Log in

No account? Create an account

Previous Entry | Next Entry

Cricket videos: baby steps

Task switching is not fun. It took me a little while to remember how to use vi. Then it took me a little while to correctly rewrite a simple script to capture 5 minutes of images from the camera. Then it took me a while longer to figure out how to get mEncoder to run under Cygwin. In case you wonder this, too, here's what has to happen. First, you have to tell Cygwin where you're keeping the executable program files for mEncoder. I put mine in the Windows Program Files directory. You can add this location to the other places where Cygwin searches for programs by appending the Path variable thusly:

PATH=$PATH:"/cygdrive/c/Program Files/mplayer"

Or what-have-you. Once this is accomplished it's possible to run mEncoder to make videos thusly:

mencoder.exe mf://*.jpg -mf w=800:h=600:fps=10:type=jpg -ovc lavc -lavcopts vcodec=mpeg4:mbd=2:trell -oac copy -o VideoFileName.avi

In a nutshell, this groups together all of the images in whatever folder you're in and turns them into a video with 10 frames per second (an arbitrary number at the moment). My images are currently 1280x720, so I adjusted the width and height values accordingly.

Now I am back to the step of figuring out how to run BioTrack with an Ubuntu virtual machine on a Windows computer. I suspect this will involve figuring out how to navigate another set of directories.

Fatigue for process setup is running high, what with what has been involved in learning how to use Cygwin and vi already (mencoder knowledge dates back to the prior cricket video project in Texas, where I set up a computer in a simpler fashion by using fake dual-boot Ubuntu). I need to remind myself that I'm making progress overall.

So, to do (for this project):
-Get BioTrack actually up and running
-First, get virtual machine working
-Collect more test video at night, preferably with a color checker to assess what kinds of paint schemes will work for distinguishing between individuals
-Test whether the current image collection method will generate adequate videos for BioTrack to actually track crickets accurately
-Figure out any file size limits I'm going to run up against
-Streamline workflow


Latest Month

September 2017


Powered by LiveJournal.com
Designed by Naoto Kishi