ArduPilot / ArduPilot/OctomapPlanner

Question regarding Octomap & Planner

Open
#9 7 comments 0 reactions 0 assignees View on GitHub
Dominant language
C++
Stars
163
Forks
43
PR merge metrics
No merged PRs in 30d

Description

Hello,
I have set up cluttered environment and playing with this repository.
What I have set-up here is world with walls, bars and trees are up ahead but found not satisfactory result.
If planner works properly, drone can fly almost straight ahead with some small course correction

![screenshot from 2019-01-14 15-14-45](https://user-images.githubusercontent.com/13569579/51099969-270d2c80-1817-11e9-8ec7-d0a57ba9b55f.png)
![image](https://user-images.githubusercontent.com/13569579/51100289-b535e280-1818-11e9-9ed4-0385d63b7524.png)

I have managed StereoSGBM parameter little bit tuned.
```
PreFilterCap: 32
sgbmWinSize: 7
MinDisparity: 0
UniquenessRatio: 65
SpeckleWindowSize: 500
SpeckleRange: 1
Disp12MaxDiff: 1
TextureThreshold: 0
Width: 640
Height: 480
Scale: 1
```

In disparity map, I can clearly see obstacles with some noise, but Octomap result is quite different.
It is mapped as some radial way regarding walls in left & right front.
![disparity_screenshot_14 01 2019](https://user-images.githubusercontent.com/13569579/51101260-49a24400-181d-11e9-8c5c-e66ea08f8249.png)

[map.bt.tar.gz](https://github.com/ArduPilot/OctomapPlanner/files/2754247/map.bt.tar.gz)

And also I'm wondering that does InformedRRT* check how much space vehicle does occupy.
It plans like it tries to move through the wall if it finds small space in thick wall
and does not avoid bars so far.
However, it does pause in front of bars for few seconds then goes straight head into.

I'd like to know more about these issues so maybe I can help to improve.
Thank you.

Contributor guide

No contributing guide indexed for this repository

Research direction

No source file or test is identified. Start by reproducing the reported Octomap output with the attached map and StereoSGBM settings, then inspect the mapping and InformedRRT* behavior; done means determining why obstacle geometry and vehicle clearance are incorrect and documenting or correcting the confirmed cause.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp, opencv
Domain
computer-vision, robotics
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.