When it comes to Big Data, aerial and satellite imagery is as big as it gets. But with the rise of cloud computing, and new tools built specifically to take advantage of its capabilities, the task of turning all that data into information is getting a lot easier.
Managing and publishing large volumes of aerial and satellite imagery has never been an easy task. Legacy desktop systems built primarily to handle vector data struggle to efficiently process imagery, and often fail completely with even modestly large data sets. Storage often becomes a maintenance and backup nightmare. Cloud computing offers obvious advantages over traditional methods of image processing and storage. Primarily:
- Low storage/management costs
- Flexible, dynamic processing resources
- High availability/reliability
But many challenges still remain. Most existing solutions for dealing with imagery are desktop platforms that do not translate well to cloud infrastructure. Security is an issue, since cloud computing often means placing data on servers that are not under the full control of the stakeholders. And then there is the problem of moving all that data around.
CubeWerx OpenImageMap is a cloud based solution for these problems, built from the ground up with an eye to using Amazon’s industry leading web services and platform. Amazon was an easy choice, based on their mature infrastructure, global reach and highly competitive pricing. These advantages transfer well to our own clients.
OpenImageMap is easy to deploy on Amazon’s servers. All functions are accessible through simple web consoles. We can deploy an installation using subscription pricing at a client’s request and hand over the keys, or we can provide a fully managed service. The system is also available as a paid AMI in Amazon’s Marketplace should clients wish to install the platform on their own Amazon infrastructure and pay by the hour.
Getting Data In
The most efficient way to store data in the Amazon Cloud is the S3 service. S3 provides triple redundancy backups, high performance access, and high service availability, all at very low storage costs. Many utilities exist to move data from local storage into Amazon’s S3 buckets. OpenImageMap can also upload directly into S3 through its web console. This is normally sufficient for small to medium installations. Should clients require a more robust solution for very large volumes of data, Amazon provides several solutions, including Amazon Direct Connect and Amazon Snowball.
The OpenImageMap web console is extremely intuitive. In most cases users can be up and running, building their own web services in a matter of minutes. No extensive training is required. CubeWerx support staff is on hand to answer any questions that may arise. Once data is uploaded to S3, a few mouse clicks are all it takes to start the publishing process. The system quickly creates overviews and image pyramids for rapid access. Once this is complete, the data is automatically available through a full complement of web APIs, including a suite of OGC Services (WMS, WMTS, WCS), Google Maps, Google Earth, Bing Maps, Cesium, OpenLayers and others. There are built-in example viewers to give developers a quick-start to integrating the services into their own applications.
Security is built in, using Secure Socket Layer (SSL) for all communications. The platform includes a sophisticated role-based authentication and access control dashboard that lets data providers grant as much or as little access to their data as they require. Content may be restricted by data set, geographic region or data resolution. Deployment on Amazon’s servers also means that clients inherit the considerable security infrastructure that Amazon uses to protect their resources.
Pushing Services to S3
S3 provides more than just great storage for source imagery. Its architecture also allows resources that are hosted there to be accessed by millions of users with little degradation in performance. Snapshots of services managed through OpenImageMap may be pushed to S3 for wider availability and scaling purposes. These snapshots may be limited to a region of the base data set, or offer a restricted resolution if required. This feature can be combined with Amazon CloudFront to further increase performance and reduce costs.
The platform provides the capability to download multi-resolution (tiled) snapshots of any data set. These offline maps can be used in the field or behind network firewalls where access to the OpenImageMap service is not available. The downloads, which may be limited in scope by region or resolution, come with fully functional viewers that require no access to the Internet.
Managing imagery does not have to be a daunting or difficult task. The benefits of cloud computing have been fully realized in Amazon’s web services. Combined with an innovative platform like OpenImageMap, all the heavy lifting is done behind the scenes. Uploading data and publishing a service can be done in just a few minutes, by someone with no specialized training or skill sets. Clients receive all the benefits of a large data center with none of the headaches of managing such complex systems. Costs are kept low by Amazon’s highly competitive pricing and CubeWerx subscription-based access to services. It is an easy win for anyone looking to publish imagery on the Web.