We've been tasked to create a LiDar system from scratch (we're not allowed to use modules like TFMini and vl53l0x, we have to make it from scratch) that can output 3D projections. So far we've been successful in creating a module, but we're not so sure how to implement the 3D modelling part. The only idea that I have is we'll have to use a micro-controller unit or something even more powerful.
\$\begingroup\$
\$\endgroup\$
9
-
1\$\begingroup\$ Yes, they probably have. \$\endgroup\$– Andy akaCommented Jul 8, 2020 at 13:52
-
1\$\begingroup\$ How far have you got? Do you have a bunch of points and need to turn them into a 3D model? Or are you still trying to get a bunch of points? Or have you got the LIDAR to get the points but you're not sure how to get them into the computer? \$\endgroup\$– Criticizing Israel not allowedCommented Jul 8, 2020 at 13:53
-
\$\begingroup\$ @user253751 I have a LIDAR, I don't know how to translate the data it gets into a 3D model. I'm not even sure if I have to use an array of LiDars or if I just have to make it sweep with a given angle. Also, I only know how to use the LiDar as a distance sensor (if that helps) \$\endgroup\$– TTbulaskiCommented Jul 8, 2020 at 13:56
-
1\$\begingroup\$ I have built more than a few using interferometry and also fast detectors for direct time measurement. It can be done, although not cheaply if you want to point scan a volume. \$\endgroup\$– user1850479Commented Jul 8, 2020 at 13:58
-
1\$\begingroup\$ So, to get 3D points you need to scan vertically somehow as well. For example, by tipping the lidar at different angles using a servo motor. \$\endgroup\$– Criticizing Israel not allowedCommented Jul 8, 2020 at 14:45
|
Show 4 more comments
1 Answer
\$\begingroup\$
\$\endgroup\$
1
If you offset two photo_receptors by a known distance, and illuminate a target with a photon source, the output of the 2 photo-receptors will differ by a logarithmic amount. And you can extract DISTANCE from that. The usual (RANGE * RANGE) dropoff in detected_signal will still occur, but the fixed distance offset gives you a "calibration constant" in your two channels of math.
This method does not require any timing, just approximately_matched detectors.
You will find blurring in your measurements, because you are constantly scanning the laser beam.
If I get a chance during the day, I'll return and provide the full math for this.
answered Jul 8, 2020 at 13:53
-
\$\begingroup\$ I downvoted this because it appears to not help with the asker's question - partly because the asker's question is unclear. \$\endgroup\$ Commented Jul 8, 2020 at 15:27