Title :
Smart city and geospatiality: Hobart deeply learned
Author :
Aryal, Jagannath ; Dutta, Ritaban
Author_Institution :
Discipline of Geogr. & Spatial Sci, Univ. of Tasmania, Hobart, TAS, Australia
Abstract :
We propose a cloud computing based big data framework using Deep Neural Networks, to learn urban objects from very high-resolution image in an abstract optimized manner. Automatic recognition of such objects would be essential to minimize big data accessibility issues and increase efficiency of urban dynamics monitoring and planning. We have shown that deep learning could be a way forward towards that complex aim with very high accuracy rates.
Keywords :
Big Data; cloud computing; image resolution; learning (artificial intelligence); neural nets; object recognition; remote sensing; smart cities; town and country planning; Big Data framework; Hobart IKONOS data; cloud computing; deep learning; deep neural network; geospatiality; image resolution; object recognition; smart city; urban dynamics monitoring; urban planning; Accuracy; Big data; Cities and towns; Data mining; Feature extraction; Knowledge based systems; Spatial resolution; Deep Learning; GEOBIA; Hobart; IKONOS; geospatiality; smart cities; ultra-high resolution;
Conference_Titel :
Data Engineering Workshops (ICDEW), 2015 31st IEEE International Conference on
Conference_Location :
Seoul
DOI :
10.1109/ICDEW.2015.7129557
