No description, website, or topics provided.
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Output
pics
.~lock.recordspdsht.xlsx#
Path Oriented Powered Wheelchair Navigation Assistance_Jason Dekarske.pdf
README.md
Raspi_mount.SLDPRT
Raspi_mount.STEP
live_sidewalk_recognize.py
record.txt
sidewalkrecognize.py
videowatcher.py
vsidewalkrecognize.py

README.md

Path-Oriented-Electric-Wheelchair-Steering-Assistance

  • See abstract for research background and see TJOE for full research paper.

Starting from scratch.

  • See the bill of materials in appendix of paper for parts used in this project.
  • The "backup image" file can be flashed to an sd card using this tutorial or your favorite method.

Adding to an existing system.

Using the scripts.

  • Add run command sudo python <script directory> & to etc/profile, use ampersand to run script in background.
  • Use sudo raspi-config and switch to "auto login command line" or errors will occur!
  • Use live_sidewalk_recogize.py for the intended application described by the research documents.
  • Use sidewalkrecognize.py to analyze images and adjust range parameters.
  • Use videowatcher.py to see frame by frame if omxplayer isn't working.
  • Use vsidewalkrecognize.py to analyze videofiles.

TODO (live_sidewalk_recognize.py):

  • Find more graceful way to use onbuttonpress().
  • Add cliff detection. When the centroid falls below a certain vertical coordinate.
  • Add shadow detection. See article: "Shadow Detection and Removal from a Single Image Using LAB Color Space"
  • Improve framerate/optimize.