[Elphel-support] Scanline sequence in compressor

[Alex Gomez]

Hi,

I have an Elphel 353 camera with a monochrome sensor. As a first step, I am
trying to implement edge detection on the 16x16 tiles that the compressor
uses. I am looking to use the pixels in the Y-buffer, but I am not sure
about the addressing convention that is used to store the tile or exactly
which pixels are stored. The comment in i_csconvert_mono says y_out ouput is
in "scanline sequence". Does that mean it is taking a line of 256 contiguous
pixels and turning them into a tile (with each pixel's address corresponding
to its location in the tile), or is it taking 16 pixels from one line, then
taking the corresponding 16 pixels from the next line (the 16 pixels
directly below it), etc. in order to form the 16x16 tile?

Thank you for your help! I have the simulation running, but it I am finding
this tricky to figure out.

Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://support.elphel.com/pipermail/support-list_support.elphel.com/attachments/20110126/87538024/attachment-0002.html>