Plenary Lecture

Novel Human Detection and Tracking Solutions

Professor Tudor Barbu
Institute of Computer Science
Romanian Academy
Iasi, Romania

Abstract: Human detection and tracking represents a high-interest computer vision research field. It is the most important sub-class of the object detection and tracking domain and it has received a considerable interest over the last decade. Significant research has been devoted to detecting, locating and tracking people in digital images and videos, since numerous applications involve humans' locations and movements.
Person detection identifies the human presence in static images and videos, differentiating humans from non-human objects. Human tracking locates the instances of each detected person in the frames of the analyzed movie. Detecting humans in images and videos represents a challenging task, being complicated by numerous factors, such as: variable people appearance, camera position, wide range of poses adopted by persons, variations in brightness, illumination, contrast levels or backgrounds, and person occlusions.
Person detection domain includes important computer vision sub-domains, such as face detection or pedestrian detection. Numerous detection techniques have been developed in recent years. Thus, face detection has been approached through skin-based techniques and Boosting algorithms using Haar features. Pedestrian detection methods include those based on Histogram of Oriented Gradients (HOGs), Partial Least Square Analysis (PLSA) or BOOST algorithms. The most important tracking techniques are based on Kalman filtering, Active Contours and template matching. Pedestrian detection and tracking has important application areas, such as robotics, video surveillance and urban trafic monitoring.
We have proposed human detection and tracking solutions that outperform state-ofthe- art techniques. First, we developed robust human skin detection approaches, which were then used for both face and pedestrian detection processes. Our face detection techniques identifies those skin segments representing faces by applying some morphological operations or correlation procedures on them and checking if some certain conditions are met.
Also, novel pedestrian detection and tracking techniques that work successfully for static-camera video sequences are proposed. They use temporal-differencing algorithms for video object detection. Then, the identified objects are classified as human or non-human by using the previously detected skin regions and a set of conditions related to human body characteristics. Person tracking is performed through a template matching process applied to the detected human objects. The human matching process may use HOG-based features or normalized correlation-based procedures.

Brief Biography of the Speaker: Dr. Tudor Barbu is currently Senior Researcher II at the Institute of Computer Science of the Romanian Academy, in Iasi, Romania. He is the coordinator of the Image and Video Processing and Analysis research collective of the institute and also member of the leading Scientific Council of this institute. Mr. Barbu has a PhD degree in Computer Science, awarded by the Faculty of Automatic Control and Computers of the University “Politehnica” of Bucharest.
He has a remarkable research profile. In the last decade he published two books and four book chapters as single or main author. Also, Dr. Tudor Barbu published more than 70 articles in prestigious international journals and volumes of international scientific events (conferences, symposiums and workshops). His prolific scientific activity also includes more than 36 research reports, elaborated with the institute research team coordinated by him or related to various research projects. His scientific publications have got over 80 citations, according to Google- Academic. He was Invited Plenary Speaker at numerous international conferences.
In recent years he also coordinated various research directions in 6 projects based on contracts/grants. Dr. Tudor Barbu received also several awards for his research results, the most important being the Romanian Academy Prize “Gheorghe Cartianu”, in the Information Science and Technology domain, awarded on December 18, 2008. He is member of several conference scientific committees and also member of scientific and technical committee and editorial review boards of some journals. He is the Editor in Chief of a book. His main scientific areas of interest are: digital media (audio, video and image) signal processing and analysis, pattern recognition, computer vision, multimedia information storage, indexing and retrieval, partial differential equations (PDE) and biometric authentication using voice, face and digital fingerprint recognition.