Assessing optical remote sensing for grave detection

Abstract

The application of an effective and ready-to-use tool for discovering clandestine graves is crucial for solving a number of cases where disappearance of people is involved. This is the case in Mexico, where the government drug war has resulted in a large number of missing people that has been estimated to be over 40,000 since the year 2006. In this article, we report results from an experimental study on simulated animal graves detection using several techniques from optical remote sensing. Results showed that several spectral indices from hyperspectral and/or multispectral sensors may be used to detect N-enriched vegetation. Thermal imagery was also effective to detect underground voids through differential thermography, although this was only effective for detecting large graves with bare terrain. Lastly, while dense pointclouds reconstructed from oblique aerial photography was able to detect vegetation regrowth over the pits, the terrain subsidence was not sufficiently large to be detected with this technique, even in the case of mechanical removal of vegetation.