Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations GregLocock on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Aerial Geo-Referencing Equipment Question for Helicopter

Status
Not open for further replies.

WGilliam

New member
Feb 18, 2004
1
I am installing a stabilized camera gimbal on a helicopter. This installation will also include another downward looking camera. I need to be able to take this video (from the downward looking camera)and geo-reference it, pulling accurate (1 meter) latitude and longitude coords for specific objects within the video (for example - I need to make a mosaic and then pull the coords for a house or another type building out of the frame). The associated coord stream will then be exported to an ESRI based software processing package. Does anyone know how to get the GPS accuracy level down to a meter in a fairly economical way? The new WAAS augementation will get the accuracy range to about 3 meters. But, I need the points on the geo-referenced frames to be in the neighborhood of a meter or less. Any suggestions would really be appreciated. Thanks.
 
Replies continue below

Recommended for you

Actually, it more complicated than that.

a> You need something like RTK or a differential GPS corrector. You can potentially get position accuracy down to less than 1 ft. Not sure if that applies to altitude though.

b> BUT, even then, that just locates your helo accurately. Now, you need to accurately reference your helo heading and accurately reference your imager to the inertial reference. AND, if you assume that all the inertial sensors related to the helo are accurate to 1 milliradian, that would automatically result in a >1 m error if the slant range is more than 1 km. In order to make that all work, you'll need to determine your heading/attitude and the line of sight for the imager to better than about 100 microradians, which means a awfully decent inertial nav system. Even if your camera is only downlooking, you'll still need to know attitude as well as any mounting error in that camera.

c> THEN, you'll need to be able to align your sensors in-flight, e.g., given your DGPS location and heading and a known land reference, back out the correct line of sight angles and attitude. This will probably require doing a large loop around your known reference point.

TTFN
 
Irstuff is right - I have heard that some people use two GPS receivers on the plane/helicopter (both ends)for instance thus helping with the orientation etc- this then gets referenced to the camera-
 
Have you looked into camera tracking software?

If you can get a couple of accurate ground coordinates visible in your video, you should be able to accurately plot your air position.

I've previously played with ICARUS from the Univ. of Manchester. It worked very well, though it has now gone commercial and is expensive-
Looks like I'm going to be going with SynthEyes which sells for $349 and is starting to build a pretty impressive credit list in the movie industry-
I'll be using the software for tracking video shot from a police helicopter to plot it's position during pursuits, surveilance video and crime / collision scene flyovers.

Maybe camera tracking is low tech compared to gps, but then again, it might do the trick.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor