Deserts are one of the major landforms on the Earth. While deserts occupy about one-fifth of Earth’s land surface, they have been studied to a much lesser extent. All over the world, desert landforms are expanding ev...Deserts are one of the major landforms on the Earth. While deserts occupy about one-fifth of Earth’s land surface, they have been studied to a much lesser extent. All over the world, desert landforms are expanding ever rapidly and more and more human settlements are finding place in desert regions for habitation. Thus, quantifying and monitoring dunes becomes more relevant from a managerial perspective. Analyzing desert areas using satellite imagery is a challenging task due to weak textural differences and nearly homogeneous spectral responses in most parts of the terrain. In this paper, a post-clustering methodology for change detection of desert sand dunes is proposed. Features based on Radon spectrum are used to cluster dunes of various orientations. These clustered boundaries are used to detect if there are any changes occurring in the dune regions. In the experiments, remote sensing data covering various dune regions of the world are observed for possible changes in dune orientations. In all the cases, it is seen that there are no major changes in desert dune orientations since three decades.展开更多
With the rapid growth of the Internet,the copyright protection problem occurs frequently,and unauthorized copying and distributing of geospatial data threaten the investments of data producers.Digital watermarking is ...With the rapid growth of the Internet,the copyright protection problem occurs frequently,and unauthorized copying and distributing of geospatial data threaten the investments of data producers.Digital watermarking is a possible solution to solve this issue.However,watermarking causes modifications in the original data resulting in distortion and affects accuracy,which is very important to geospatial vector data.This article provides distortion assessment of watermarked geospatial data using wavelet-based invisible watermarking.Eight wavelets at different wavelet decomposition levels are used for accuracy evaluation with the help of error measures such as maximum error and mean square error.Normalized correlation is used as a similarity index between original and extracted watermark.It is observed that the increase in the strength of embedding increases visual degradation.Haar wavelet outperforms the other wavelets,and the third wavelet decomposition level is proved to be optimal level for watermarking.展开更多
文摘Deserts are one of the major landforms on the Earth. While deserts occupy about one-fifth of Earth’s land surface, they have been studied to a much lesser extent. All over the world, desert landforms are expanding ever rapidly and more and more human settlements are finding place in desert regions for habitation. Thus, quantifying and monitoring dunes becomes more relevant from a managerial perspective. Analyzing desert areas using satellite imagery is a challenging task due to weak textural differences and nearly homogeneous spectral responses in most parts of the terrain. In this paper, a post-clustering methodology for change detection of desert sand dunes is proposed. Features based on Radon spectrum are used to cluster dunes of various orientations. These clustered boundaries are used to detect if there are any changes occurring in the dune regions. In the experiments, remote sensing data covering various dune regions of the world are observed for possible changes in dune orientations. In all the cases, it is seen that there are no major changes in desert dune orientations since three decades.
文摘With the rapid growth of the Internet,the copyright protection problem occurs frequently,and unauthorized copying and distributing of geospatial data threaten the investments of data producers.Digital watermarking is a possible solution to solve this issue.However,watermarking causes modifications in the original data resulting in distortion and affects accuracy,which is very important to geospatial vector data.This article provides distortion assessment of watermarked geospatial data using wavelet-based invisible watermarking.Eight wavelets at different wavelet decomposition levels are used for accuracy evaluation with the help of error measures such as maximum error and mean square error.Normalized correlation is used as a similarity index between original and extracted watermark.It is observed that the increase in the strength of embedding increases visual degradation.Haar wavelet outperforms the other wavelets,and the third wavelet decomposition level is proved to be optimal level for watermarking.
