LD06: Record only 2deg slices #27045
Conversation
mw46d
commented
May 12, 2024
mw46d
commented
- Some recording changes were taken from the RPLidar driver code.
- Using the average of the 12 readings/package is not a good idea. Those packages don't describe the same 'angle' for each rotation. So you end up with a lot of different distances per package. I tried to fix things to well-defined 2deg slices and take the minimum reading reported.
There was a problem hiding this comment.
OK, this improves the driver out-of-sight - thanks for this!
I've just tested this on an LD06 locally, and it makes the driver actually usable!
I've taken the liberty of adjusting a few patches in here to conform to our style guidelines, and adjusted the message on your commit to also conform to our requirements.
I've also adjusted the handling of a bad size received to avoid the nasty buffer overflow, rebasing on a previous PR of mine.
Thanks!
@lvale you said you had some of these. I've no idea how they're at all useful for you before this PR :-)
incorrect distances+angles being returned
the length field is actually the count of 3-byte data elements
|
I've created a simulator for this device now. And corrected the sanity check again. |
CraigElder
changed the title
| const float degrees_per_ms = degrees_per_s / 1000.0f; | ||
| const float degrees_per_sample = degrees_per_ms / samples_per_ms; | ||
|
|
||
| // ::fprintf(stderr, "Packing %u samples in for %ums interval (%f degrees/sample)\n", sample_count, time_delta_ms, degrees_per_sample); |
There was a problem hiding this comment.
let's get rid of this debug
|
@Adithya-Patil, are you OK with this? Any comments on the orientation? The wiki docs say the wire should be facing backwards |
That's how my sensor is oriented as well;-) |
|
Hi @mw46d, Ah, so you've got one of these sensors? Do you find that it doesn't actually work with the code in 4.5, 4.6 and/or master? |
|
That's why I started that PR;-) Some time ago |
