About this document
Monocular Video Obstacle Detection by Kishan Sachdeva is a document available to read on EtoBox.
This document proposes a method for real-time detection of obstacles from monocular video in a moving vehicle. It describes a two-step approach: 1) Estimating scene depth from the scaling of tracked image regions between frames to generate obstacle hypotheses. 2) Testing the hypotheses by analyzing the perspective transformation over time and comparing it to what would be expected for a free road vs. an obstacle. The key insight is that scene depth can be directly calculated from the scaling factor of image
- Author
- Kishan Sachdeva
- Language
- EN