@inproceedings{cbbfbb016eab43059e36cb58d5c241ad,
title = "Real time semi-dense point tracking",
abstract = "This paper presents a new algorithm to track a high number of points in a video sequence in real-time. We propose a fast keypoint detector, used to create new particles, and an associated multiscale descriptor (feature) used to match the particles from one frame to the next. The tracking algorithm updates for each particle a series of appearance and kinematic states, that are temporally filtered. It is robust to hand held camera accelerations thanks to a coarse-to-fine dominant movement estimation. Each step is designed to reach the maximal level of data parallelism, to target the most common parallel platforms. Using graphics processing unit, our current implementation handles 10 000 points per frame at 55 frames-per-second on 640 ×480 videos.",
author = "Matthieu Garrigues and Antoine Manzanera",
year = "2012",
month = jan,
day = "1",
doi = "10.1007/978-3-642-31295-3\_29",
language = "English",
isbn = "9783642312946",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
number = "PART 1",
pages = "245--252",
booktitle = "Image Analysis and Recognition - 9th International Conference, ICIAR 2012, Proceedings",
edition = "PART 1",
note = "9th International Conference on Image Analysis and Recognition, ICIAR 2012 ; Conference date: 25-06-2012 Through 27-06-2012",
}