Determine the location of an object being viewed by the camera.

Has anyone tried this, and where would be the best place to calculate it (APM, OSD, or GCS)?


Situation: You are flying around looking at things with your camera and see something of interest. Let's say a a broken fence on your farm. You point your camera's crosshair on it and read a location for where the camera is pointing so you can go out and repair it.


Given: The location of the aircraft (Heading, Altitude, and Position) and orientation of the camera (Azimuth and Look Angle).


Problem: Determine the location of the item being viewed by calculating it form the know aircraft position/direction and the camera’s azimuth/look angle. Then out put the location in a format that can be displayed in the Mission Planner software or video stream.


Views: 3397

Reply to This

Replies to This Discussion

I have an interest in the use of UAVs in SAR as well as damage assessment during a disaster. I have seen software somewhere that allows the locking on of a stationary target with a camera. I believe the gimble had it built in. In short, it used position and altitude data to give a lock once someone visually spotted a target, and then used the lock on the point in space to keep the camera pointed that direction. It seems if that can be done, the position in space could be given, and at least further refined on ground. 

I have a background in emergency preparedness and response, so this aspect of the hobby interest me quite a bit. I am relatively new here on diydrones. Is there a subgroup which discusses this topic, by chance?

Craig J., I could not find a sub group for this topic so I put it here. I too am new to this programing Lang.

As for gimbals doing the auto track, there are a bunch of them but most are too large to fit on some of these airframes never mind the integration aspect and most companies will claim proprietary software.

I figured the gimbals were heavy, and not open. My point, which I failed to make, is if its being done proprietary, it could be done open source. I think I implied pulling the data from a commercially available system. Maybe it can be done lighter and cheaper. 

Hi Jay, I tried this by first calculating distance to broken fence by using regular trig, you have height and downward angle therefore horizontal distance can be calculated. If height is 100 the downward angle was 45 degree ( easy one to check ) use tan ( 90-45 )*height. This ought to give 100 as well.

Then I use gps visualizer from a google search, I entered lat and long, distance and bearing, it then came up with a lat and long and even produced a google map.

Hope this helps, I expect all these calculations could be included within the osd by code, but that is beyond me. At least it can be done by paper by looking at the osd as long as the screen shows the downward angle of th camera. These calculation are only based on the fact the targets are on level ground.
Hi Jay, Have a look at this website, there is this guy from Sweden who has coded osd, maybe he can help you.
I think I have a handle on the math now can someone help me with the codeing part.

Here's where I am so far


Hi Jay,

Did you get any success with the Target Localisation issue?


Ashutosh Kumar

Reply to Discussion



Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service