Land Use/Land Cover (LULC) mapping plays a major role in land management applications such as to generate community map, proper urban planning, and disaster risk management. Proposed algorithm efficiently segments different land use/land cover classes such as buildings, trees, bare land, and water body. RMS value based multi-thresholding technique is used to segment various land use/land cover classes and consequently using the binning technique to accurately estimate the utilization of earth’s surface. The proposed algorithm is tested on two different data sets of Bengaluru city, India. The percentage utilization of surface objects for grid 7 image of dataset I is found to be 96.68% building, 1.05% vegetation, and 0.22% barren land, the area covered in grid 7 of dataset I is identified as overutilized land. Percentage utilization of surface objects for grid 8 of dataset II is found to be 68.95% building, 11.93% vegetation, 0.15% bare land, and 1.14% water body. The area covered in grid 8 of dataset II is identified as overutilized land.
CITATION STYLE
Sowmya, D. R., Kulkarni, A. N., Sandeep, S., Deepa Shenoy, P., & Venugopal, K. R. (2018). Land use/land cover segmentation of satellite imagery to estimate the utilization of earth’s surface. In Advances in Intelligent Systems and Computing (Vol. 712, pp. 27–36). Springer Verlag. https://doi.org/10.1007/978-981-10-8228-3_4
Mendeley helps you to discover research relevant for your work.