Configuring VLFeat in the MATLAB R2018b

Configuring VLFeat in the MATLAB R2018b

Author: Kailugaji - blog Park  http://www.cnblogs.com/kailugaji/

VLFeat official website: http://www.vlfeat.org

    VLFeat open source libraries focus on achieving the most current popular image understanding and computer vision algorithms to extract local features and matching. Fisher algorithms include vector, VLAD, SIFT, MSER, k-means, hierarchical k-means, gather information bottleneck, SLIC super-pixel, ultra-fast shifting pixels, large-scale SVM training. It is written in C, to improve efficiency and compatibility, and interfaces with MATLAB, easy to use, and provide detailed documentation throughout. It supports Windows, Mac OS X and Linux. The latest version of VLFeat is 0.9.21. Here explain VLFeat library configuration in MATLAB.

1. Download the latest version of VLFeat 0.9.21

2. extract to a path, such as: D: \ Program Files \ MATLAB \ vlfeat-0.9.21-bin

3. Open D: \ Program Files \ MATLAB \ R2018b \ toolbox \ local, startup.m new file, write run ( 'D: \ Program Files \ MATLAB \ vlfeat-0.9.21-bin \ toolbox \ vl_setup') and save

4. D: \ Program Files \ MATLAB \ R2018b \ toolbox in the entire local folder and all the files inside to the path

    Found in the current folder in the local folder, right-click the local folder and start the path to delete local folders and sub-folders, then right-click local, all added to the path, so startup.m file is added to the path the open MATLAB will automatically run the file. Of course, can also startup.m file into another position as long as the file can be added to the path.

5. 关闭MATLAB,再重新打开MATLAB,在命令行窗口输入vl_version,如果出现版本号,则说明配置成功

6. 注意

1). startup.m文件是用户自己定义的,MATLAB中没有的,要自己新建脚本文件,写进程序。

2). 关闭MATLAB后vl_version,如果提示未找到文件,多半是因为startup.m文件没有添加到路径,找到startup.m文件存放的文件夹,添加文件夹及其子文件到路径中即可。

3). VLFeat库文件可以存放到其他位置,相应地startup.m中的路径要改,startup.m文件也可以存放到其他位置,相应地startup.m所在的文件夹要添加到路径中。

4). VLFeat.org官网中提供了教程以及MATLAB API,方便查询。

Guess you like

Origin www.cnblogs.com/kailugaji/p/12127191.html