A novel model of the deep neural network approach in coal mining surface pattern to assess land use classification using remote sensing image

Abstract

Abstract Computer vision usage in coal mining land use (LU) pattern classification is an exigent application for accurate accuracy. In the context of the satellite image pattern classification, the artificial intelligence (AI) for the development of a deep neural network (DNN), we have expressed an interest in land use (LU) for coal mining activities. Also, this image is addressed an area of interest as a spatial feature pattern that characterises coal mining regions. Moreover, the DNN classification algorithm's performance depends on the quality of the dataset. In the satellite image dataset, these practices of supervised-based learning are used for the accuracy assessment of mining activities area that is categorized into five classes coal area, built-up area, barren area, vegetation area, and water area respectively. The suit of mining activities area is selected from a case study of Talcher, Odisha, India. Further, we have found performances of training, testing, and validation like 88%, 69.7%, and 73.6%, respectively. Also, the overall accuracy is 79.4%. Therefore, the potential of DNNs learning is introduced for LU classification over mining activities area.