Hough-Transform-Based Interpolation Scheme for Generating Accurate Dense Spatial Maps of Air Pollutants from Sparse Sensing

Air pollution is a significant health risk factor and causes many negative effects on the environment. Thus, arises the need for studying and assessing air-quality. Today, air-pollution assessment is mostly based on data acquired from Air Quality Monitoring (AQM) stations. These AQM stations provide continuous measurements and considered to be accurate; however, they are expensive to build and operate, thus scattered sparingly. To cope with this limitation, typically, the information obtained from those measurements is generalized with interpolation methods such as IDW or Kriging. Yet, the mathematical basis of those schemes defines that pollution extremum values are obtained at the measuring points. In addition, they are not considering the location of the pollution source or any physicochemical characteristics of pollutant hence do not reveal the real spatial air-pollution patterns. This research introduces a new interpolation scheme which breaks the interpolation process into two stages. At the first stage, the source of pollution and its estimated emission rate are inferred through a detection procedure which is based on the Hough Transform. At the second stage, based on the detected source location and emission, spatial dense pollution maps are created. The method requires, for its computation, to assume a dispersion model. To this end, any model can be used as sophisticated as it may be. Spatial maps created with simplified dispersion models in a computational simulation, show that the suggested interpolation scheme manages to create more accurate and more physically reasonable maps than the state-of-the-art.