Phanideep, Kallepalli (2018) Mineral Identification Using Hyperspectral Image: a Case Study of Gua Iron Ore Mine, Jharkhand. MTech thesis.
|PDF (Restricted upto 31/03/2019) |
Restricted to Repository staff only
Hyperspectral imaging plays a noticeable role in the remote sensing group.Recent investigations state that Spectral alongside spatial handling increases the classification precision of hyperspectral images (HSI).The present studydemonstrates the use of hyperspectral images(EO-1 Hyperiondata)for mineral identification over Gua Iron Ore Mine, Jharkhand, India. The reflectance and emittance spectroscopyin the VNIR and SWIR band soffer an effective path in distinguishing the mineral zones from the other land cover types like vegetation, water body, etc. The first requirement is to retrieve the hyperspectral data for the area of interest in the desired bands from the original dataset. The retrieved image/data may contain errors and thus the primary prerequisite is to conduct the image preprocessing to remove the errors. The preprocessing of the retrieved image was done for
atmospheric correction, cloud removal, and removal of water bands. The hyperspectral image classification for identifying the iron ores bearing zones has been carried out in different stages viz.minimum noise fraction (MNF)analysis for removal of noisy bands, pixel purity index (PPI for obtaining the unadulterated pixels,characterization of the unadulterated pixels to identify the object, n-D visualization of unadulterated pixels to get the spectral information for identifying the objects by comparing with the standard library designed by United State Geological Survey (USGS). The classification of the image has been done using two classifiers (Spectral Angle Mapping (SAM) and Binary encoding classifier based on the spectral information of the ide notified objects.The results revealed that the Hyperion data can be successfully used in mineral identification. The study results also indicated that both the classifiers Spectral Angle Mapping (SAM) and Binary encoding classifier performs equally well in mineral identification.
|Item Type:||Thesis (MTech)|
|Uncontrolled Keywords:||Hyperspectral remote sensing; Gua iron ore mine; Hyperion; MNF; PPI; n-D visualizer; Spectral angle mapping(SAM); Binary encoding mapping|
|Subjects:||Engineering and Technology > Mining Engineering > Underground Mining|
Engineering and Technology > Mining Engineering > Mining Industry
|Divisions:||Engineering and Technology > Department of Mining Engineering|
|Deposited By:||IR Staff BPCL|
|Deposited On:||28 Mar 2019 11:05|
|Last Modified:||28 Mar 2019 11:05|
|Supervisor(s):||Gorai, Amit Kumar|
Repository Staff Only: item control page