Expression of Concern for: A Deep Belief Network Based Land Cover Classification

• Published in: 2021 Innovations in Power and Advanced Computing Technologies (i-PACT) p. 1
• Main Authors: Chintada, Koteswara Rao, Yalla, Surya Prakash, Uriti, Archana
• Format: Conference Proceeding
• Language: English
• Published: IEEE 27.11.2021
• DOI: 10.1109/i-PACT52855.2021.10702985

Land cover is the noticed actual cover on the Earth's surface which fills in as an ideal info boundary for various agricultural, hydrological and biological models. Thus, to foster feasible land use frameworks, there is a need to arrange the land cover. The principle objective of this work is to perform land cover arrangement by utilizing Hyper Spectral Image information and by applying Deep Belief Network. In this work, a combination approach is utilized to join the spatial and unearthly data in the arrangement cycle. An epic significant plan is proposed to get high portrayal accuracy in the wake of checking the capability of Restricted Boltzmann machine and Deep Belief Network. Trial results with broadly utilized hyper spectral information show those classifiers, inherent this profound learning-based system give serious execution. Also, the work reports to resolve the issues looked in utilizing 1-D information of the first Deep Belief Networks. To conquer this limit, another system is proposed for 3-D hyper unearthly picture which joins Principal Component Analysis; various leveled learning-based element extraction with calculated relapse.