I am working on a robot but I am looking for a cheaper but locally available optical sensor but all I can really get are web cams so I am wondering can web cams some how be interface with a microcontrolers directly or do you know of a fairly decent camera that is affordable and can be interfaced with a microcontroller.
Also dose any one know of a thermal camera that interfaces with a microcontroler thanks.
See I am wanting to build some drones I wanted to use the camera so I can see what is going on and thermal so it can follow certain heat signatures.
Ok due to how expensive and rare thermal cams are I am going to drop it and go on to plan B. Luckily I have been experimenting with some artificial neural networks in C# that can track shapes so I will use that instead. So lets talk regular cameras . Is there a fairly decent, not microscopic, med to low cost camera that can be interfaced into a microcontroler. I saw the ones so far on the sight but they just don’t look like they have the power I want and in needs to connect to a microcaotroler because everything needs to be self contained with no comp connections. Any one know of one or a sight that kind of specializes in that stuff?
Thanks
See the camras I am seeing so far just don’t seem powerful enough. See they do not look like they have a good zoom. Also the Devantech 8 Pixel Thermal Array Sensor would work for me but it can only detect temps form 6 ft away and I need it to see farther then that. Thanks for the help sofar. Also is there a differnce between inferred Range finders and ultersonice ones except for their price and detection range?
Unless your application is PC based, using machine vision may require a more complete solution then a simple camera, such as the CMUCam or POB-EYE vision sensors. This really does depend on what you need to do with the video.
The same deal applies to thermal imagery, although no vision systems are available at the moment for thermal imaging. However, a microcontroller a thermal array sensor may prove useful. Thermal cameras are very expensive. Hopefully their price will go down in the next few years. It would be great to have such a thermal imagery camera with integrated image processing as seen with the CMUCam.
The main difference between infrared and sonar rangefinders is the core technology behind them.
Generally speaking, sonar rangefinders use sound so their weakness could be porous materials such as lace curtains. Similarly, Infrared rangefinders use light so a flat black mate surface may not be detected if it does not reflect enough light. It is suggested to use a combination of technologies to obtain the best results. I many cases this is a non-issue.
Be forewarned however: The Logitech orbit sits within the base or can be elevated using the stand. Only the connection holds it in place, and the camera housing is very fragile (I learned this when turning the camera upside down and having the camera fall out of the base onto tile). I would suggest a more rugged camera (even if it remains in one place). The “neck” makes it even more precarious. My suggestion - use a servo pan and tilt with a standard webcam.
It looks like you are trying to make your own CMUcam?
The little CMOS camera could probably work but this is very advanced:
robotshop.ca/sfe-cmos-camera-640-480.html
People are still figuring out how to get these to work. You are probably better off with a CMUcam if you don’t want to reinvent the wheel. It offers alot of features you are looking for.
This is one of the reason why I chose to use a PC for a main controler. USB cameras like this one : logitech.com/index.cfm/webcam_communications/webcams/devices/3480&cl=ca,fr
can follow movements and has a real physical autofocus with a high quality definition. Also, plugin it to a PC, you don’t have to search for an energy source… I can develop a program that will be able to control the camera (autofocus, visual recognition, etc).
Good luck in your search.
Jeff
Thermal sensors are rare and expensive… But you don’t want your drones to be seen if I understand. Also, heat is a good way to detect movement.
What about an IR sensor ? I think… well… it can be possible. If you ‘draw’ an IR signal on the PC. It’s possible to detect and draw forms. And then, you can compare each image to extrapolate on a more detailed image (a head, a leg, whatever). But you’ll need a lot of IR sensors.
Good luck,
Jeff
