PDA

View Full Version : Speecys SPC-101C direct Servo Control



cdraptor
10-13-2007, 05:58 PM
The Speecys SPC-101C has an SDK called Open Roads - I have been working on translating the documentation, understanding the calls and building my own C# wrapper to simplify using the SPC-101C. The Open Roads SDK is a very powerful and relatively simple approach - you communicate to the SPC-101C via TCP/IP socket datastreams - each command has a specific data packet format. So far from example code and question/answers with Koji at Speecys I've been able to setup in my app calls to the Motion Files and now I have a direct servo control function which you can see in this video I am using to control the neck servo by providing the angle - currently I am not setting the speed, so it's default - my next step is to provide the speed parameter. I am going to be setting up control for the midsection body servo for looking down and up.

I've added image snapshot capture to the application as well which I am planning to use with some facial recognition and object recognition libraries (more info on that to come soon). What I should be able to do now with servo control and using the object/facial libraries is to provide active object tracking for the SPC-101C. So enjoy the vid, more to come, maybe before the weekend is out


http://www.youtube.com/watch?v=vdwlNpXA1PU

Alex
10-15-2007, 04:19 PM
Great stuff cdraptor, thanks for the update!!

Wait a sec.. I knew the SPC-101C was going to have voice recognition (http://forums.trossenrobotics.com/showthread.php?t=1202), but facial and object recognition?!? That's sick man! I don't remember hearing about that feature...

:D

cdraptor
10-16-2007, 01:56 PM
Well the SPC-101C doesn't have that on board - but it's a platform that lets you develop super powerful applications on the PC which directly control the SPC-101C via WiFi. The video from the robot is via wireless camera, that you can hook into any analog video capture - I am using the Pinnacle HDTV Pro Stick on my laptop which allows me to use any WDM device library such as the Image Control 3.0 component you see on my videos. I found some Facial and Object recognition libraries so for processing information like that I have the full power of the PC. The voice recognition I show is also on the PC communicating via TCP/IP to tell the robot - I am using the .NET 3.0 Framework libraries on Vista for speech synthesis and voice recognition.