Mapping Industrial Poultry Operations at Scale With Deep Learning and Aerial Imagery

Concentrated Animal Feeding Operations (CAFOs) pose serious risks to air, water, and public health, but have proven to be challenging to regulate. The U.S. Government Accountability Office notes that a basic challenge is the lack of comprehensive location information on CAFOs. We use the USDA’s National Agricultural Imagery Program (NAIP) 1m/pixel aerial imagery to detect poultry CAFOs across the continental United States. We train convolutional neural network (CNN) models to identify individual poultry barns and apply the best performing model to over 42 TB of imagery to create the first national, open-source dataset of poultry CAFOs. We validate the model predictions against held-out validation set on poultry CAFO facility locations from 10 hand-labeled counties in California and demonstrate that this approach has significant potential to fill gaps in environmental monitoring.

[1]  Nikos Komodakis,et al.  Rotation Equivariant Vector Field Networks , 2016, 2017 IEEE International Conference on Computer Vision (ICCV).

[2]  Sang Michael Xie,et al.  Combining satellite imagery and machine learning to predict poverty , 2016, Science.

[3]  P. Peppard,et al.  Residential proximity to concentrated animal feeding operations and allergic and respiratory disease. , 2019, Environment international.

[4]  Frank Hutter,et al.  Decoupled Weight Decay Regularization , 2017, ICLR.

[5]  Steve Wing,et al.  Air Pollution, Lung Function, and Physical Symptoms in Communities Near Concentrated Swine Feeding Operations , 2011, Epidemiology.

[6]  Michele Volpi,et al.  Land cover mapping at very high resolution with rotation equivariant CNNs: towards small yet accurate models , 2018, ISPRS Journal of Photogrammetry and Remote Sensing.

[7]  Susan A. Maroney,et al.  Using object-based image analysis to map commercial poultry operations from high resolution imagery to support animal health outbreaks and events. , 2020, Geospatial health.

[8]  Daniel E. Ho,et al.  Deep learning to map concentrated animal feeding operations , 2019, Nature Sustainability.

[9]  Kathleen A. Smith,et al.  Antimicrobial residues in animal waste and water resources proximal to large-scale swine and poultry feeding operations. , 2002, The Science of the total environment.

[10]  Michael A. Mallin,et al.  Impacts of Industrial Animal Production on Rivers and Estuaries , 2000 .

[11]  Jaki S. McCarthy,et al.  Combining Multiple Methods in Establishment Questionnaire Testing: The 2017 Census of Agriculture Testing Bento Box , 2018, Journal of Official Statistics.

[12]  Cassandra Handan-Nader,et al.  Deep learning with satellite imagery to enhance environmental enforcement , 2021 .

[13]  Denise A. Abreu,et al.  The 2012 Census of Agriculture: A Capture–Recapture Analysis , 2017 .

[14]  D. Opitz,et al.  Object recognition and image segmentation: the Feature Analyst® approach , 2008 .

[15]  Sheryl Magzamen,et al.  Modelling the domestic poultry population in the United States: A novel approach leveraging remote sensing and synthetic data methods. , 2020, Geospatial health.

[16]  Ow,et al.  National Pollutant Discharge Elimination System (NPDES) , 2014 .

[17]  J. Barker,et al.  LIVESTOCK MANURE NUTRIENT ASSESSMENT IN NORTH CAROLINA , 2009 .

[18]  S. Myint,et al.  Spatiotemporal land use change and environmental degradation surrounding CAFOs in Michigan and North Carolina. , 2021, The Science of the total environment.

[19]  Christopher Ré,et al.  Snorkel: Rapid Training Data Creation with Weak Supervision , 2017, Proc. VLDB Endow..

[20]  Anthony Ortiz,et al.  Temporal Cluster Matching for Change Detection of Structures from Satellite Imagery , 2021, COMPASS.

[21]  J. Burkholder,et al.  Impacts of Waste from Concentrated Animal Feeding Operations on Water Quality , 2006, Environmental health perspectives.

[22]  A. Borchers,et al.  Major Uses of Land in the United States, 2012 , 2012 .

[23]  Zhiguo Cao,et al.  In-field automatic observation of wheat heading stage using computer vision , 2016 .

[24]  Minimum-Area Rectangle Containing a Set of Points , 2016 .

[25]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[26]  Ben Chugg,et al.  Enhancing Environmental Enforcement with Near Real-Time Monitoring: Likelihood-Based Detection of Structural Expansion of Intensive Livestock Farms , 2021, Int. J. Appl. Earth Obs. Geoinformation.

[27]  Nima Tajbakhsh,et al.  UNet++: A Nested U-Net Architecture for Medical Image Segmentation , 2018, DLMIA/ML-CDS@MICCAI.

[28]  H. H. Wooten.  Major Uses of Land in the United States , 1953 .

[29]  Kathe E. Bjork,et al.  Simulating the Distribution of Individual Livestock Farms and Their Populations in the United States: An Example Using Domestic Swine (Sus scrofa domesticus) Farms , 2015, PloS one.

[30]  J. Kline,et al.  School proximity to concentrated animal feeding operations and prevalence of asthma in students. , 2006, Chest.

[31]  C. Gerba,et al.  Sources of pathogenic microorganisms and their fate during land application of wastes. , 2005, Journal of environmental quality.

[32]  Herbert Freeman,et al.  Determining the minimum-area encasing rectangle for an arbitrary closed curve , 1975, CACM.

[33]  D. R. Raman,et al.  Estrogen content of dairy and swine wastes. , 2004, Environmental science & technology.

[34]  Ray Purdy,et al.  Using Earth Observation Technologies for Better Regulatory Compliance and Enforcement of Environmental Laws , 2010 .

[35]  R. Johansson,et al.  The Manure Hits the Land: Economic and Environmental Implications When Land Application of Nutrients is Constrained , 2004 .

[36]  Geoff Boeing,et al.  OSMnx: New Methods for Acquiring, Constructing, Analyzing, and Visualizing Complex Street Networks , 2016, Comput. Environ. Urban Syst..

[37]  Carrie Hribar.  Understanding concentrated animal feeding operations and their impact on communities , 2010 .