Probabilistic Occupancy Grid Generation from Lidar Data
Objectives
To understand how to use lidar data to generate occupancy probabilities
Apply logodds updates based on given Bresenham raytracing outputs
Convert a logodds grid to a probabilistic occupancy grid
You are provided with lidar measurement data (Measurements.txt), as well as a Python 3 Jupyter Notebook which contains the required supplementary code (thanks to Paul Balzer). Your task is to complete each TODO section of the notebook in order to generate a probabilistic occupancy grid, as well as answer the given written questions.
Resources and Instructions
There are 3 TODO sections to complete in the given Jupyter notebook:
Write code to convert lidar data in spherical coordinates to Cartesian coordinates in the function ibeo2XYZ().
Perform the logodds update for the `grid` global variable in insertPointcloudBRESENHAM(). Make sure to complete both TODOs in this section.
Convert the logodds grid to a probabilistic occupancy grid.
In addition, there are two written questions you must answer:
What are the computational advantages of using logodds when generating our occupancy grid?
Is the angle phi in our Spherical to Cartesian calculation the same as the polar angle in standard Spherical coordinates? Why?
Deliverables
HTML output: In the Jupyter Notebook, go to File > Download as > HTML (.html).
In addition, include a PDF file of your answers to the two written questions. Submit a ZIP file containing the HTML output and the PDF file.
Run all code blocks before downloading the HTML.
Please follow the naming convention for your zip file: a5_<user_id>.zip .
