Interpretation of multispectral satellite imagery to meet the challenges of search and identification of archaeological sites
The paper presents the results of the development and testing of algorithms for automated search and identification of archaeological sites using multispectral imagery. These examples are based on the experimental data obtained on representative sites of Altai (watershed Chuya). To achieve the goals and objectives of the proposed technology algorithmized interpretation, based on two approaches. The first is based on the mathematical description of the spatial distribution and correlation of luminance elements of satellite images for the implementation of the second algorithm of multistage classification. The experiments reveal problems that reduce the possibility of using the developed algorithms and modules - seasonal shooting and high differentiation of the spectral curve of the images on the index, low spatial resolution at a relatively small size objects of the search , the lack of spectral library search target areas , etc. However, the results demonstrate the feasibility of using the developed approaches and algorithms, as evidenced by a stable result in significant variations in parameters, and search terms.