I should note that when I wrote my first response I was under the impression that we had been using the PandaBoard. It turns out we are using the PandaBoard ES, which uses the OMAP4460 and not the DuoVero's OMAP4430.
Also because people seem to be curious, I'll elaborate on the performance differences. We are doing pupil detection using a 9-step process which includes converting the image to grayscale and using a histogram to crop the image around the pupil. Then we use a haar feature detection and the starburst algorithm to find where edge of the ellipse might be, followed by an ellipse fitting algorithm. We do this using TBB + Boost + OpenCV libraries. We have compiled both OpenCV and the algorithm with gcc (4.7.1) using -mfpu=neon, -mfpu=vfpv3, -mfpu=vfpv3-fp1, and no additional flags at all.
With ~95% accuracy, this algorithm ran at ~1-2 fps on the DuoVero and ~6 fps on the PandaBoard ES. With ~60% accuracy, we were able to run two instances of the algorithm on the DuoVero at ~8 fps (each) and ~17 fps (each) on the PandaBoard ES. We are currently looking into using the armcc compiler to improve this even more (anyone know how to do this?)
Now, I don't know the reasoning behind this, but I suspect it may be because the PandaBoard includes the drivers for the GPU which means support for OpenGL, which may allow the system to use OpenGL calls to accelerate matrix operations.
PS: If anyone has any ideas/suggestions/criticisms for things we are doing, please let me know.

On Wed, Feb 6, 2013 at 2:31 AM, William Bryan <william.bryan91@gmail.com> wrote:
For the DuoVero I have used the image provided by Steve Sakoman, which works relatively well (Thanks Steve!) It is just lacking in the driver domain, which is frustrating since I need to use some of the systems that are missing driver support. I have recently installed the Ubuntu Core on it as well, but I have not gotten it to boot properly yet. The uboot image provided by Ubuntu doesn't even remotely work on the DuoVero and the uboot image that is provided with Steve's image works a little better, but fails after loading most of the modules (which I expected). I haven't really attempted to fix that issue yet though. I'm hoping to get it working in the near future, as it comes with a lot of the proprietary TI drivers that I need.

As for the PandaBoard, we are using the Ubuntu image provided, and it works great.


On Feb 6, 2013, at 2:09 AM, James Ang <angweiyang@gmail.com> wrote:

Hi Will,

On Wed, Feb 6, 2013 at 4:17 AM, William Bryan <william.bryan91@gmail.com> wrote:
Good luck. The hardware itself is great, but the software support is lacking. As it stands, all provided system images lack support for the SGX540 PowerVR, the DSP, the ISS, et cetera.

I am currently using the DuOvero for image acquisition and processing as well. The results are mediocre at best, and I don't believe the kernels available utilize the hardware available to them properly. For instance, compiling and running one program on the DuOvero and the PandaBoard yielded a 10x better performance on the PandaBoard (in terms of frames processed per second -- loaded from a video on identical model microSDs)

So basically, if you are looking for something that requires little work right out of the box, the DuOvero probably isn't what you're looking for.

Just my 2 cents.


What is the OS that run on your Pandaboard and DuOvero?

Free Next-Gen Firewall Hardware Offer
Buy your Sophos next-gen firewall before the end March 2013
and get the hardware for free! Learn more.
gumstix-users mailing list

Will Bryan
Iowa State University | Computer Engineering
william.bryan91@gmail.com | 712.212.5027