Improved Approach of Seed Point Selection in RPCCL Algorithm for Aerial Remote Sensing Hyper-spectral Data Clustering with Data Dimensionality Reduction

With the number of channels in the hundreds instead of in the tens Hyper spectral imagery possesses much richer spectral information than multispectral imagery. The increased dimensionality of such Hyper spectral data provides a challenge to the current technique for analyzing data. Conventional classification methods may not be useful without dimension reduction pre-processing. So dimension reduction has become a significant part of Hyper spectral image processing. This paper presents a comparative analysis of the efficacy of Haar and Daubechies wavelets for dimensionality reduction in achieving image classification. Spectral data reduction using Wavelet Decomposition could be useful because it preserves the distinction among spectral signatures. Daubechies wavelets optimally capture the polynomial trends while Haar wavelet is discontinuous and resembles a step function. The performance of these wavelets are compared in terms of classification accuracy and time complexity. This paper shows that wavelet reduction has more separate classes and yields better or comparable classification accuracy. In the context of the dimensionality reduction algorithm, it is found that the performance of classification of Daubechies wavelets is better as compared to Haar wavelet while Daubechies takes more time compare to Haar wavelet. The experimental results demonstrate the classification system consistently provides over 84% classification accuracy.

Download Full-text

A Review of Dimensionality Reduction Techniques for Processing Hyper-Spectral Optical Signal

Light & Engineering ◽

10.33383/2019-017 ◽

2019 ◽

pp. 85-98 ◽

Cited By ~ 7

Author(s):

Ana del Águila ◽

Dmitry S. Efremenko ◽

Thomas Trautmann

Keyword(s):

Remote Sensing ◽

Dimensionality Reduction ◽

Principal Component ◽

Optical Signal ◽

Spectral Radiance ◽

Electromagnetic Spectrum ◽

Spectral Processing ◽

Reduction Techniques ◽

Hyper Spectral ◽

Dimensionality Reduction Techniques

Hyper-spectral sensors take measurements in the narrow contiguous bands across the electromagnetic spectrum. Usually, the goal is to detect a certain object or a component of the medium with unique spectral signatures. In particular, the hyper-spectral measurements are used in atmospheric remote sensing to detect trace gases. To improve the efficiency of hyper-spectral processing algorithms, data reduction methods are applied. This paper outlines the dimensionality reduction techniques in the context of hyper-spectral remote sensing of the atmosphere. The dimensionality reduction excludes redundant information from the data and currently is the integral part of high-performance radiation transfer models. In this survey, it is shown how the principal component analysis can be applied for spectral radiance modelling and retrieval of atmospheric constituents, thereby speeding up the data processing by orders of magnitude. The discussed techniques are generic and can be readily applied for solving atmospheric as well as material science problems.

Download Full-text

HYPER-SPECTRAL DATA CLUSTERING METHOD BASED UPON THE SENSITIVE SUBSPACE

New Mathematics and Natural Computation ◽

10.1142/s1793005707000756 ◽

2007 ◽

Vol 03 (02) ◽

pp. 271-280

Author(s):

WENCANG ZHAO ◽

WEI WANG

Keyword(s):

Probability Density ◽

Spectral Data ◽

Data Clustering ◽

Red Tide ◽

Competitive Learning ◽

Clustering Method ◽

High Dimensions ◽

Large Numbers ◽

Hyper Spectral ◽

Computational Speed

We put forward a quick clustering method for large numbers of data, with high dimensions, which is based on sensitive subspace consisting of the data set's sensitive dimensions. We first estimate the probability density of each dimension by the parzen window algorithm, enhance its optional ability through extracting zeroes and smoothness processing, then through the recognition of the number of the rallying points and the gain of the sensitive dimensions in order to compose the sensitive subspace, and lastly, we perform the Rival Penalized Competitive Learning (RPCL) clustering in the subspace. Moreover, we detected the red tide of hyper-spectral data using this method. Furthermore, the overall improvement in terms of the computational speed is about nine times faster.

Download Full-text

Application of an airborne hyper-spectral survey system CASI/SASI in the gold-silver-lead-zinc ore district of Huaniushan, Gansu, China

Geologia Croatica ◽

10.4154/gc.2021.04 ◽

2021 ◽

Vol 74 (1) ◽

pp. 73-83

Author(s):

Yuhai Fan ◽

◽

Yuiqing Wan ◽

Hui Wang ◽

Xingke Yang ◽

...

Keyword(s):

Remote Sensing ◽

Spectral Data ◽

Spectral Characteristics ◽

Hyperspectral Remote Sensing ◽

Alteration Zones ◽

Hyper Spectral ◽

Gold Silver ◽

Survey System ◽

Lead Zinc ◽

Spectral Survey

The airborne hyper-spectral survey system CASI/SASI, which has an integrated system for gathering both image an spectral data, is at the cutting edge developments in the remote-sensing field. It can be used to directly identify surface objects based on diagnostic spectral characteristics. In this paper, the CASI/SASI were used in the Huaniushan gold-silver-lead-zinc ore district–Gansu to produce a lithologic map, identify altered minerals, and map the mineralized-alteration zones. Radiometric correction, radiometric calibration, atmospheric correction (spectral reconstruction), and geometric corrections were carried out in ENVI to pre-process the measured data. A FieldSpec ® Pro FR portable spectrometer was used to obtain the spectral signatures of all types of rock samples, ore deposits, and mineralized-alteration zones. We extracted and analyzed the spectral characteristics of typical alteration minerals. On the basis of hyper-spectral data, ground-spectral data processing, and comparative analysis of the measured image spectrum, we used the spectral-angle-mapping (SAM) and mixture-tuned matchedfiltering (MTMF) methods to perform hyperspectral-alteration mineral mapping of wall rock and mineralized-alteration-zone hyperspectral identification. Hyperspectral- remote- sensing geological- classification maps were produced as well as distribution maps of all kinds of alteration minerals and mineralized-alteration zones. Based on geological comprehensive analysis and field investigations, the range of mineral alteration was proven to be the same as shown by the remote-sensing imagery. Indications are that airborne hyperspectral- remote-sensing -image CASI/SASI offer good application results and show a promising potential as a tool in geological investigations. The results will provide the basis for hyperspectral remote-sensing prospecting in the same or similar unexplored areas.

Download Full-text