This project implements a few SLAM algorithms in two dimensions using rangefinder data for vehicle localization and for environment mapping. It is based on a Matlab implementation by Tim Bailey et al [OpenSLAM.org]. This is a work-in-progress port to Julia for fun and self-education, and (hopefully) eventual deployment on a real embedded system.
Current status: the extended Kalman Filter algorithm is written, and FastSLAM is ongoing.
A browser-based visualization of the EKF-SLAM simulation is runnable by launching wsserver.jl and navigating to localhost:8000. It depends on the HttpServer, WebSockets, and JSON packages, and makes use of D3.js for basic 2-D visualization.