Title :
High Spatial Resolution remote sensing Image segmentation using Temporal Independent Pulse Coupled Neural Network
Author :
Liwei, Li ; Jianwen, Ma ; Xue, Chen ; Qi, Wen ; Xiaoyan, Xi
Author_Institution :
CAS, Beijing
Abstract :
Temporal independent pulse-coupled neuron network (TI-PCNN) has been developed and shows its usefulness on digital image segmentation. However, Due to its heavy computational cost and over-segmentation of objects within the range of low intensity, the original TI-PCNN method is ineffective at segmenting High Spatial Resolution remotely sensed Images (HSRI). By taking into account of spatial and spectral characteristics of HSRI, an improved method based on the TI-PCNN was developed and used to segment HSRI. Experiment was carried out on a subset of an aerial image. Result showed that the improved method largely overcomes the drawbacks of the original method and provided a promising approach for HSRI segmentation.
Keywords :
geophysics computing; image segmentation; neural nets; remote sensing; computational cost; digital image segmentation; high spatial resolution remote sensing; spatial characteristics; spectral characteristics; temporal independent pulse-coupled neural network; Computational efficiency; Digital images; Earth; Image segmentation; Joining processes; Neural networks; Neurons; Pixel; Remote sensing; Spatial resolution; High spatial resolution remote sensing image; Neuron Network; Pulse-Coupled; Segmentation; Temporal-Independent;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2007. IGARSS 2007. IEEE International
Conference_Location :
Barcelona
Print_ISBN :
978-1-4244-1211-2
Electronic_ISBN :
978-1-4244-1212-9
DOI :
10.1109/IGARSS.2007.4423200
