About this document
Incremental Object Tracking Across Cameras by Lihi Alter is a document available to read on EtoBox.
This paper presents an approach for tracking objects across multiple uncalibrated and non-overlapping cameras. The technique uses incremental learning to model both the color variations between cameras and the probability distributions of spatio-temporal links between camera views. It learns these relationships over time without requiring pre-calibration or batch processing. As more data is accumulated, the accuracy of tracking objects across camera views improves.
- Author
- Lihi Alter
- Language
- EN