Video Processing Techniques for Traffic Information Acquisition Using Uncontrolled Video Streams

摘要

This paper reports on the first steps taken in search of a solution that uses public video streams available on the Internet to address the increasing need for monitoring transportation networks with the intent of returning added value to the community, either by allowing a better understanding of the network and its needs or by feeding applications with real-time information for various purposes, such as simulation, decision-making support and updated route guidance. After the introduction of the field, we present our findings from a survey that briefly describes several works with related studies and explain the algorithms that can be adopted to get relevant information from video streams. This is followed by an analysis of the issues that may arise and the best ways to address them. Next it reports on the results achieved so far, draws some conclusions on what has been done and suggests the next steps of our research.