Detection and extraction features for signatures images via different techniques
Signature is one of the most important features to identify individuals. It represents a specific mark that includes handwritten characters or symbols. Also, signing takes place in a wide range of businesses, such as bank transactions and government documents so it provides a good way to maintain...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2019
|
| Subjects: | |
| Online Access: | http://eprints.unisza.edu.my/2586/1/FH03-FIK-19-28018.pdf http://eprints.unisza.edu.my/2586/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Signature is one of the most important features to identify individuals. It represents a specific mark that includes
handwritten characters or symbols. Also, signing takes place in a wide range of businesses, such as bank transactions
and government documents so it provides a good way to maintain security, in biometric systems. Signature is used as
a feature to identify the user by extracting a set of features. Over time, a number of techniques have been developed to
identify and extract a set of features from the signature image. Although there are many of these techniques, there is a
set of elements that determines the feasibility of using a particular technique, such as accuracy, computational
complexity, and the time needed to extract features. In this paper, three widely used feature detection algorithms,
SURF, BRISK and FAST, these algorithms are compared to calculate the processing time and accuracy for set of
signatures correctly. Three techniques have been applied using (UTSig) dataset; the results showed that the BRISK
algorithm got the best result among the feature detection algorithm in terms of accuracy and the FAST algorithm got
the best result among the feature detection algorithm in terms of run time. |
|---|