OSD Board (Parallax Propeller Backback)
by marcwolf
Writer, Inventor, Creator
14 years ago
This little board performs the comples task of taking the video feed from the right camera and overlaying text onto it, and then sending the merged signal back to the video goggles.
At the moment the wores on the left go to my computer running a serial comms program. But ultimately it will be controlled by a 28X2 PicAxe chip that will do the folloing
1. Read GPS Lat, Long, Time, Date, Distance, Direction, and Alt
2. Read 4 different temperature sensors around the suit
3. Read the local clock and stopwatch
4. Collect power usage and calculate battery time
5. Check for Voice Command Activation
And then send this to this board to display on the goggles.
At the moment the wores on the left go to my computer running a serial comms program. But ultimately it will be controlled by a 28X2 PicAxe chip that will do the folloing
1. Read GPS Lat, Long, Time, Date, Distance, Direction, and Alt
2. Read 4 different temperature sensors around the suit
3. Read the local clock and stopwatch
4. Collect power usage and calculate battery time
5. Check for Voice Command Activation
And then send this to this board to display on the goggles.
401
Views
4
Comments
1
Favorites
General
Rating
Category
Sub-Category
Species
Resolution
File Size
All
All
Unspecified / Any
640 x 480
53.9 kB
FA+

And..
Do you do commissioned work?
Marc
How would using a board like this stack up to something like a RaspberryPi?
I remember looking in the Propeller boards years ago, bout couldn't think of anything to do with them.
Ok - The Prop has one main advantage and that is it is excellent at making a TV signal (built in NTSC waveforms software), and it can be reasonably programmed to have additional features on that functionality i.e. reverse graphics.
For example it can take a video stream, break it into pixels, add additional pixels, and then regenerate the video stream. As this is a task that can benefit through pararellism the COG concept works fine.
The Pi already has all of the video generation hardware built in so it is very difficult to err. leap into the video stream and hack it o do something like a OSD.
My current experiments with blob following etc is using the SimpleCV libraries which work very well on the Beaglebone Black. I did try using the Pi for that bit the Pi does not that much processing power. Its main claim to fame is that the has the H264 codec built in so to make playing video's etc very easy.
If you know how James Cameron did the facial mapping for Avatar then I will be doing the same except with 8 point around the lips for real time lipsyncing
I have more of my work at https://www.marcwolf.org/gallery
Thanks for commenting.
Marc