Description

Category

Computer Vision, Deep Learning, Image Processing

Scope

Two students completing a Master's Thesis (30 hp each).

Background

At Axis Communications, we are pushing the boundaries of surveillance technology beyond traditional video. As part of this effort, we are exploring how 3D sensors can be used in static surveillance setups – think traffic intersections or perimeter monitoring.
 

A key challenge in this domain is to get actionable insights from the raw sensor data: separating relevant events from the static environment and correctly classifying different objects, with the end goal of having a complete analytics pipeline for automated alarm decisions. While such perception problems are well-studied for 2D video, they remain less explored for static 3D surveillance sensors.
 

This thesis will investigate a specific topic within the 3D sensor analytics pipeline – pushing th...