Open AccessLemp: a Robust Image Feature Descriptor for Retrieval Applications
Author(s) -
Lanxiang Rao,
P Rohni,
M Narayana
Publication year - 2019
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.a1821.109119
Subject(s) - enhanced data rates for gsm evolution , pixel , feature (linguistics) , artificial intelligence , computer science , line (geometry) , pattern recognition (psychology) , image (mathematics) , precision and recall , computer vision , mathematics , philosophy , linguistics , geometry
line edge magnitude pattern (lemp) is proposed in this paper. Line edge distribution is used to denote local region of an image. Popular texture descriptors such as lbp deal with a comparison of centre pixel with neighbors and thus encode the information. In lemp ,pixel at the centre is replaced by edge values of neighbors. Discriminating information provided by line edges makes this method different from many of the existing methods. Magnitude is also added to the line edge information in order to make the feature descriptor more effective and robust. Performance of lemp method is estimated with corel database. Standard metrics such as recall, precision and average retrieval rate are determined for comparison purpose. Experimental values exhibit a notable improvement in the performance.