Brief Introduction

`mpiv' is a PIV (Particle Image Velocimetry) toolbox written in MATLAB. The main purpose of this toolbox is for educational and research. It is intended for both the undergraduate and graduate students working on PIV methods, or need to use PIV for their study. It is also a convenient and useful tool for engineers and scientists who use PIV in their research. Writing the program in MATLAB is, in our opinion, believed to be easier to understand and use, although it may be slower in computation. The project started in Fall 2002, and since then several algorithms have been modified and added. Moreover several options have be included for users of different levels. mpiv has also been tested and proved to be sufficient accuracy and robust.

There are many PIV programs existing today. Some of the programs are written in MATLAB and already available to users (e.g., URAPIV and MatPIV). These programs have excellent capability for research propose and have many options for users of different level and optimization for fast computation. In comparison, mpiv is more general and portable, and easy to use and modify. It is the developers' intention to keep the codes short and simple. In brief, mpiv consists of two main programs (one for image processing and the other for post-processing) and several external functions (.m files) while most of them are very short and are based on simple algorithms. Thus, not only are the algorithms and codes of mpiv easy to understand for undergraduate and graduate students, they are also quite general and easy to modify for use in specific research area for engineers and scientists.

The present version of mpiv has the following features:

• Cross-correlation and MQD (Minimum Quadratic Difference)
• Super-resolution (recursive/hierarchical) algorithm
• Local mean, median and global filtering
• Vector interpolation using linear, spline and Kriging methods
• Simple GUI menu is available
• Free of charge

To run the program, you need the followings:

• MATLAB version 6 (Release 12) or later
• MATLAB Image Processing Toolbox
• DACE (Design and Analysis of Computer Experiments) Toolbox

Note that the earlier versions of MATLAB may work fine as well though we did not test the program on them, and MATLAB Image Processing and DACE Toolbox are necessary to run the program. Image Processing Toolbox is the one of Toolbox developed by Mathworks, while DACE Toolbox (see next section for more details) is a GPL based toolbox and can be downloaded from http://www.imm.dtu.dk/$\tilde{\ }$hbn/dace/. In addition, you can use MATLAB Compiler to reduce computational time, but it is not a must. The original images in this document are taken from VSJ-PIV (the PIV Standard Project) distributed by the Visualization Society of Japan (http://www.vsj.or.jp/).

mpiv has been tested (MATLAB 6.5.0.180913a, R13) on Linux platform running on Vine Linux 2.6 (http://www.vinelinux.org/), equivalent to Red Hat Linux 7.2, and Microsoft Windows 2000.