PyData Berlin
Lane detection in self-driving using only NumPy
Talk
Abstract
Are you a scientist or a developer looking to understand how to use NumPy to solve computer vision problems? NumPy is a Python package that provides the multidimensional array object which you can use to solve the lane detection problem in computer vision for self-driving cars or autonomous driving. You can apply non-machine learning techniques using NumPy to find the straight lines on street images. No other external libraries, just python with NumPy.
Prerequisites
Basic knowledge of Python and NumPy.
Description
Car accidents happen every day. Lane detection is a common use case in computer vision and self-driving which can help prevent accidents by identifying the boundaries of driving lanes on an image of a road scene. It is a fundamental requirement for autonomous vehicles to drive safely on the roads.
In my talk, “Lane detection in self-driving using only numpy”, I will share my learning experience on how to solve a computer vision problem with a small dataset.
Outline of my talk: 25-27 minutes
2 minutes - the problem
5 minutes - What is NumPy and the steps of the algorithm, & the methods (Create Arrays using NumPy functions, Array slicing, indexing, math using NumPy Broadcasting)
10 minutes - edge detection algorithms
10 minutes - ROI extraction, hough transform and detecting the lines
Speaker
Emma Saroyan
Emma Saroyan is a developer advocate and an author. She has worked in startups at the intersection of technology and education, and is a passionate community builder within the open source ecosystem. Emma is running a local Postgres User Group in Yerevan, Armenia, where she organizes meetups and conferences. She currently serves on Diversity & Inclusion Committees of PgUS and NumFOCUS. She enjoys giving talks, and sharing her knowledge at tech conferences.