An SVM Ensemble Approach Combining Spectral, Structural, and Semantic Features for the Classification of High-Resolution Remotely Sensed Imagery

By Xin Huang, Liangpei Zhang. IEEE Transactions on Geoscience and Remote Sensing, 2013

441 citations recorded by the source corpus.

Subfields: Atmospheric Science, Ecology, Media Technology

View Paper: https://www.semanticscholar.org/paper/f6071eb167ac70099d357976c0a065ad685193b8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Atmospheric Science

1. Antonio J. Plaza
2. Liangpei Zhang
3. Bo Du
4. Qian Du
5. Lefei Zhang
6. Javier Plaza
7. Shihong Du
8. Juan Mario Haut

In Ecology

1. Wenzhi Zhao
2. Shihong Du
3. Qian Du
4. Liangpei Zhang
5. Bo Du
6. Lefei Zhang
7. Gong Cheng
8. Junwei Han

In Media Technology

1. Liangpei Zhang
2. Gong Cheng
3. Antonio J. Plaza
4. Junwei Han
5. Qian Du
6. Bo Du
7. Javier Plaza
8. Juan Mario Haut

Citations in the corpus

• Local Binary Patterns and Extreme Learning Machine for Hyperspectral Imagery Classification (2015, methodology): “For various classification tasks, these features have their own advantages and disadvantages, and it is difficult to determine which one is always optimal [16].”
• Learning multiscale and deep representations for classifying remotely sensed imagery (2016, background): “However, due to the complex properties in both spatial and spectral domains, it is commonly recognized that higher resolutions do not naturally result in higher classification accuracies (Huang and Zhang, 2013), especially when dealing with urban area structures.”
• Unmanned Aerial Vehicle for Remote Sensing Applications - A Review (2019, background): “More advanced classifiers, including an ensemble of simpler classifiers [103,104] or deep learning-based methods [105,106], still present a huge potential to be explored for this ultra-high-resolution, multi-modal data (height, thermal, and hyperspectral information).”
• Sensing spatial distribution of urban land use by integrating points-of-interest and Google Word2Vec model (2017, background): “2011, Huang and Zhang 2013, Mordelet and Vert 2014).”
• Object Detection in Optical Remote Sensing Images Based on Weakly Supervised Learning and High-Level Feature Learning (2015, positive): “Unlike existing works to extract structural features solely based on human design [9, 10], the proposed approach derives high-level features by applying unsupervised representation learning approach, where spatial and structural patterns from the low-level and middle-level features can be…”
• Assessing the robustness of Random Forests to map land cover with high resolution satellite image time series over large areas (2016, background): “Currently, hundreds of features are computed (Dalla Mura et al., 2010; Huang and Zhang, 2013), and then the best subset is selected (Gressin et al.”
• Deep learning classifiers for hyperspectral imaging: A review (2019, methodology): “(2018b) introduce the LSS-RNN, a RNN model with a local spatial sequential method (LSS) that includes a low-level FE step, implemented using Gabor filtering and differential morphological profiles (DMPs) (Huang and Zhang, 2013), whose corresponding features are stacked together and passed through th”
• Ensemble manifold regularized sparse low-rank approximation for multiview feature embedding (2015, background): “HYDICE hyperspectral data set The pixel-based image classification task of HYDICE hyperspectral data set is challenging because many pixels of the roof, road, and path are similar in spectral domain (1st view) since they may be made of similar materials [46].”