# CoMaL Tracking: Tracking Points at the Object Boundaries

## TL;DR

This paper introduces a novel, efficient tracking algorithm for CoMaL points that improves accuracy at object boundaries by matching level line segments to MSER segments, outperforming traditional methods.

## Contribution

The paper presents a new tracking method that leverages shape and texture matching of level line segments, enhancing boundary point tracking over existing re-detection frameworks.

## Key findings

- Improved tracking accuracy at object boundaries.
- Enhanced speed compared to re-detect-and-match methods.
- Outperforms KLT in boundary point tracking tasks.

## Abstract

Traditional point tracking algorithms such as the KLT use local 2D information aggregation for feature detection and tracking, due to which their performance degrades at the object boundaries that separate multiple objects. Recently, CoMaL Features have been proposed that handle such a case. However, they proposed a simple tracking framework where the points are re-detected in each frame and matched. This is inefficient and may also lose many points that are not re-detected in the next frame. We propose a novel tracking algorithm to accurately and efficiently track CoMaL points. For this, the level line segment associated with the CoMaL points is matched to MSER segments in the next frame using shape-based matching and the matches are further filtered using texture-based matching. Experiments show improvements over a simple re-detect-and-match framework as well as KLT in terms of speed/accuracy on different real-world applications, especially at the object boundaries.