Skip to content

An obstacle avoidance system for warehouse robots using an RGB camera, image segmentation, and CLF-CBF controls

Notifications You must be signed in to change notification settings

ctallum/S.A.M.I.A.M

Repository files navigation

S.A.M.I.A.M

This project was completed for Boston Unviversity EK505 as a Final Project.

Summary

The purpose of this project was to design a software system that would allow a warehouse robot with an overhead rgb camera to detect and drive around obstacles. We explored using Meta AI's Segment Anything Model and other segmentation algorithms to segment our live camera feed, from which we detect obstacles. We then use a CLF-CBF formulation to drive the robot around the obstacles to the goal point.

Demo

EK505.-.video1.mp4

About

An obstacle avoidance system for warehouse robots using an RGB camera, image segmentation, and CLF-CBF controls

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages