 |
 |
|||
|
||||
|
|
||||
| Title: | Intelligent Surveillance and Pose-invariant 2D Face Classification | |
| Source: | MACHINE INTERPRETATION OF PATTERNS Image Analysis and Data Mining (pp 207-229) |
|
| Author(s): | Brian C. Lovell
NICTA, 300 Adelaide St, Brisbane, QLD 4000, Australia SAS, ITEE, University of Queensland, Brisbane, QLD 4072, Australia Conrad Sanderson NICTA, 300 Adelaide St, Brisbane, QLD 4000, Australia Ting Shan NICTA, 300 Adelaide St, Brisbane, QLD 4000, Australia |
|
| Abstract: | We describe recent advances in a project being undertaken to trial and develop advanced surveillance systems for public safety. One goal of the project is to trial commercial technologies in public spaces to evaluate their performance. Another is to develop and trial enhanced capabilities that will lead to more effective surveillance systems. A key technology being developed within the group is reliable face in the crowd identification from conventional CCTV cameras. While this is acknowledged to be a challenging problem, we have made considerable progress on several fundamental issues including recognition robust to large pose angles. We also describe a reconfigurable smart camera we are developing to handle the problem of obtaining high resolution face images while simultaneously surveilling large crowds in real-time. | |
| Full Text: | View full text in PDF format (5461KB) | |
| TOC: | Back to Table of Contents | |
|
||