Skip to content

Latest commit

 

History

History
18 lines (15 loc) · 1.43 KB

README.md

File metadata and controls

18 lines (15 loc) · 1.43 KB

Palmprint_Recognition

This is the matlab implementation for BLPOC, Location based matching and DoG for reference points selection, the dataset used in this implementation is CASIA Palmprint database. For more detailed information about this techniques applied here, please refer to this paper: Performace Improvement of Phase-based Correspondence Matching for Palmprint Recognition.

Notice: Database prerocessing (ROI cropping, rotation, normalization) is not in this implementation.

Getting started

You can start this code by typing:

[dis, blpoc] = matching('filename',mode)

filename is actually a txt file, filled with lines of image paths in the following forms path\to\image\xxx0.jpg path\to\image\xxx1.jpg with a space in between representing two images that are gonna be matched together. mode specifies the approaches applied. 0 means location based method and 1 means applying DoG for reference points selection. and this will return two vectors dis the score gained from the locations and blpoc for the average for BLPOC values of each corresponding image patches.

##Future work Since this is the only starting version of this implementation of the paper, further reconstruction and revison will be done.