What question did this study set out to answer?

This research aims to enhance feature tracking by leveraging temporal context through a novel model.

March 14, 20261 citationsOpen Access

KLT Revisited: Spatio-Temporal Affine Regression for Feature Tracking

NDNigel Joseph Bandeira DiasUniversidade Federal de Goiás GLGustavo Teodoro LaureanoUniversidade Federal de Goiás RCRonaldo Martins Da Costa

Key Points

This research aims to enhance feature tracking by leveraging temporal context through a novel model.
Proposed a differentiable model based on the KLT feature tracker.
Developed a fully convolutional neural network for spatio-temporal feature learning.
Utilized experimental comparisons to evaluate performance against traditional methods.
The proposed method shows superior performance in feature tracking compared to the KLT method.
Improvements are particularly notable in challenging environments.

Abstract

Feature association is an essential step for vision-based localization methods. These methods rely on feature matching to estimate relative motion between consecutive frames using projective geometry. Regardless of the advances in feature association, most existing methods still rely on pairwise feature matching approaches and ignore the rich temporal context in image sequences. In this paper, we revisit the well-known Kanade-Lucas-Tomasi (KLT) feature tracker algorithm and propose a differentiable tracker model. The Proposed method is a fully convolutional neural network that learns spatio-temporal features to track keypoints across videos. Experimental results show that the proposed method outperforms the KLT method, especially in challenging environments.

Perguntar à IA

Bookmark

View Full Paper