VocMatch: Efficient Multiview Correspondence for Structure from Motion

  title={VocMatch: Efficient Multiview Correspondence for Structure from Motion},
  author={Michal Havlena and Konrad Schindler},
Feature matching between pairs of images is a main bottleneck of structure-from-motion computation from large, unordered image sets. [] Key Result The proposed vocabulary-based matching has been integrated into a standard SfM pipeline, and delivers results similar to those of the conventional method in much less time.

PAIGE: PAirwise Image Geometry Encoding for improved efficiency in Structure-from-Motion

A learning-based approach is proposed, the PAirwise Image Geometry Encoding (PAIGE), to efficiently identify image pairs with scene overlap without the need to perform exhaustive putative matching and geometric verification.


To achieve high accuracy and robustness, image triplets are employed and (an approximate) camera calibration is assumed to be known and additional links are determined and employed to improve the accuracy of the pose estimation.

Progressive Large-Scale Structure-from-Motion with Orthogonal MSTs

A progressive SfM method to tackle the completeness, robustness and efficiency problems in a united framework, where two loops are contained, and finds both loops converge fast and a large number of redundant pairs are excluded.

Efficient Covisibility-based Image Matching for Large-Scale SfM

A novel efficient image matching method by using the transitivity of region covisibility to find overlapping image pairs that can be efficiently found in an iterative matching strategy even only with few inlier feauture matches.

An efficient method to detect mutual overlap of a large set of unordered images for structure-from-motion

A new method which was inspired by and improves upon methods employing random k-d forests for low-cost 3D reconstruction based on images is proposed, which first derive features from the images and then a random forest is used to find the nearest neighbours in feature space.

Recent Developments in Large-scale Tie-point Search

Recent developments in this field are reviewed, which make it possible to generate tie-points and reconstruct unordered image sets with thousands of images.

A Fast and Robust Large-Scale Structure from Motion Using Auxiliary Information

  • Wenxiang DuYao LeeYue Qi
  • Computer Science
    2017 International Conference on Virtual Reality and Visualization (ICVRV)
  • 2017
The paper will use the GPS data to select potential matching images, and some distant image pairs will be discarded directly without calculating feature matching to greatly accelerate the reconstruction process.

Hyperpoints and Fine Vocabularies for Large-Scale Location Recognition

An orthogonal strategy is explored, which uses all the 3D points and standard sampling, but performs feature matching implicitly, by quantization into a fine vocabulary, and achieves state-of-the-art performance, while the memory footprint is greatly reduced, since only visual word labels but no 3D point descriptors need to be stored.

Line3D: Efficient 3D Scene Abstraction for the Built Environment

This work uses purely geometric constraints to match 2D line segments from different images, and forms the reconstruction procedure as a graph-clustering problem, which generates accurate 3D models, with a low computational overhead compared to SfM alone.



Randomized structure from motion based on atomic 3D models from camera triplets

Using three views instead of two allows us to reveal most of the outliers of pairwise geometries at an early stage of the process hindering them from derogating the quality of the resulting 3D structure at later stages.

Active Matching

This paper shows that the dramatically different approach of using priors dynamically to guide a feature by feature matching search can achieve global matching with much fewer image processing operations and lower overall computational cost.

Fast image-based localization using direct 2D-to-3D matching

This paper derives a direct matching framework based on visual vocabulary quantization and a prioritized correspondence search that efficiently handles large datasets and outperforms current state-of-the-art methods.

Optimal Reduction of Large Image Databases for Location Recognition

Experiments with several image sets show that the greedy solution already performs remarkably well, and that the optimal solution achieves roughly 5% smaller key frame sets which perform equally well in location recognition and SfM tasks.

Efficient representation of local geometry for large scale object retrieval

This work proposes a novel method for learning discretized local geometry representation based on minimization of average reprojection error in the space of ellipses and shows that if the gravity vector assumption is used consistently from the feature description to spatial verification, it improves retrieval performance and decreases the memory footprint.

Building Rome in a day

A system that can match and reconstruct 3D scenes from extremely large collections of photographs such as those found by searching for a given city on Internet photo sharing sites and is designed to scale gracefully with both the size of the problem and the amount of available computation.

Visual Modeling with a Hand-Held Camera

A complete system to build visual models from camera images is presented and a combined approach with view-dependent geometry and texture is presented, as an application fusion of real and virtual scenes is also shown.

Object retrieval with large vocabularies and fast spatial matching

To improve query performance, this work adds an efficient spatial verification stage to re-rank the results returned from the bag-of-words model and shows that this consistently improves search quality, though by less of a margin when the visual vocabulary is large.

Location Recognition Using Prioritized Feature Matching

This work devise an adaptive, prioritized algorithm for matching a representative set of SIFT features covering a large scene to a query image for efficient localization, based on considering features in the scene database, and matching them to query image features, as opposed to more conventional methods that match image features to visual words or database features.

Modeling and Recognition of Landmark Image Collections Using Iconic Scene Graphs

This article presents an approach for modeling landmarks based on large-scale, heavily contaminated image collections gathered from the Internet. Our system efficiently combines 2D appearance and 3D