/MatchBench

Evaluation code for "MatchBench: A System-level Benchmark and Evaluation for Image Matching"

Primary LanguageMATLAB

MatchBench: A System-level Benchmark and Evaluation for Image Matching

alt tag

Publication:

MatchBench: A System-level Benchmark and Evaluation for Image Matching, JiaWang Bian, Le Zhang, Yun Liu, Wen-Yan Lin, Ming-Ming Cheng and Ian D. Reid. [Project Page] [pdf(old)] [Dataset (Baidu Yun)] [Dataset (Google Drive)]

We make significant changes in the new version of the paper which will come soon.

Introduction

There exists a "ReadMe.txt" for describing the data and code in each folder.

Requirements:

1. Matlab (Evaluate results and plot figures)
2. OpenCV (Run feature matching algorithms)

Dataset:

1. Download the all subsets from the project page (https://jwbian.net/matchbench) and put them in the folder "Dataset".
2. The data format is described in "ReadMe.txt".

Algorithms:

1. Many feature matchers are provided.
2. The Matlab script for running algorithms are given.

Evaluation:

You can use the code in this folder to plot figures as presented in the paper.

If you like this work, please cite our paper

@inproceedings{bian2018matchbench,
    title={MatchBench: A System-level Benchmark and Evaluation for Image Matching},
    author={JiaWang Bian and Le Zhang and Yun Liu and Wen-Yan Lin and Ming-Ming Cheng and Ian D. Reid},
    year={2018}
}