Analyzing patterns, motion and images

Got a question, and perhaps Panda3D isnt the best visualization tool for the job.

Id like to analyze video, animation, images within a visual environment for say edge detection, pattern recognition, through data streams (streaming video for instance or packet analysis).

Would PANDA be an appropriate tool for such visualizations?

Thanks for any info!

i’d say… no.
panda doesnt really have many buildin functions providin pattern or movement analysation of images or videos.
with the exception of the AR toolkit which got into panda lately. it can recognize certain patterns, their position and rotation relative to the camera.
but you can use it outside of panda,too.

just my 2 cents. but i bet there are better tools for image analysation

Thanks for the reply.

Looking around I also agree with you…3D engines typically arent used for any type of recognition systems ( be cool if there was one, could create some kind of low-low-low- budget motion capture based on moving people)

Any way, I do know I want to stick with Python. Thanks for your input.

well you could try to stick those ARToolkit signs all over your body and record it with several cameras, merging the result… you’d have a quite low cost motion capturing.
if you just wanna capture motions for animations (non realtime) you can simply take 2 cameras, set up in a 90° angle. make 2 movies of your animations and put them into your editor-background… it’s fast, cheap, and produces good results. it’s not fully automated but you can make a most basic movements like run,jump and so on all on a single day.

there are VERY few tools beeing able to detect motion. tracking camera motion already is tricky. but motion capturing like it’s known from movies is really tricky with low-low-low-budget. only choises i would know would be the AR-signs or manual-tracking with the movie in background.

I have been playing with OpenCV lately. It has python bindings and a set of python example programs that show how to do motion tracking and pattern recognition. You can find it at .

Thank you Laurens,

I havent heard of those libraries before. I appreciate the link.