Ask Your Question
0

Long Distance object detection

asked 2019-01-09 19:59:21 -0500

I want to detect objects - any object at long distances. To do so, I would like to use several high-powered lasers, with each beam separated by a degree or two. I would flash the lasers and turn them on ONLY on every other video frame (I would synchronize them with the camera). The camera would recognize that it was MY lasers by seeing the laser "dot" or "dots" only on every other frame. Of course, these dots would move in the video field a bit between frames, but not a large amount. I have no idea how to do this in OpenCV, and wonder if it is possible, and if so - how.

Manually using high-powered green and red lasers, I can easily see reflections from trees and houses in broad daylight that are hundreds of feet away. Of course, if the lasers are aimed at open space, there is no reflection. (unless the air is very dirty or foggy). I don't care how far away the object is, only if there is any object at all.

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted
0

answered 2019-01-09 21:58:53 -0500

Tetragramm gravatar image

There's a couple of questions in your question. So let me know if I got everything.

The camera synchronization is not really an OpenCV thing. If it is possible through OpenCV, it's almost certainly easier through the camera API.

Looks like you've got a plan for identifying the dots, so that's good.

For detecting the dots, step 1 is manual. Point laser at object, take picture. Can you see the dot in the picture? If the answer is no, you'll need to improve things. Better sensitivity, brighter laser, narrow filter over the lens, ect.

Then, take a couple of sample pictures to see if you can detect the dots automatically. Lots of ways to do it, so it's up to you to figure out which is best for your use case. It's better to take some sample images, try the simple stuff like thresholding and color filtering (try the more complicated stuff too, if you can), and then post another question with images so we can actually help.

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

1 follower

Stats

Asked: 2019-01-09 19:59:21 -0500

Seen: 31 times

Last updated: Jan 09