SPACENET DIGITALGLOBE MANUAL
SPACENET DIGITALGLOBE SERIES
SPACENET DIGITALGLOBE DOWNLOAD

Understanding the effect this has upon algorithm performance has applications not only in remote sensing analysis, but also in the general computer vision domain. The exact same geography looks very different depending upon the look direction and angle. Negative numbers correspond to South-facing looks, positive numbers to North-facing looks.

The look angle is shown on the left, along with a bin ID: Nadir, ≤25 degrees off-nadir Off-nadir, >25 and <40 degrees very off-nadir, ≥40 degrees off-nadir. Sample images from the SpaceNet Multi-View Overhead Imagery (MVOI) dataset (described below). The satellite took 27 images, ranging from 7 degrees to 54 degrees off-nadir, including both North- and South-facing looks:

SPACENET DIGITALGLOBE SERIES

The dataset is derived from a unique series of collections from a DigitalGlobe WorldView-2 satellite during a single pass over Atlanta, GA USA.

SPACENET DIGITALGLOBE DOWNLOAD

To explore how much look angle influences model performance, we released the SpaceNet MVOI Dataset, which is open source and freely available on AWS S3 - all you need is an AWS account! Download instructions and additional metadata are available here. The SpaceNet Multi-View Overhead Imagery (MVOI) Dataset To find out how much these factors influence model performance, we need a well-labeled dataset that controls for all variables except look angle. This is obvious in the animation at the top of this story - as the image gets more off-nadir, resolution gets worse. In the SpaceNet Multi-View Overhead Imagery (MVOI) dataset described below, images taken “at nadir” had a resolution of 0.51 m/pixel, whereas resolution very off-nadir dropped to 1.67 m/pixel. Images taken at a greater angle cover more ground, but still contain the same number of pixels, reducing ground resolution.

A building whose roof is visible (though cloaked in shadow) in nadir imagery (left, red arrow) is obscured by the skyscraper in off-nadir imagery (right). Occlusion can make it impossible to see some objects in off-nadir imagery. Why don’t we automatically map after natural disasters?Ī major barrier still exists to automated analysis of overhead imagery in disaster response scenarios: look angle. Alongside these developments, automated mapping challenges have seen steadily improving performance, as evidenced by the SpaceNet competition series: building extraction scores on imagery taken directly overhead improved almost three-fold from the first challenge in 2016 to the most recent challenge (discussed here) at the end of 2018. This is by no means a criticism of the HOT-OSM team or their fantastic community of labelers - they had 950,000 buildings and 30,000 km of roads to label! Even a preliminary automated labeling step, corrected manually afterward, could improve map delivery time.Ĭomputer vision-based map creation from overhead imagery has come a long way as deep learning models have grown from the nascency of AlexNet, implemented before TensorFlow even existed, to today’s advanced model architectures including Squeeze and Excitation Networks, Dual Path Networks, and advanced model training methods implemented in easy-to-use packages like Keras.

SPACENET DIGITALGLOBE MANUAL

But manual labeling is time-consuming and labor-intensive: even with 5,300 mappers working on the project, the first base map of Puerto Rico was not delivered until over a month after the hurricane hit, and the project was not officially concluded for another month. At present, this is done manually by teams in government, the private sector, or volunteer organizations like the Humanitarian Open Street Maps (HOT-OSM) team, who created a base map (roads and buildings) of Puerto Rico after Hurricane Maria at the request of the USA’s Federal Emergency Management Agency (FEMA). roads) needed to coordinate response efforts may be disrupted. Rapid disaster response could be aided by automated overhead image analysis: new maps are often essential after disasters, as infrastructure ( e.g. The current state of affairsĪs computer vision methods improve and overhead imagery becomes more accessible, scientists are exploring ways to unite these domains for many applications: among them, monitoring de-forestation and tracking population dynamics in refugee situations. The key takeaway: though algorithms are fantastic at mapping from “ideal” imagery taken directly overhead, the types of imagery that exist in urgent collection environments - such as after natural disasters - pose a currently unsolved problem for state-of-the-art computer vision algorithms. In this post, we’ll describe the challenges associated with automated mapping from overhead imagery. A series of images taken by DigitalGlobe’s WorldView-2 satellite, from the SpaceNet Multi-View Overhead Imagery (MVOI) dataset described below, illustrating how look angle changes image appearance.

Spacenet digitalglobe