TY - GEN
T1 - Aerial image classification using texture and color-based descriptors
AU - Cortes, Daniel
AU - Calderón, Gustavo
AU - Arista, Antonio
AU - Toscano, Karina
AU - Nakano, Mariko
N1 - Publisher Copyright:
© 2016 IEEE.
PY - 2017/7/19
Y1 - 2017/7/19
N2 - In this paper, we evaluate eleven image descriptors for urban-rural classification using aerial images. The eleven image descriptors are composed by seven texture-based descriptors and four combinations of texture and color descriptors. The classification is carried out using Support Vector Machine (SVM) with radial basis function as kernel function. The performance of these images descriptors are evaluated using accuracy, precision, sensitivity and specificity. From the evaluation, the combination of Gabor descriptor and Dominant Color descriptor provides a better performance, obtaining accuracy more than 91%.
AB - In this paper, we evaluate eleven image descriptors for urban-rural classification using aerial images. The eleven image descriptors are composed by seven texture-based descriptors and four combinations of texture and color descriptors. The classification is carried out using Support Vector Machine (SVM) with radial basis function as kernel function. The performance of these images descriptors are evaluated using accuracy, precision, sensitivity and specificity. From the evaluation, the combination of Gabor descriptor and Dominant Color descriptor provides a better performance, obtaining accuracy more than 91%.
KW - Image descriptor
KW - aerial image classification
KW - color-based descriptor
KW - texture-based descriptor
KW - urban-rural classification
UR - http://www.scopus.com/inward/record.url?scp=85027852460&partnerID=8YFLogxK
U2 - 10.1109/CNCG.2016.7985077
DO - 10.1109/CNCG.2016.7985077
M3 - Contribución a la conferencia
AN - SCOPUS:85027852460
T3 - 2016 IEEE 1er Congreso Nacional de Ciencias Geoespaciales: Sustainable Geospatial Technology at Service of Society, CNCG 2016 - Proceedings
SP - 9
EP - 12
BT - 2016 IEEE 1er Congreso Nacional de Ciencias Geoespaciales
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 1st IEEE Mexican Geospatial Science Congress, CNCG 2016
Y2 - 7 December 2016 through 9 December 2016
ER -