AI algorithms in photogrammetry and remote sensing (GeoAI)

Author: Sławomir Mikrut

In the past decade, the Polish market for image analysis has seen the emergence of highly specialized hyperspectral cameras capable of capturing spectral information across hundreds of channels simultaneously. This innovation significantly enhances the interpretive potential of obtained results by increasing the density of sampling for spectral curves, thereby improving their accuracy. Develop a unique Big Data database leveraging artificial intelligence (AI) algorithms to create proprietary algorithms for processing photogrammetry and remote sensing data, including hyperspectral data. Currently, many analyses in photogrammetry and remote sensing are performed manually.

The use of our AI will enable the automation of processes such as:
• Environmental monitoring (examining soil, water, and air pollution)
• Precision agriculture (accurate fertilization forecasting, predicting crop yield and quality)
• Crop type identification (projects for the Agency for Restructuring and Modernization of Agriculture, verifying crops in programs related to direct subsidies for farmers)
• Detection of hazardous materials (e.g., roofs containing asbestos) and identification of illegal discharges (e.g., detecting unauthorized sewage disposal from small and large facilities)
• Automatic determining the volume of earth masses using a drone and AI