Building Extraction using SLIC from High Resolution Satellite Image

Abstract

Data extraction has always been a priority for any company or government agency in satellite imagery. This release of Data in the form of architecture from satellite imagery has played a very important role in the military, national security, agriculture, and understanding of global geography. It is a field of ongoing research into computer vision and remote hearing. In this paper we present the default model for creating automated method of extracting buildings structures from a high-resolution satellite image using deep learning and segmentation. Here, we perform a separate segmentation using the SLIC algorithm. Segmenting the image before training increase accuracy and efficiency of the model.