DocumentCode
153036
Title
Plant identification using local invariants
Author
Yildiran, S. Tolga ; Yanikoglu, Benin ; Abdullah, Emran
Author_Institution
Muhendislik ve Doga Bilimleri Fak., Sabanci Univ., Istanbul, Turkey
fYear
2014
fDate
23-25 April 2014
Firstpage
2094
Lastpage
2097
Abstract
We present a plant image recognition system geared towards plants with flowers. The system uses local invariants with Dense SIFT features and Bag of Visual Words representation, while the classification is done using Support Vector Machines. Our approach contains a pre-classification stage where images are categorized into color subgroups, to reduce the complexity of the problem. Using a 161-class subset of the ImageClef´2013 flower dataset, the classification accuracy is measured as %42.68, compared to %18 eithout the pre-classification.
Keywords
image classification; support vector machines; transforms; bag of visual words representation; dense SIFT feature; local invariant; plant identification; plant image recognition system; preclassification stage; scale invariant feature transform; support vector machines; Computer vision; Conferences; Histograms; Image color analysis; Signal processing; Support vector machines; Visualization; Bag of Visual Words (BOVW); Dense SIFT; K Means Clustering; PHOW; SVMs;
fLanguage
English
Publisher
ieee
Conference_Titel
Signal Processing and Communications Applications Conference (SIU), 2014 22nd
Conference_Location
Trabzon
Type
conf
DOI
10.1109/SIU.2014.6830674
Filename
6830674
Link To Document