Census to use satellite imagery to count new home builds
The method entails running satellite images through "convolutional neural networks," or CNNs, to identify construction.
The U.S. Census Bureau unrolled a new experimental method of using images from space to count new housing construction.
"For select areas of interest, Survey of Construction (SOC) data collected by field representatives have been replaced with estimates obtained from satellite imagery to demonstrate the efficacy of this data collection method," Census said in a Dec. 19 release.
The method entails running satellite images through "convolutional neural networks," or CNNs, to identify construction.
The programs will be able to differentiate excavation, foundation, framing, unfinished roof, and finished roofs, per the Census. It is expected to be able to distinguish between single-family detached, single-family attached, multi-family, and non-residential.
This data will be combined with conventional data Census received from local government offices to give a clearer picture of home construction data moving forward.
The bureau is only using the experimental method in District 7, which includes Texas, Oklahoma, Arkansas and Louisiana.
Data from the district shows 14,100 new housing starts in November.
"This product demonstrates a new computer-vision based approach to monthly economic measurement that can improve geographic granularity and accuracy while also reducing respondent burden," the release said.