Dani Unit Seminar: Brief overview of machine vision and assessment of a stereo vision algorithm performance

Dani Unit Seminar: Oliver Hamilton, Cranfield University

Title: A brief overview of machine vision and assessment of at stereo vision algorithm performance

Abstract: Machine vision is a rapidly advancing field. The reduction in size of cameras and the increase in processing power available mean machine vision can now be found on a variety of devices from phones to televisions to cars. There has been significant recent interest in stereo correspondence algorithms for use in the automotive environment. In this work we evaluate a range of dense stereo algorithms, using a unique evaluation criterion which provides quantitative analysis of accuracy against range, based on ground truth 3D annotated object information. This study extends prior work on dense stereo evaluation of variety of real-time stereo algorithms, from the very basic to the latest creations, using a novel quantitative metric relative to object range.