Hawaiʻi's Technology Community
Positioning System Accuracy
I understand that standard GPS devices such as are used in vehicles and iPods and the like have a resolution of, at best, about ten feet. Are there positioning devices with resolutions better than that? Are there any that can determine position to with inches, or even sub-inch resolution?
Views: 99
Replies to This Discussion
-
Permalink Reply by Nate Sanders on
-
Roger,
Are you interested only in something that is applicable almost anywhere (like GPS) or is within-room tracking good enough?
InterSense (http://www.intersense.com/) makes ultrasonic trackers where they claim resolution of 1.5mm untethered. I used one of their older models in 2003 in an immersive projector-camera environment and I certainly experienced extremely good tracking. We were using precise head tracking to correctly render first-person views in our immersive environment and it sure seemed to be sub-inch resolution (which makes sense, given that 1 inch = 25.4 mm and they're now claiming 1.5mm resolution).
There are other trackers, too, but I remember InterSense being an especially big player in this field and having high-quality products, good customer service, and knowledgeable salespeople.
-- Nate
-
Permalink Reply by Roger Garrett on
-
Nate,
The application is agriculture and other outdoor, multi-acre areas. So, InterSense products won't work. And ultrasonic sensors in general won't work. I'm actually working on developing a device using RF and multiple transponders positioned around the field. I've searched around the Internet for anything similar and haven't seen anything, so I think my approach is unique. But you never know, there may already be something out there. That's why I asked on this forum, to see if anyone might know of an existing device.
- Roger
-
-
Roger,
What precision do you need? And is the thing being tracked taller than the stuff around it? In other words, is it like a 6 foot tall person in a field of 3 foot tall plants or are the plants much taller?
Would cameras mounted on poles (high enough to see most of the field) be a possibility? Maybe using multiple cameras to alleviate occlusion and also to add as sanity checks to the camera with the largest number of pixels on target?
Camera-based metrology is an active area of research in computer vision. It might be worth a look.
I really think there's probably something already made for what you want, though -- I could be wrong. If you find anything or want to discuss your method more, I'd certainly like to know about either.
-- Nate
Roger Garrett said:Nate,
The application is agriculture and other outdoor, multi-acre areas. So, InterSense products won't work. And ultrasonic sensors in general won't work. I'm actually working on developing a device using RF and multiple transponders positioned around the field. I've searched around the Internet for anything similar and haven't seen anything, so I think my approach is unique. But you never know, there may already be something out there. That's why I asked on this forum, to see if anyone might know of an existing device.
- Roger
-
-
Also, I would have thought that tracking of animals (which I thought used RF + triangulation) would apply to your situation. Does it not?
-- Nate
-
-
Nate,
There are two types of animal tracking. One simply identifies the animal, as when a cow passes through gate, and it does that using RFID technology. That does not actually determine the position of the animal. The other type is when a transmitter is attached to the animal. This can include a GPS device, in which case the resolution is still limited to about ten feet, or without a GPS unit it requires directional antennas used by human trackers and that only gives very general direction and range information.
So, no, the animal tracking techniques would not work.
- Roger
-
-
Nate,
The immediate application, or at least the one that prompted my interest in it, is for a robotic mower for Christmas tree farms. Mowing of the weeds between the trees is a major part of the labor for growing the trees. The robot mower would need to know the precise position of every single tree in the field, particularly the newly planted seedlings because they are shorter than the weeds (which also reduces the possibility that any vision system could handle the job, since it's even difficult for a human operator to distinguish the seedlings from the trees). I'm taking an RF, stationary transponders, mobile unit (on the robot mower) approach. If it works it becomes fairly straightforward to identify the tree locations, as well as the field boundaries, obstacles, gates, etc. and to then have the mower efficiently cut the weeds between the trees.
And there are mutiple other agricultural and non-agricultural applications.
Every few weeks I do another check of the Internet to see if there's anything like this and have yet to come across it.
- Roger
-
-
Roger,
There is animal tracking involving RF that is older than both RFID and GPS. It may not suit your purposes, however. I know very little about it, other than the fact that biologists have been tracking animals using RF for a long time.
If I have time later, I may look into this.
-- Nate
Roger Garrett said:Nate,
There are two types of animal tracking. One simply identifies the animal, as when a cow passes through gate, and it does that using RFID technology. That does not actually determine the position of the animal. The other type is when a transmitter is attached to the animal. This can include a GPS device, in which case the resolution is still limited to about ten feet, or without a GPS unit it requires directional antennas used by human trackers and that only gives very general direction and range information.
So, no, the animal tracking techniques would not work.
- Roger
-
-
I think I was thinking of LORAN, maybe, or something similar. GPS is more precise, evidently.
The problem you're talking about is really interesting. I think there are a lot of angles to attack it from.
I'd be interested in hearing updates.
Good luck,
Nate
© 2024 Created by Daniel Leuck.
Powered by
