Detection of Buried Remote Monitoring Objects Using Multispectral Data

Multispectral data is obtained using special sensors that measure reflected or emitted energy in certain spectral ranges. This data represents information about various properties of objects, such as color, texture, structure, and more. Various methods of multispectral data processing can be used to detect buried objects. The article deals with the issues of image segmentation using deep learning, in order to obtain estimates of the thermophysical parameters of hidden subsurface objects, the coefficient inverse problem of thermal conductivity was solved using a genetic algorithm.