Powering the future of location intelligence: Earth imaging basics

By: Josh Winer, DigitalGlobe Senior Sales Manager

We’ve all interacted with imagery of our planet in one form or another. Whether on our mobile devices, televised news coverage, in-dash navigation screens, even gaming platforms – viewing the earth from above has become commonplace.

Technology advancements, largely enabled by cloud storage and compute capabilities, have changed the way we access and interact with geospatial data. Streaming tile services rapidly serve up only the pixels we want to see. Artificial intelligence, computer vision, and machine learning have made it possible to extract information from imagery at scale – supporting novel use cases every day.

These capabilities help us develop a better understanding of our planet and have become critical for business growth, asset security, and disaster response. But how do we capture the geospatial data required to remotely guide autonomous vehicles, analyze coastline changes, or monitor assets? Satellites.

Satellite imagery comes in many forms. Object identification and consumer applications often rely on the highest spatial resolution imagery available. Through use of DigitalGlobe’s WorldView constellation, European Space Imaging is able to deliver 30 cm resolution imagery – the highest resolution available from a commercial satellite. At that resolution you can not only count cars from space – but identify direction of travel and vehicle type. Just as with resolution, positional accuracy is critical for applications such as autonomous vehicles, asset tracking and monitoring, and object identification. European Space Imaging’s industry leading accuracy is able to pinpoint the location of its imagery to within less than 5 m of its known location on earth. Balancing these unique capabilities with frequent revisit and looking beyond the visible light spectrum, opens a world of opportunity.

These advances in satellite imagery have enabled critical change detection applications for everything from environmental and population shifts to illegal structure identification and military mobilization. Frequency of collection keeps our navigational apps up to date. Elevation data (3D models) and building footprints provide crucial information for regulated industries, map-based consumer services, and global development organizations. Benefits of these advances transcend almost all industries.

It gives a whole new meaning to the phrase, a picture is worth a thousand words. A single satellite image contains vast amounts of information that can be used for various purposes. We refer to this information as geospatial big data, and geospatial big data is what transforms satellite imagery into actionable location intelligence . Even if you haven’t heard of location intelligence, it’s likely you’re already using it.

But not all commercial satellite imagery is equal. Different spatial, spectral, and temporal resolutions provide different types of information. From acquisition to analysis, we’ve developed a foundational video series to help you understand these differences and get a better idea of what your potential applications may require.

This article was written by Josh Winer, DigitalGlobe Senior Sales Manager as part of the Earth Imaging Basics Series. To see the original article, click here>>

Share on facebook
Share on twitter
Share on linkedin

Related Stories

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