Performance Evaluation of Linear and Nonlinear Scale Descriptor Based Algorithms

Feature extraction and object matching is one of the important and significant task in many computer vision and robotic applications. Finding the corresponding points between the two images is very crucial in image analysis. The KAZE and SURF algorithms are found to perform well on feature matching....

Full description

Saved in:
Bibliographic Details
Published in2018 International Conference on Computing, Power and Communication Technologies (GUCON) pp. 561 - 565
Main Authors Vishwaraj, Vishwaraj, Ali, Layak
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2018
Subjects
Computational efficiency
Feature detection
Feature extraction
Gaussian noise
KAZE
linear scale
non linear scale
Performance evaluation
Speckle
Strain
SURF
Online AccessGet full text
DOI10.1109/GUCON.2018.8674950

Cover

More Information
Summary:Feature extraction and object matching is one of the important and significant task in many computer vision and robotic applications. Finding the corresponding points between the two images is very crucial in image analysis. The KAZE and SURF algorithms are found to perform well on feature matching. This paper investigates the performance evaluation of linear scale and nonlinear scale based algorithms namely KAZE and SURF respectively. The experimental results show that KAZE is more efficient than SURF but takes more computational resources.
DOI:10.1109/GUCON.2018.8674950