Basic Robot Vision-Fez Panda

Hello,

I want to know if it will be possible to run basic real time robot/computer vision algorithms on the Fez Panda.
I know that these things usually require huge amounts(relatively) of processing overhead and advanced algorithms are impossible.

I’m talking about basic stuff like line following and color blob tracking.

This guy managed to do it on a parallax propeller(everything runs on the propeller) - [url]http://www.circuitcellar.com/archives/viewable/224-Sander/index.html[/url].

I don’t want to but the propeller because I already know programming in
C++ and I don’t want to learn an entire new language

I was wondering if the same thing was possible using a serial camera.
like this- [url]http://www.sparkfun.com/products/8739[/url]

He used a camera with analog output and just used the gray scale images.

I was wondering if similar filters could be used to filter much of the data
and use the remaining for line following and color blob tracking etc.

I realize that there are many fundamental differences between the two
controllers like the cogs etc.

Note-I’m posting on another forum as well because I want to buy a board which will be able to implement a similar setup.
Also English isn’t my first language,so if you find this hard to read, my apologies.

Thanks in advance.

Yes, GHI products can do that. You can do ti in C# and if your algorithmic are complex then you can do it in RLP (native C)

I don’t think the Panda will work for you. Ram is limited and it doesn’t contain JPG or BMP libraries.

Also, the camera you linked to isn’t serial, it’s a normal video camera.

You would need something like this: http://www.sparkfun.com/products/10061

But, for the panda, you will have to add your own jpeg decoding. I think the EMX/Cobra contains jpeg decoding…

I think misunderstood the question. If you want to make a line follower robot then this can be done easily but not by using a camera but using reflective sensors. Color blob tracking can’t be directly accomplished but there are cameras that have such feature built internally then you only have to get the position.

You can’t read (or it will be extremely difficult) to read an analog camera using any microcontroller. You can do it on propeller chip because it is different than typical microcontroller. You can use it to generate audio video or also to device video and audio. propeller chip is a hybrid between a micro and FPGA. But then propeller is not a replacement for micros, it doesn’t even have interrupts…each device has its area where it excels at.