel_t black() { return pixel_t( 0, 0, 0, 0, 0 ); } }; template< typename View > void foo( const View& dst ) { typedef pixel

...
comp_pixel_t; typedef image comp_image_t; typedef typename view_type_from_pixel::type comp_view_t; typedef typename comp_view_t::x_iterator tmp_iterator;

typedef black_pixel_factory< bits32f, num_channels<View>::value > black_pixel_factory_t; typename black_pixel_factory_t::pixel_t black = black_pixel_factory_t::black(); comp_image_t itmp( dst.dimensions( ) ); comp_view_t tmpv( view( itmp ) ); fill_pixels( tmpv, black); return; } int main() { rgb32f_image_t img; foo( view( img )); return 0; } Christian

Reply

Sign in to reply online Use email software

Nicolas Lelong

14 Jan 14 Jan

7:24 a.m.

Depending on if you can rely on a boost gil extension, numeric/pixel_numeric_operations.hpp has a zero_channels function. This would give (untested) : comp_pixel_t black; zero_channels( black ); fill_pixels( tmpv, black ); Cheers, Nicolas On Wed, Jan 13, 2010 at 9:45 PM, Eloi Du Bois wrote:

Okay, I see. But the problem is that I don't know the pixel number of channel. It have to be full generic. That's why I am trying this (unsuccessfully): comp_pixel_t black; color_convert( bgil::gray8_pixel_t( 0 ), black ); fill_pixels( tmpv, black);

Reply

Sign in to reply online Use email software

Lubomir Bourdev

10:58 a.m.

GIL core provides channel-level equivalents for some STL algorithms, including std::fill. These have the “static_” prefix since they are compile-time recursive: rgb8_pixel_t p; static_fill(p, 10); assert(p == rgb8_pixel_t(10,10,10)); http://stlab.adobe.com/gil/html/group___color_base_algorithm_fill.html Lubomir On 1/13/10 11:24 PM, "Nicolas Lelong" wrote: Depending on if you can rely on a boost gil extension, numeric/pixel_numeric_operations.hpp has a zero_channels function. This would give (untested) : comp_pixel_t black; zero_channels( black ); fill_pixels( tmpv, black ); Cheers, Nicolas On Wed, Jan 13, 2010 at 9:45 PM, Eloi Du Bois wrote: Okay, I see. But the problem is that I don't know the pixel number of channel. It have to be full generic. That's why I am trying this (unsuccessfully): comp_pixel_t black; color_convert( bgil::gray8_pixel_t( 0 ), black ); fill_pixels( tmpv, black);

Reply

Sign in to reply online Use email software

Eloi Du Bois

1:41 p.m.

That is a very good idea :) Thank you very much ! Eloi. 2010/1/14 Lubomir Bourdev

GIL core provides channel-level equivalents for some STL algorithms, including std::fill. These have the “static_” prefix since they are compile-time recursive:

rgb8_pixel_t p; static_fill(p, 10); assert(p == rgb8_pixel_t(10,10,10));

http://stlab.adobe.com/gil/html/group___color_base_algorithm_fill.html

Lubomir

On 1/13/10 11:24 PM, "Nicolas Lelong" wrote:

Depending on if you can rely on a boost gil extension, numeric/pixel_numeric_operations.hpp has a zero_channels function.

This would give (untested) :

comp_pixel_t black; zero_channels( black ); fill_pixels( tmpv, black );

Cheers, Nicolas

On Wed, Jan 13, 2010 at 9:45 PM, Eloi Du Bois wrote:

Okay, I see. But the problem is that I don't know the pixel number of channel. It have to be full generic. That's why I am trying this (unsuccessfully): comp_pixel_t black; color_convert( bgil::gray8_pixel_t( 0 ), black ); fill_pixels( tmpv, black);

_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

Reply

Sign in to reply online Use email software

5364

Age (days ago)

5365

Last active (days ago)

Download

8 comments

4 participants

tags

participants (4)

Christian Henning
Eloi Du Bois
Lubomir Bourdev
Nicolas Lelong