Faculty Publications
Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736
Publications by NITK Faculty
Browse
Search Results
Item Virtual Sample Generation Of Hyperspectral Mineral Data(Institute of Electrical and Electronics Engineers Inc., 2023) Yadav, P.P.; Shetty, A.; Raghavendra, B.S.; Narasimhadhan, A.V.Identification of earth surface features or useful resources even at inaccessible locations for humans made possible with the help of spatial-spectral information provided by remote sensing data especially hyperspectral images. At present deep learning (DL) based models have become a best choice to address the issues and provide better solutions in many fields including remote sensing data analysis. Due to very subtle differences exhibited by mineral signatures and also lack of sufficient training samples, geo-science and remote sensing research community has not much explored the DL techniques in analyzing mineral data. Therefore, in this paper, a virtual sample generation technique using vector rotation is proposed to increase the mineral data for training DL models. The proposed virtual sample generation technique is explored on a spectral library of 5 minerals that is formed from Cuprite scene, a benchmark mineral data set. The quality of the mineral samples generated are assessed using visual inspection as well as a relative spectral discriminating power of target minerals with respect to non-target or remaining minerals. Results show that mineral samples generated by the proposed virtual sample generation technique are not only qualitative in nature but also helpful or encouraging in exploration of DL in mineral data analysis. © 2023 IEEE.Item 1-D CNN for Mineral Classification using Hyperspectral Data(Institute of Electrical and Electronics Engineers Inc., 2023) Yadav, P.P.; Shetty, A.; Raghavendra, B.S.; Narasimhadhan, A.V.Hyperspectral Image (HSI) is a potent remote sensing (RS) technique, capturing images over numerous narrow, contiguous spectral bands. Unlike traditional RS methods, HSI offers detailed spectral insights for each pixel, enhancing comprehension of the Earth's surface and its contents. Initially intended for mining and geology, its application has expanded across various domains. Yet, mineral identification poses challenges due to spectral signature variations and limited ground truth. Despite various advanced algorithms, including machine learning, no dedicated Deep Learning (DL) expert system exists for mineral classification in HSI. DL models require abundant training data and ground-truth, which are scarce in hyperspectral mineral data. Introducing the 1-D CNN model as a proposed method, we focus on enhancing mineral classification by increasing the available training data. The utilization of augmented training samples through the 1-D CNN model tackles the challenge of limited ground truth data, enabling accurate classification of mineral classes. © 2023 IEEE.Item Performance evaluation of dimensionality reduction techniques on hyperspectral data for mineral exploration(Springer Science and Business Media Deutschland GmbH, 2023) C, D.; Shetty, A.; Narasimhadhan, A.V.With recent advances in hardware and wide range of applications, hyperspectral remote sensing proves to be a promising technology for analysing terrain. However, the sheer volume of bands, strong inter band correlation and redundant information makes interpretation of hyperspectral data a tedious task. Aforementioned issues can be addressed to a considerable extent by reducing the dimensionality of hyperspectral data. Though plethora of algorithms exist to downsize hyperspectral data, quality assessment of these techniques remains unanswered. Since Dimensionality Reduction (DR) is a special case of unsupervised learning, classification accuracy cannot be directly used to compare the performance of different dimensionality reduction techniques. As a consequence, a different type of goodness measure is essential which is expected to be easily interpretable, robust against outliers and applicable to most algorithms and datasets. In this paper, fifteen popular dimensionality reduction algorithms are reviewed, evaluated and compared on hyperspectral dataset for mineral exploration. The performance of various DR algorithms is tested on hyperspectral mineral data since the extensive study of DR for mineral mapping is scarce compared to land cover mapping. Also, DR techniques are evaluated based on coranking criteria which is independent of label information. This facilitates to demonstrate the robust technique for mineral mapping and also provides meaningful insight into topology preservation. These techniques play a vital role in mineral exploration since in field observation is expensive, time consuming and requires more man power. From experimental results it is evident that, deep autoencoders provide better embedding with a quality index value of 0.9938, when K = 120 compared to other existing nonlinear techniques. The conclusions presented are unique since previous studies have not evaluated the results qualitatively and comparison between conventional machine learning and deep learning algorithms is limited. © 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.