Aggregation of binary feature descriptors for compact scene model representation in large scale structure-from-motion applications

TL;DR

This paper introduces a method to aggregate binary feature descriptors into compact prototypes, significantly reducing memory usage and speeding up matching in large-scale structure-from-motion and SLAM applications.

Contribution

It proposes a novel aggregation technique that converts binary descriptors into low-dimensional real-valued prototypes for efficient large-scale 3D scene modeling.

Findings

Memory usage is significantly reduced.

Matching speed is improved using approximate nearest neighbor search.

The method is effective for large-scale structure-from-motion.

Abstract

In this paper we present an efficient method for aggregating binary feature descriptors to allow compact representation of 3D scene model in incremental structure-from-motion and SLAM applications. All feature descriptors linked with one 3D scene point or landmark are represented by a single low-dimensional real-valued vector called a \emph{prototype}. The method allows significant reduction of memory required to store and process feature descriptors in large-scale structure-from-motion applications. An efficient approximate nearest neighbours search methods suited for real-valued descriptors, such as FLANN, can be used on the resulting prototypes to speed up matching processed frames.