[mythtv] Comm Flag questions/confusion

[Mark Spieth]

On 9/6/2012 1:02 AM, Chris Pinkham wrote:
> * On Tue Sep 04, 2012 at 12:53:14PM +1000, Mark Spieth wrote:
>> On 9/4/2012 11:07 AM, Bryan wrote:
>>> 2) The comm flag code passes kDecodeLowRes to the player, which causes
>>> the frames coming out of the player to be 1/4 their original size
>>> (1920x1080 -> 420x270).  Since the player does this, these are
>>> scaled/blended so not all of the data is lost.  This size&data is used
>>> for logo detection.  But for Scene Change and Blank Frames, the comm
>>> flagger then only uses every 4th byte, ignoring 3/4 of the scaled down
>>> data.  Could it be that comm flag's build in "every 4th pixel" is
>>> supposed to do the same thing as kDecodeLowRes, and thus using both at
>>> the same time was never intended?
>> I read somewhere that this could be why logo det doesnt work very well
>> any more.
I think there may be a root cause here as VideoFrame's buffer is 16 byte 
aligned and the display width is not guaranteed to be so. Thus all 
processing that handles pixels with display width as row step will be wrong.
Not sure what this means yet as fixing this doesnt make classic logo det 
edge detector work for the one tested so far.
For a SD 720x576 the divide by 4 of x and y converts to 180x144 which is 
not much for any detail in any case.
The logo for GO! is only 18x11 pixels though it does stand out with 
canny on a time averaged image (10 @ 2 sec/frame).

I'm playing with these algos currently
http://www.busim.ee.boun.edu.tr/~sankur/SankurFolder/EUSIPCO2009_Automatic_TV_Logo.pdf
and later Ill try
http://www.nlpr.ia.ac.cn/2007papers/gjhy/gh40.pdf
http://avss2012.org/2006papers/gjhy/gh40.pdf
which use color for image gradient.

see what works.

cheers
mark