Access to energy pictures [edit: included since v0.4.0]
Forum » General / Feature requests » Access to energy pictures [edit: included since v0.4.0]
Started by: Julien Narboux (guest)
On: 1239197109|%e %b %Y, %H:%M %Z|agohover
Number of posts: 6
rss icon RSS: New posts
Access to energy pictures
Julien Narboux (guest) 1239197109|%e %b %Y, %H:%M %Z|agohover

It would be nice to have function to access to the energy array, to allow the gui to give an idea of the way the selected energy function works.

Thanks,

Julien

unfold Access to energy pictures by Julien Narboux (guest), 1239197109|%e %b %Y, %H:%M %Z|agohover
Re: Access to energy pictures
UnNeuroneUnNeurone 1239214429|%e %b %Y, %H:%M %Z|agohover

This was actually already planned for next version 0.4, but I am also considering to add new energy functions, or maybe some interface for user-defined energy functions, so I don't know how long this will take.
I was thinking to add a function like lqr_carver_get_energy(carver, x, y), and maybe some other function like lqr_energy_compute(image_buffer, image_type, x, y, energy_function, energy_function_data) which could be used before creating the carver; if you have other suggestions you're welcome.

unfold Re: Access to energy pictures by UnNeuroneUnNeurone, 1239214429|%e %b %Y, %H:%M %Z|agohover
Re: Access to energy pictures
Julien Narboux (guest) 1239863800|%e %b %Y, %H:%M %Z|agohover

It would be also nice to have access to the Lqr energy functions. The Gui would be able to compute thumbnails of the energy pictures for quick preview before choosing the energy function.

As far as the new energy function are concerned, maybe a convolution function would be usefull, then you could have Sobel and Prewit filters.

Julien

unfold Re: Access to energy pictures by Julien Narboux (guest), 1239863800|%e %b %Y, %H:%M %Z|agohover
Re: Access to energy pictures
UnNeuroneUnNeurone 1239895399|%e %b %Y, %H:%M %Z|agohover

I will do that I guess if I implement the function lqr_energy_compute as sketched above, but wouldn't the preview be misleading if the energy is computed over a thumbnailed image? I think one should compute the energy over the full image and then thumbnail the result.

I will think about a general framework for convolutions, it seems a good idea; however, it should just be a special case of a user-defined functions framework - I think I will implement that first (as soon as I can, it requires some work).

unfold Re: Access to energy pictures by UnNeuroneUnNeurone, 1239895399|%e %b %Y, %H:%M %Z|agohover
Re: Access to energy pictures
Julien Narboux (guest) 1239900285|%e %b %Y, %H:%M %Z|agohover

ok thanks !

unfold Re: Access to energy pictures by Julien Narboux (guest), 1239900285|%e %b %Y, %H:%M %Z|agohover
Re: Access to energy pictures [edit: included since v0.4.0]
UnNeuroneUnNeurone 1242006076|%e %b %Y, %H:%M %Z|agohover

Note: the functionality is now included in the library, but the API is different from the one sketched above.

last edited on 1242006103|%e %b %Y, %H:%M %Z|agohover by UnNeurone + show more
New post
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License