Play Video

GEOSeries episode 3 - Leveraging Artificial Intelligence To Create Global HD Vector Maps

Ecopia Global Feature Extraction (GFX) Powered by European Space Imaging is a unique partnership that utilises the freshest, highest quality satellite imagery along with the most advanced artificial intelligence from Ecopia.AI (Ecopia) to offer accurate geospatial feature extraction at continent-wide scale. It takes away the challenge of obtaining up to date VHR imagery and manually creating polygons by hand – both of which are labour intensive and can be quite costly. With 13 core features eligible for extraction, users receive comprehensive, up-to-date and accurate land cover maps with zero in-house GIS work and with automation-level speed and efficiency.

In this GEOSeries episode, Ecopia and European Space Imaging will discuss the many benefits of GFX and the solutions it provides. From land-cover capabilities for maintaining smart city digital twins, impervious surface mapping for water drainage assessment, detecting and analysing urban sprawl and flood risk modelling, and the software’s newest application – solar panel identification to aid tax rebates and provide accurate assessments for insurance purposes. Highlighting real use cases to demonstrate the power of these datasets, we will dive deeper into the software and how both government and commercial customers can benefit from this advanced product.

Including Special Guests:

Vitor Pessoa Colombo, PhD, The Swiss Federal Institute of Technology in Lusanne (EPFL)
Vitor is an architect. He grew up in Brazil before moving to Switzerland, where he graduated from École Polytechnique Fédérale de Lausanne (Bachelor of Science) and Accademia di Architettura di Mendrisio (Master of Science). Since 2016, he has conducted research on urbanization processes in cities of the global South, focusing on habitat and health issues related to rapid urbanization. His work is largely based on geographic information systems (GIS), including participatory mapping methods based on mobile data collection tools. Throughout his work, Vitor has developed cross-sector partnerships, collaborating with different research institutes, non-governmental organizations and the private sector. Today, Vitor conducts a PhD at École Polytechnique Fédérale de Lausanne (EPFL) : his project investigates the impact of urban morphology on the risk of diarrheal diseases in Abidjan and Nairobi, two cities marked by rapid and largely informal urban growth. He is also a lecturer at EPFL, where he teaches the Bachelor degree course “Urban Planning in the Global South”.

Nate Herold, Physical Scientist, NOAA’s Office for Coastal Management (OCM)
Nate has been with NOAA for 17 years and is located in Charleston, South Carolina where he is responsible for leading NOAA’s national land cover mapping and monitoring activities through its Coastal Change Analysis Program, or C-CAP. He is an active member and the former chair of the Multi-Resolution Land Characteristics (MRLC) Consortium, the coordinating body for federal land cover mapping in the U.S.

More Videos

Scroll to Top

Architecture of ResNet34-UNet model

UNet architecture for semantic segmentation with ResNet34 as encoder or feature extraction part. ResNet34 is used as an encoder or feature extractor in the contracting path and the corresponding symmetric expanding path predicts the dense segmentation output.

Architecture of VGG16-UNet model

UNet architecture for semantic segmentation with VGG16 as the encoder or feature extractor. VGG16 is used as an encoder or feature extractor in the contracting path and the corresponding symmetric expanding path predicts the dense segmentation output.

Architecture of ResNet34-FCN model

In this model, ResNet34 is used for feature extraction and the FCN operation remains as is. The feature of ResNet architecture is exploited where just like VGG, as the number of filters double, the feature map size gets halved. This gives a similarity to VGG and ResNet architecture while supporting deeper architecture and addressing the issue of vanishing gradients while also being faster. The fully connected layer at the output of ResNet34 is not used and instead converted to fully convolutional layer by means of 1×1 convolution.

Architecture of VGG16-FCN model

In this model, VGG16 is used for feature extraction which also performs the function of an encoder. The fully connected layer of the VGG16 is not used and instead converted to fully convolutional layer by means of 1×1 convolution.

X