2010 Fiscal Year Final Research Report
Tracking people in crowds and its application for behavior analysis
| Project/Area Number |
20300061
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Perception information processing/Intelligent robotics
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
SATO Yoichi The University of Tokyo, 生産技術研究所, 教授 (70302627)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SUGIMOTO Akihiro 国立情報学研究所, コンテンツ科学研究系, 教授 (30314256)
OKABE Takahiro 東京大学, 生産技術研究所, 助教 (00396904)
|
|---|
| Project Period (FY) |
2008 – 2010
|
| Keywords | 人物追跡 / 歩容特徴 / 特徴点軌跡 / 時系列フィルタ |
|---|
| Research Abstract |
Tracking people in a scene plays an essential role in human behavior analysis for various applications including security systems and marketing。However, the existing techniques for tracking people perform have difficulties when they are applied to crowded scenes。In this research project, we achieved robust tracking of multiple people even in crowds by introducing three novel ideas: integration of a sequential Bayesian filter and face/head detectors, clustering of trajectories of low level features, and tracking of a non-stationary sound source based on audio-visual correlation analysis。
|
