LiDAR Server - serving LIDAR data via the Internet

During the ILMF 2008 conference that was held in Denver, QCoherent released a prototype of a web service that allows users to visualize as well as work with LIDAR data hosted on a central server.

The service which is currently called "LiDAR Server" (http://www.lidarserver.com/) is running off of a 1.3 GHz Celeron D processor and 1GB of RAM machine. LiDAR Server allows the user to visualize LIDAR data using TINS and even allows you to analyze the data using a profile viewer. In addition the service also allows you to extract derivative products from the point cloud - such as contours, TINs, etc.

LidarServer is still a prototype and is sort of clunky - but the concept is truly ground breaking. The idea of hosting of LIDAR data on a central server is not new - USGS Click is one such service. But being able to interact with the data itself and in addition to be able to save derivative products from the original data - is new and I don't think has been done before.

Such a service would be of enormous use to government users - such as cities or the forest service - who might want their data to be available to a wide range of users. It also allows the users of such a service to quickly disseminate their LIDAR data as soon as its available to them. Finally being able to use the data from within a web browser makes the data accessible to even those who don't have a GIS application such as ArcMAP or QTModeller.

My thoughts on improvements to this service:

• Viewer: Currently the view into the data is static. Each click requires the page to be reloaded with a new set of images.
  • The interaction with the viewer should be more in tune with Web2.0. It should be dynamic, and data should automatically get cached behind the scenes - so that I can move around in the data just as I would with Google Maps.
  • It would be even more awesome if QC would implement their viewer much like Microsoft's Virtual Earth 3D viewer. This would allow one to work with their data in a 3D environment, instead of just a 2d view.
• Web services: The way to go forward in my mind is to actually break this technology into 2 pieces.
  1. A web page based viewer.
  2. Actual web services that would allow developers to interact with the data via REST (or SOAP)

Implementing LiDAR Server as these two separate components will then allow other developers to create more plugins based on the web service. The web service would be the key to opening up the data to a myriad other uses. The LIDAR data could then be consumed within applications such as ArcMAP or QTModeller in addition to the web page based viewer for visualization. Applications such as LIDAR Analyst could directly download the data and generate derivative products and upload it back to the server. One might even be able write plugins that will allow for visualization of the data within Virtual Earth or Google Earth.
In addition if the web-service can serve out images (just as it does now), it would be possible to load the data directly into Google Maps (or Google Earth) via image overlay tiles.

LiDAR Server is definitely one of those technologies that makes me exclaim "Now why didn't I think of that!"

LiDAR Server showing an overview of the LIDAR data, an in-depth view of the data as well as a profile of the data.