Block-based Watermarking for Robust Authentication and Integration of GIS Data

Abstract

A Geographic Information System (GIS) is a computer system for gathering, storing, transmitting, and presenting data related to positions on Earth’s surface. This research aims to authenticate the GIS data during transmission via internet based on transform-based invisible watermarking. The proposed framework uses the singular value decomposition and discrete cosine transformation in the frequency domain. The proposed framework is evaluated on National DEM images obtained from the Geospatial Information Agency's (Badan Informasi Geospatial—BIG) geoportal, under several types of attacks. Two performance metrics (Peak to Signal Noise Ratio (PSNR) and Mean Square Error (MSE)) were considered for the evaluation of the security of the designed framework. Likewise, Normalized Correlation (NC) was computed to assess the robustness by calculating the similarity between the original and the extracted images. The experimental outcomes show that the extracted logos are readable even if altered, which guarantees that the received DEM data are authentic.