Object Recognition for Economic Development from Daytime Satellite Imagery

Reliable data about the stock of physical capital and infrastructure in developing countries is typically very scarce. This is particular a problem for data at the subnational level where existing data is often outdated, not consistently measured or coverage is incomplete. Traditional data collection methods are time and labor-intensive costly which often prohibits developing countries from collecting this type of data. This paper proposes a novel method to extract infrastructure features from high-resolution satellite images. We collected high-resolution satellite images for 5 million 1km x 1km grid cells covering 21 African countries. We contribute to the growing body of literature in this area by training our machine learning algorithm on ground-truth data. We show that our approach strongly improves the predictive accuracy. Our methodology can build the foundation to then predict subnational indicators of economic development for areas where this data is either missing or unreliable.

[1]  Xiaoling Xia,et al.  Inception-v3 for flower classification , 2017, 2017 2nd International Conference on Image, Vision and Computing (ICIVC).

[2]  Stefano Ermon,et al.  Predicting Economic Development using Geolocated Wikipedia Articles , 2019, KDD.

[3]  D. Johnston,et al.  Statistical Tragedy in Africa? Evaluating the Data Base for African Economic Development , 2015 .

[4]  W. Nordhaus,et al.  Using luminosity data as a proxy for economic statistics , 2011, Proceedings of the National Academy of Sciences.

[5]  Tushar Agarwal,et al.  Multi-Task Deep Learning for Predicting Poverty From Satellite Images , 2018, AAAI.

[6]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[7]  Mathias Thoenig,et al.  This Mine is Mine! How Minerals Fuel Conflicts in Africa , 2014, SSRN Electronic Journal.

[8]  A. Storeygard,et al.  The View from Above: Applications of Satellite Data in Economics , 2016 .

[9]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[10]  Kathryn Baragwanath Vogel,et al.  Detecting Urban Markets with Satellite Imagery: An Application to India , 2018 .

[11]  Thomas M. Stoker,et al.  The Economics of Slums in the Developing World , 2013 .

[12]  Francesco Visin,et al.  A guide to convolution arithmetic for deep learning , 2016, ArXiv.

[13]  P. Sutton,et al.  SPECIAL ISSUE: The Dynamics and Value of Ecosystem Services: Integrating Economic and Ecological Perspectives Global estimates of market and non-market values derived from nighttime satellite imagery, land cover, and ecosystem service valuation , 2002 .

[14]  Xiao Chen,et al.  Infrastructure Quality Assessment in Africa using Satellite Imagery and Deep Learning , 2018, KDD.

[15]  Kush R. Varshney,et al.  Targeting direct cash transfers to the extremely poor , 2014, KDD.

[16]  P. Sutton,et al.  Estimation of Gross Domestic Product at Sub-National Scales using Nighttime Satellite Imagery , 2007 .

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

[18]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[19]  C. Elvidge,et al.  VIIRS night-time lights , 2017, Remote Sensing of Night-time Light.

[20]  Gabriel Cadamuro,et al.  Predicting poverty and wealth from mobile phone metadata , 2015, Science.

[21]  J. Henderson,et al.  Measuring Economic Growth from Outer Space , 2009, The American economic review.

[22]  Bistra N. Dilkina,et al.  A Deep Learning Approach for Population Estimation from Satellite Imagery , 2017, GeoHumanities@SIGSPATIAL.

[23]  Simon Johnson,et al.  provided that full credit, including © notice, is given to the source. Is Newer Better? Penn World Table Revisions and Their Impact on Growth Estimates , 2009 .

[24]  Stefano Ermon,et al.  Learning to Interpret Satellite Images Using Wikipedia , 2018, IJCAI.

[25]  Anne Driscoll,et al.  Using publicly available satellite imagery and deep learning to understand economic well-being in Africa , 2020, Nature Communications.

[26]  Budhendra L. Bhaduri,et al.  A global poverty map derived from satellite data , 2009, Comput. Geosci..

[27]  H. Overman,et al.  Causes of Sprawl: A Portrait from Space , 2006 .

[28]  R. Goldblatt,et al.  Detecting urban markets with satellite imagery: An application to India , 2019 .

[29]  Stefano Ermon,et al.  Tile2Vec: Unsupervised representation learning for spatially distributed data , 2018, AAAI.

[30]  James H. Faghmous,et al.  Equitable development through deep learning: The case of sub-national population density estimation , 2016, ACM DEV.

[31]  Jonathan Hersh,et al.  Evaluating the relationship between spatial and spectral features derived from high spatial resolution satellite data and urban poverty in Colombo, Sri Lanka , 2017, 2017 Joint Urban Remote Sensing Event (JURSE).