On Mon, Jun 04, 2012 at 01:54:34AM -0700, Shervin Emami wrote:
> I was going to recommend OpenCV but you say you had no luck with it.
> Was it an install / configuration problem? And if so, did you try
> following the 2 official Android tutorials?
System limitations currently prevent me from using the NDK (that
problem goes away when my Mac arrives late this week or early next
week---I'm just doing the research part early so I'll be ready to
get back to work on my app when it's here), so I didn't go through
the NDK version of the setup, but the other, yes. I tried to figure out
the conversion from a few samples how to use stuff not in the samples,
and just didn't have any luck. I'll probably try OpenCV again when I CAN
use the NDK, as I can just use the C++ code in the existing non-Android
tutorials without having to change it.
The next question, however, is whether or not that will bloat my app
like a lot of the samples were the last time I checked. Simple C
or C++ code probably wouldn't do that.
Honestly, I'm just looking for simple C or C++ code to do the filters
for my app, such as brightness, contrast, hue, tint, color balance,
sharpen, blur, diffusion, a stronger version of diffusion that covers
only a selected area ... also known as a fog filter, edge enhance, find
contours, and other weird stuff like the last two. I'm also looking
for a C or C++ version of blending two images like the PorterDuff
SR_OVER method in the SDK. Oh, and also converting a byte[] array
to a bitmap (up to the maximum photo size for whatever is the maximum
Android camera resolution for whatever device has the highest), so
the user can then do some post-processing on it, save it as a PNG
(or maybe even TIFF) instead of JPEG, etc.
I've got those in Java, and they work, but some of them are SLOW (even
on my Acer Iconia A500 tablet), and they suffer from Java's limited
memory. I'm hoping to solve both by using native C/C++ for all of
those operations.
I've found some tutorials and libraries in C++ and in C, but they
all either have licenses that prevent use in a paid app (or in an
app, period, for that matter), or they assume that nobody likes
full-color photos (after all, everybody prefers grayscale over
color for ALL photos[1], right?).
Thanks,
--jim
[1] Yes, it is better from an artistic point in SOME photography,
but not always.....
--
THE SCORE: ME: 2 CANCER: 0
73 DE N5IAL (/4) MiSTie #49997 < Running FreeBSD 7.0 >
spook...@gmail.com ICBM/Hurricane: 30.44406N 86.59909W
"'Wrong' is one of those concepts that depends on witnesses."
--Catbert: Evil Director of Human Resources (Dilbert, 05Nov09)