• Corpus ID: 239016693

Mapping illegal waste dumping sites with neural-network classification of satellite imagery

  title={Mapping illegal waste dumping sites with neural-network classification of satellite imagery},
  author={Maria Roberta Devesa and H. Antonio Vazquez Brust},
Public health and habitat quality are crucial goals of urban planning. In recent years, the severe social and environmental impact of illegal waste dumping sites has made them one of the most serious problems faced by cities in the Global South, in a context of scarce information available for decision making. To help identify the location of dumping sites and track their evolution over time we adopt a data-driven model from the machine learning domain, analyzing satellite images. This allows… 

Figures and Tables from this paper


The Impacts of Illegal Toxic Waste Dumping on Children’s Health: A Review and Case Study from Pasir Gudang, Malaysia
The findings of this study indicate that more effective policies and preventive actions are urgently needed to protect human health, especially children from improper hazardous waste management.
A Scalable Approach for Detecting Dumpsites using Automatic Target Recognition with Feature Selection and SVM through Satellite Imagery
  • 2020
Heavy metal pollution and health risk assessment of soils from open dumps in the Metropolitan Area of Buenos Aires, Argentina
The most sensitive variables in the stochastic model were exposure frequency and soil ingestion rate, together explaining 70% of the variation in the forecasts and calculated risks were highly dependent on Pb concentrations.
Detection of waste dumping locations in landfill using multi-temporal Landsat thermal images
In this study, it is demonstrated how to utilize thermal remote sensing techniques to measure the land surface temperature (LST), which can aid in outlining the waste dumping regions within a landfill.
The nature and detection of unauthorized waste dump sites using remote sensing
Now a day’s waste disposal has become a massive tricky for environmentalist, as it is responsible for a various issue such as sanitary hygienic, ecological security, illegal dumps, lack of paved or
Multiclass Weighted Loss for Instance Segmentation of Cluttered Cells
Two novel weight maps applied to the weighted cross entropy loss function are presented which take into account both class imbalance and cell geometry and show superior results when compared to other similar schemes, outperforming binary class models with significant improvement of boundary adequacy and instance detection.
Methods of detection in satellite images of illegal dumps by using a method based on tree classifier
A tree-like neural network classifier consisting of a set of simple (binary) classifiers in problems of detection of unauthorized garbage dumps and the use of principal component analysis as a dimension reduction algorithm space to treat the symptoms is considered.
U-Net: Convolutional Networks for Biomedical Image Segmentation
It is shown that such a network can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks.
Waste management – still a global challenge in the 21st century: An evidence-based call for action
  • David C. Wilson, C. Velis
  • Political Science
    Waste management & research : the journal of the International Solid Wastes and Public Cleansing Association, ISWA
  • 2015
What is the status of waste management around the world? Why should waste management be seen as a global issue and a political priority? What novel insights could we obtain by sketching for the very
Atlas de la basura del Área Metropolitana de Buenos Aires
  • 2012