Abstract
The paper proposes a transparent approach for mapping the status of environmental phenomena from multisource information based on both soft computing and machine learning. It is transparent, intended as human understandable as far as the employed criteria, and both knowledge and data-driven. It exploits remote sensing experts’ interpretations to define the contributing factors from which partial evidence of the environmental status are computed by processing multispectral images. Furthermore, it computes an environmental status indicator (ESI) map by aggregating the partial evidence degrees through a learning mechanism, exploiting volunteered geographic information (VGI). The approach is capable of capturing the specificities of local context, as well as to cope with the subjectivity of experts’ interpretations. The proposal is applied to map the status of standing water areas (i.e., water bodies and rivers and human-driven or natural hazard flooding) using multispectral optical images by ESA Sentinel-2 sources. VGI comprises georeferenced observations created both in situ by agronomists using a mobile application and by photointerpreters interacting with a geographic information system (GIS) using several information layers. Results of the validation experiments were performed in three areas of Northern Italy characterized by distinct ecosystems. The proposal showed better performances than traditional methods based on single spectral indexes.
Highlights
In the age of bigdata, we are faced with the new challenge of exploiting multisource information for several purposes that span from territorial monitoring to planning and recovery after critical events
It is based on soft computing, a branch of artificial intelligence founded on fuzzy set theory [6], a formal framework well-suited to model and process imperfect information
In the case expert’s interpretation of a phenomenon, possibly incomplete and imprecise, a soft constraint can be defined by the domain expert to specify a criterion to compute a partial evidence of the phenomenon given the information on the value of a variable v, which is selected as a contributing factor of the phenomenon
Summary
In the age of big (geo)data, we are faced with the new challenge of exploiting multisource information for several purposes that span from territorial monitoring to planning and recovery after critical events. This may cause doubts on both data reliability and the suitability for making decisions to benefit territory management To solve such an impasse, especially with respect to the exploitation of the huge data flow of remote sensing-derived information, flexible approaches for big (geo)data synthesis are needed to generate environmental status maps in near real-time [4]. We propose a knowledge and data-driven synthesis of environmental status indicator (ESI) maps that evolves our first original proposal [5] by aggregating remote sensing data and georeferenced observations It is based on soft computing, a branch of artificial intelligence founded on fuzzy set theory [6], a formal framework well-suited to model and process imperfect information (i.e., information affected by uncertainty, imprecision and vagueness). We evolve this method by incorporating a machine-learning mechanism, adapting the mapping to a specific region of interest (ROI) by exploiting available VGI [5]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have