Learning to Understand Remote Sensing Images

• Hoofdauteur: Wang, Qi (auth)
• Formaat: Elektronisch Hoofdstuk
• Taal: Engels
• Gepubliceerd in: MDPI - Multidisciplinary Digital Publishing Institute 2019
• Onderwerpen: metadata; image classification; sensitivity analysis; ROI detection; residual learning; image alignment; adaptive convolutional kernels; Hough transform; class imbalance; land surface temperature; inundation mapping; multiscale representation; object-based; convolutional neural networks; scene classification; morphological profiles; hyperedge weight estimation; hyperparameter sparse representation; semantic segmentation; vehicle classification; flood; Landsat imagery; target detection; multi-sensor; building damage detection; optimized kernel minimum noise fraction (OKMNF); sea-land segmentation; nonlinear classification; land use; SAR imagery; anti-noise transfer network; sub-pixel change detection; Radon transform; segmentation; remote sensing image retrieval; TensorFlow; convolutional neural network; particle swarm optimization; optical sensors; machine learning; mixed pixel; optical remotely sensed images; object-based image analysis; very high resolution images; single stream optimization; ship detection; ice concentration; online learning; manifold ranking; dictionary learning; urban surface water extraction; saliency detection; spatial attraction model (SAM); quality assessment; Fuzzy-GA decision making system; land cover change; multi-view canonical correlation analysis ensemble; land cover; semantic labeling; sparse representation; dimensionality expansion; speckle filters; hyperspectral imagery; fully convolutional network; infrared image; Siamese neural network; Random Forests (RF); feature matching; color matching; geostationary satellite remote sensing image; change feature analysis; road detection; deep learning; aerial images; image segmentation; aerial image; multi-sensor image matching; HJ-1A/B CCD; endmember extraction; high resolution; multi-scale clustering; heterogeneous domain adaptation; hard classification; regional land cover; hypergraph learning; automatic cluster number determination; dilated convolution; MSER; semi-supervised learning; gate; Synthetic Aperture Radar (SAR); downscaling; conditional random fields; urban heat island; hyperspectral image; remote sensing image correction; skip connection; ISPRS; spatial distribution; geo-referencing; Support Vector Machine (SVM); very high resolution (VHR) satellite image; classification; ensemble learning; synthetic aperture radar; conservation; convolutional neural network (CNN); THEOS; visible light and infrared integrated camera; vehicle localization; structured sparsity; texture analysis; DSFATN; CNN; image registration; UAV; unsupervised classification; SVMs; SAR image; fuzzy neural network; dimensionality reduction; GeoEye-1; feature extraction; sub-pixel; energy distribution optimizing; saliency analysis; deep convolutional neural networks; sparse and low-rank graph; hyperspectral remote sensing; tensor low-rank approximation; optimal transport; SELF; spatiotemporal context learning; Modest AdaBoost; topic modelling; multi-seasonal; Segment-Tree Filtering; locality information; GF-4 PMS; image fusion; wavelet transform; hashing; machine learning techniques; satellite images; climate change; road segmentation; remote sensing; tensor sparse decomposition; Convolutional Neural Network (CNN); multi-task learning; deep salient feature; speckle; canonical correlation weighted voting; fully convolutional network (FCN); despeckling; multispectral imagery; ratio images; linear spectral unmixing; hyperspectral image classification; multispectral images; high resolution image; multi-objective; convolution neural network; transfer learning; 1-dimensional (1-D); threshold stability; Landsat; kernel method; phase congruency; subpixel mapping (SPM); tensor; MODIS; GSHHG database; compressive sensing
• Online toegang: DOAB: download the publication; DOAB: description of the publication