TY - GEN
T1 - Memetic three-dimensional gabor feature extraction for hyperspectral imagery classification
AU - Zhu, Zexuan
AU - Shen, Linlin
AU - Sun, Yiwen
AU - He, Shan
AU - Ji, Zhen
PY - 2012
Y1 - 2012
N2 - This paper proposes a three-dimensional Gabor feature extraction for pixel-based hyperspectral imagery classification using a memetic algorithm. The proposed algorithm named MGFE combines 3-D Gabor wavelet feature generation and feature selection together to capture the signal variances of hyperspectral imagery, thereby extracting the discriminative 3-D Gabor features for accurate classification. MGFE is characterized with a novel fitness evaluation function based on independent feature relevance and a pruning local search for eliminating redundant features. The experimental results on two real-world hyperspectral imagery datasets show that MGFE succeeds in obtaining significantly improved classification accuracy with parsimonious feature selection.
AB - This paper proposes a three-dimensional Gabor feature extraction for pixel-based hyperspectral imagery classification using a memetic algorithm. The proposed algorithm named MGFE combines 3-D Gabor wavelet feature generation and feature selection together to capture the signal variances of hyperspectral imagery, thereby extracting the discriminative 3-D Gabor features for accurate classification. MGFE is characterized with a novel fitness evaluation function based on independent feature relevance and a pruning local search for eliminating redundant features. The experimental results on two real-world hyperspectral imagery datasets show that MGFE succeeds in obtaining significantly improved classification accuracy with parsimonious feature selection.
KW - Gabor Feature Extraction
KW - Hyperspectral Imagery Classification
KW - Memetic Algorithm
UR - http://www.scopus.com/inward/record.url?scp=84875084892&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-30976-2_58
DO - 10.1007/978-3-642-30976-2_58
M3 - Conference contribution
AN - SCOPUS:84875084892
SN - 9783642309755
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 479
EP - 488
BT - Advances in Swarm Intelligence - Third International Conference, ICSI 2012, Proceedings
T2 - 3rd International Conference on Swarm Intelligence, ICSI 2012
Y2 - 17 June 2012 through 20 June 2012
ER -