Locally Orderless Tracking
- 1 June 2012
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 10636919,p. 1940-1947
Abstract
Locally Orderless Tracking (LOT) is a visual tracking algorithm that automatically estimates the amount of local (dis)order in the object. This lets the tracker specialize in both rigid and deformable objects on-line and with no prior assumptions. We provide a probabilistic model of the object variations over time. The model is implemented using the Earth Mover's Distance (EMD) with two parameters that control the cost of moving pixels and changing their color. We adjust these costs on-line during tracking to account for the amount of local (dis)order in the object. We show LOT's tracking capabilities on challenging video sequences, both commonly used and new, demonstrating performance comparable to state-of-the-art methods.Keywords
This publication has 19 references indexed in Scilit:
- Earth Mover Distance on superpixels2010
- PROST: Parallel robust online simple tracking2010
- Object trackingACM Computing Surveys, 2006
- Real-Time Tracking via On-line Boosting2006
- Geometric context from a single image2005
- Probabilistic tracking in joint feature-spatial spaces2003
- Learning a classification model for segmentation2003
- The Earth Mover's distance is the Mallows distance: some insights from statistics2002
- The Earth Mover's Distance as a Metric for Image RetrievalInternational Journal of Computer Vision, 2000
- A unified approach to the change of resolution: space and gray-levelIEEE Transactions on Pattern Analysis and Machine Intelligence, 1989