A5019014062- Remote-Sensing Image Classification
- Remote Sensing and Land Use
- Astronomy and Astrophysical Research
- Advanced Image Fusion Techniques
- Galaxies: Formation, Evolution, Phenomena
- Advanced Chemical Sensor Technologies
- Face and Expression Recognition
- Stellar, planetary, and galactic studies
- Remote Sensing in Agriculture
- Image Retrieval and Classification Techniques
- Optical measurement and interference techniques
- Advanced Computational Techniques and Applications
- Adaptive optics and wavefront sensing
- Machine Learning and ELM
- Network Traffic and Congestion Control
- Land Use and Ecosystem Services
- Sparse and Compressive Sensing Techniques
- Astronomical Observations and Instrumentation
- Scientific Research and Discoveries
- Geochemistry and Geologic Mapping
- Image Processing Techniques and Applications
- Software-Defined Networks and 5G
- Astrophysics and Star Formation Studies
- Interconnection Networks and Systems
- History and Developments in Astronomy
Hohai University
2016-2025
State Forestry and Grassland Administration
2023
ORCID
2021
Armstrong Atlantic State University
2002-2017
Shenzhen University
2016
State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering
2014-2015
Chinese Academy of Sciences
1995-2015
Nanjing University
2012-2015
The 180th Hospital of PLA
2014
Nanjing University of Science and Technology
2013
The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST, also called the Guo Shou Jing Telescope) is a special reflecting Schmidt telescope. LAMOST's design allows both large aperture (effective of 3.6 m–4.9 m) and wide field view (FOV) (5°). It has an innovative active configuration which continuously changes mirror's surface that adjusts during observation process combines thin deformable mirror optics with segmented optics. Its primary (6.67 m × 6.05 (5.74m 4.40m) are...
The Large sky Area Multi-Object Spectroscopic Telescope (LAMOST) General Survey is a spectroscopic survey that will eventually cover approximately half of the celestial sphere and collect 10 million spectra stars, galaxies QSOs. Objects both in pilot first year general are included LAMOST First Data Release (DR1). started October 2011 ended June 2012, data have been released to public as Pilot August 2012. September completed its operation 2013. DR1 includes total 1202 plates containing...
It is of great interest in exploiting texture information for classification hyperspectral imagery (HSI) at high spatial resolution. In this paper, a paradigm to exploit rich HSI proposed. The proposed framework employs local binary patterns (LBPs) extract image features, such as edges, corners, and spots. Two levels fusion (i.e., feature-level decision-level fusion) are applied the extracted LBP features along with global Gabor original spectral where involves concatenation multiple before...
Band selection is often applied to reduce the dimensionality of hyperspectral imagery. When desired object information known, it can be achieved by finding bands that contain most information. It expected these provide an overall satisfactory detection and classification performance. In this letter, we propose a new supervised band-selection algorithm uses known class signatures only without examining original or need training samples. Thus, complete task much faster than traditional methods...
Extreme learning machine (ELM) is a single-layer feedforward neural network based classifier that has attracted significant attention in computer vision and pattern recognition due to its fast speed strong generalization. In this paper, we propose integrate spectral-spatial information for hyperspectral image classification exploit the benefits of using spatial features kernel ELM (KELM) classifier. Specifically, Gabor filtering multihypothesis (MH) prediction preprocessing are two...
This paper describes the data release of LAMOST pilot survey, which includes reduction, calibration, spectral analysis, products and access. The accuracy released information about FITS headers spectra are also introduced. set 319 000 a catalog these objects.
Hyperspectral imagery can obtain hundreds of narrow spectral bands ground objects. The abundant and detailed information offers a unique diagnostic identification ability for targets interest. anomaly detection aims to find without prior knowledge, which has attracted attention as branch target location. In this article, current hyperspectral methods, performance evaluation techniques, data sets are widely investigated. Among them, methods be classified into seven categories:...
view Abstract Citations (354) References (49) Co-Reads Similar Papers Volume Content Graphics Metrics Export Citation NASA/ADS Nearby Seyfert galaxies. Simkin, S. M. ; Su, H. J. Schwarz, P. The paper contains preliminary results from a photographic study of 30 nearby galaxies (VR less than 5000 km/s). appearance these well-resolved objects suggests that the morphology both type 1 and 2 is characterized by three-step distribution in radial surface brightness disk. Superposed on this continuum...
A particle swarm optimization (PSO)-based system is proposed to select bands and determine the optimal number of be selected simultaneously, which near-automatic with only a few data-independent parameters. The includes two swarms, i.e., outer one for estimating inner corresponding band selection. To avoid employing an actual classifier within PSO so as greatly reduce computational cost, criterion functions that can gauge class separability are preferred; specifically, minimum estimated...
A firefly algorithm (FA) inspired band selection and optimized extreme learning machine (ELM) for hyperspectral image classification is proposed. In this framework, FA to select a subset of original bands reduce the complexity ELM network. It also adapted optimize parameters in (i.e., regularization coefficient C, Gaussian kernel σ, hidden number neurons L). Due very low ELM, its accuracy can be used as objective function during parameter optimization. experiments, two datasets acquired by...
Recently, collaborative representation classification (CRC) has attracted much attention for hyperspectral image analysis. In particular, tangent space CRC (TCRC) achieved excellent performance in a simplified space. this article, novel Bagging-based TCRC (TCRC-bagging) and Boosting-based (TCRC-boosting) methods are proposed. The main idea of TCRC-bagging is to generate diverse results using the bootstrap sample method, which can enhance accuracy diversity single classifier simultaneously....
An improved firefly algorithm (FA)-based band selection method is proposed for hyperspectral dimensionality reduction (DR). In this letter, DR formulated as an optimization problem that searches a small number of bands from data set, and feature subset search using the FA developed. To avoid employing actual classifier within searching process to greatly reduce computational cost, criterion functions can gauge class separability are preferred; specifically, minimum estimated abundance...
Anomaly detection has become an important remote sensing application due to the abundant spectral and spatial information contained in hyperspectral images. Recently, anomaly methods based on collaborative representation model have attracted significant attention. Nevertheless, these face two main challenges: (1) all features (spectral signatures) are constrained share same coefficient, which ignores differences among features; (2) existing dictionaries for pixel-by-pixel usually not...
Band clustering is applied to dimensionality reduction of hyperspectral imagery. Different from unsupervised using all the pixels or supervised requiring labeled pixels, proposed semisupervised band needs class spectral signatures only. After clustering, a cluster selection step select clusters be used in following data analysis. Initial conditions and distance metrics are also investigated improve performance. The experimental results show that algorithm can outperform other existing...
Recently, collaborative representation detector (CRD) has been popularly used for hyperspectral anomaly detection. For the original CRD, least squares solution becomes more unstable when classes, i.e., samples detection are involved, and error is likely to happen if test pixel an anomalous several from background similar anomalous. In this paper, we propose a method that uses CRD with principal component analysis (PCA) removing outlier (PCAroCRD). According different modeling methods, global...
Recently, multifeature learning in collaborative representation classification (CRC) for hyperspectral images has generated promising performance. In this paper, two novel algorithms that update dictionary directly and indirectly are proposed. order to offer the complementarity of multifeature, four different types features-global feature (i.e., Gabor feature), local binary pattern), shape extended multiattribute profiles), spectral feature-are adopted paper. Under hypothesis most features...