IP - 1 Based Video Analysis for Traffic Flow Estimation Based on Machine learning Technologies

NICTA Project - Dr. Jing Chen (Jing.Chen@nicta.com.au)

Recent research in content based video analysis has explored the usage of machining learning technologies to achieve the semantic understanding at video sequence. In this project, you will get opportunity to participate in some research activities at the Sensor & Survelliant component of the NICTA STaR project. This project will involve video low-level feature extraction, feature expression and modelling, and feature classification and recognition. You will have a chance to learn those advanced machine learning technologies, such as K-Mean/C-Mean, HMM, CRM, and Mean-Shift etc, and further to use some of them in the project to auto-recognize the traffic situation at an intersection or freeway point. You will be working closely with those NICTA research staff at this project to develop novel algorithm in image process and Patten recognition. An expected outcome of the research is the implementation and setup of a demo system.

For further information, please contact the supervisor or consider the following:

Fatih M. Porikli and Xiaokun Li. Traffic Congestion Estimation Using HMM Models Without Vehicle Tracking. IEEE Proc. Intelligent Vehicles. Dec 2004.

Xiaokun Li and Fatih M. Porikli. A Hidden Markov Model Framework for Traffic Event Detection Using Video Features. IEEE Int. Conf. on Image Processing. Dec 2004.

IP - 2 Intelligent Traffic Dog

NICTA Project - Dr. Jing Chen (Jing.Chen@nicta.com.au)

Recent research in content based video analysis has explored the usage of machining learning technologies to achieve the semantic understanding at video sequence. In this project, you will be involved in researching and developing an intelligent traffic monitoring system, namely traffic DOG, to auto-estimate the traffic flow at a free way or intersection sport. The project will involve semi-automatic classification at the traffic flows under all weather conditions based on advanced machine learning technologies. The project requires to develop an algorithm for running the system based on a HMM method. You will be working closely with those NICTA research staffs and get some solid knowledge and experience in computer vision and multimedia technologies at the end of this project. An expected outcome of the research is the implementation and setup of a demo system.

For further information, please contact the supervisor or consider the following:

Fatih M. Porikli and Xiaokun Li. Traffic Congestion Estimation Using HMM Models Without Vehicle Tracking. IEEE Proc. Intelligent Vehicles. Dec 2004.

Xiaokun Li and Fatih M. Porikli. A Hidden Markov Model Framework for Traffic Event Detection Using Video Features. IEEE Int. Conf. on Image Processing. Dec 2004.

[Top of Page]