7819 Rate this article:
4.7

ENVI Deep Learning Identifies Damaged Building Structures in Wake of Nashville Tornado

In the aftermath of a tornado, flood or other disaster, the focus of emergency response agencies is to reach the hardest hit areas first. Speed is of the essence. The damage must be assessed quickly so that search and rescue personnel – along with vital resources such as food, water, and medical aid – can be deployed in a prioritized manner to help those most in need.

Evaluating the extent of destruction at ground level in the midst of rubble and debris is challenging, especially when a large geographic area is involved. Fortunately, imaging systems aboard manned and unmanned aircraft can rapidly acquire high-resolution imagery of the region, often within hours of the event. Assessing damage in these images, however, can be a time-consuming process.

An Artificial Intelligence technology called Deep Learning is changing the paradigm in image analysis by quickly and accurately finding hundreds or even thousands of targets of interest, such as damaged building structures and blue tarps, in imagery data sets.

As a demonstration project, L3Harris Geospatial applied this automated technology to airborne images captured over Nashville, Tennessee, following a deadly March 2020 tornado. Using the ENVI Deep Learning module, which is offered as an extension to ENVI geospatial analysis software, L3Harris technicians identified where building damage had been inflicted by the tornado and charted its precise path of destruction from the aerial photos – in less than two hours.

Assessing Damage with Deep Learning

On Tuesday, March 3, 2020, a tornado with wind speeds reaching 165 mph swept through the Nashville suburbs killing more than 20 people and reducing hundreds of homes and buildings to rubble.

“Identifying building damage with Machine Learning is a challenge because ‘damage’ doesn’t have a specific appearance or shape,” explained Austin Coates, Sales Engineering Manager at L3Harris. “Damage will look different for just about every building or home.”

Finding houses, airports or roadways with Machine Learning is much simpler because each of these features – although different in many ways – have multiple common characteristics, including their shapes and construction materials, for example. But Coates was confident ENVI Deep Learning would pinpoint damaged or destroyed structures thanks to the complexity of its algorithms.

“The Machine Learning algorithms just need to be shown examples of the features to be found in the multispectral image data, and they evaluate the spatial, spectral and textural characteristics of that feature,” he said. “The algorithms use this training information to find other features or objects with the same characteristics.”

Working in ENVI Deep Learning, Coates manually examined small sections of the imagery to detect damaged buildings with his naked eye. He defined these examples, or training sites, by drawing polygons around them. Due to the wide variation in type and degree of structural destruction, he selected more training samples than he would normally.

In setting up Deep Learning training models, it’s standard procedure to pick clusters of examples in several small areas of the overall image. This allows the user to run quick tests on the model to see how it is performing. Coates followed this procedure and learned the algorithms were identifying damaged structures accurately but were also delivering false positives.

“Piles of debris in yards and rock quarries were identified as damaged buildings,” he said.

To further refine his models, Coates obtained open-source building footprint data from Nashville’s online GIS. By importing this data into the ENVI model, he focused the algorithms to look for damage in spaces defined by known building footprints. The result was an extremely accurate map showing the damaged structures in Nashville.

In addition, Coates created a similar, but much simpler, Deep Learning model to find blue tarps in the image. These tarps are traditionally used by FEMA and other relief agencies to cover damaged roofs after a storm. It should be emphasized that blue tarps would not be in place in imagery captured a day or two after the storm, so they could not be used for rapid response planning.

With the building destruction mapped, Coates used an automated ENVI function to indicate the centerline of the tornado’s path. He also drew buffer lines at one-eighth and one-quarter miles from the centerline to show where the heaviest damage likely occurred.

“Information like this is what federal agencies and local governments need to rapidly coordinate their emergency response efforts so that resources are deployed properly,” said Coates.

Coates added that Deep Learning could be applied to extract other information from imagery that could prove helpful to first responders – including the locations of destroyed utility infrastructure or blocked transportation routes.

To learn more about ENVI Deep Learning, visit
www.harrisgeospatial.com/Software-Technology/ENVI-Deep-Learning

DISCOVER MORE

ENVI DEEP LEARNING MODULE

AUTOMATE ANALYTICS

Automate analytics with deep learning for fast, accurate results.

Solve Geospatial Problems with Deep Learning

RECORDED WEBINAR

See how ENVI Deep Learning is used for environmental, natural resource, and disaster response.

SEE ENVI® DEEP LEARNING IN ACTION

Situational Awareness Webinar

SHORT DEMO VIDEO

Real-time, actionable intelligence to relief organizations.

 

 

 

 

 

Please login or register to post comments.