A model for environmental data extraction from multimedia and its evaluation against various chemical weather forecasting datasets

摘要: Abstract Environmental data analysis and information provision are considered of great importance for people, since environmental conditions strongly related to health issues directly affect a variety everyday activities. Nowadays, there several free web-based services that provide in formats with map images being the most commonly used present air quality pollen forecasts. This format, despite intuitive humans, complicates extraction processing underlying data. Typical examples this case chemical weather forecasts, which usually encoded heatmaps (i.e. graphical representation matrix colors), while forecasted numerical pollutant concentrations unavailable. work presents model semi-automatic such based on template configuration tool, methodologies reconstruction from images, as well text Optical Character Recognition (OCR). The aforementioned modules integrated standalone framework, is extensively evaluated by comparing extracted heat maps against real values produced forecasting models. results demonstrate satisfactory performance terms recovery positional accuracy.