Multilateration improvments #164
Conversation
ghost
assigned 
| self.vec_samples = deque(maxlen=buffer_len) | ||
| self.enabled = False | ||
| self.global_fame = 'enu' | ||
| self.last_pos = None | ||
| self.global_frame = 'enu' |
There was a problem hiding this comment.
As this will be a mil_common script (likely to be used on Subjugator), the global frame should also be a param (Subjugator uses "map")
| if distance1 < self.filter_dist or distance2 < self.filter_dist: | ||
| line_intersections += 1 | ||
|
|
||
| if line_intersections < self.filter_dist: |
There was a problem hiding this comment.
Should this be self.filter_count?
There was a problem hiding this comment.
Good catch! I've fixed this now. I guess it passed my testing since the default values are close.
| line_array = np.array([[0, 0, 0, 0]]) | ||
| line_intersecting_array = np.array([]) | ||
|
|
||
| for line1 in lines_array: |
There was a problem hiding this comment.
Comment noting that this section filters out nearby intersections
|
|
||
| # If no point could be determined, don't publish | ||
| if where is None: | ||
| return |
There was a problem hiding this comment.
This is a nice place to do a rospy.logwarn_throttle(10.0, "could not ..."). So if nothing gets published the logs show why
|
Don't worry about it for now (there are more important things before competition), but I ticketed #165 to eventually have CI test this code |
- Generalized global frame - Bug with filter - Additional documentation - Additional feedback
|
Requested changes have been made. Good catch on the filter_count vs filter_dist thing. |
closes #148