A New Environmental Image Processing Method for Chemical Weather Forecasts in Europe

It is common practice to present environmental information of spatial nature (like atmospheric quality patterns) in the form of pre-processed images. The current paper deals with the harmonization, comparison and reuse of Chemical Weather (CW) forecasts in the form of pre-processed images of varying quality and informational content, without having access to the original data. In order to compare, combine and reuse such environmental data, an innovative method for the inverse reconstruction of environmental data from images, was developed. The method is based on a new, neural adaptive data interpolation algorithm, and is tested on CW images coming from various European providers. Results indicate a very good performance that renders this method as appropriate to be used in various image-processing problems that require data reconstruction, retrieval and reuse.