Sunday, May 18, 2008

ERDAS Image Manager WCS

I'd like to present the utility and feature "highlights" of the ERDAS Image Manager Web Coverage Service (WCS). For proper delivery to image exploitation clients, the Image Manager has the requirement to deliver the original pixel values and bit depth of gridded data to client applications for analysis. A WMS provides a "portrayed" map of the gridded data, so it effectively stylizes and delivers an "encoded" image focuses at a visualization experience for the user, so the pixel values are "changed", as well as the bit depth.

An example Image Analysts goal would be to "calculate NDVI for a user specified area".

The Lieca Image Manager functionally works just like a WMS, it has a "GETCAPABILITIES" request and it lists the aggregates of data based on the username/password that has requested the capabilities document. The difference is that for the GET requests, the WCS will deliver the raw pixels of the gridded data in the client specified encoding format.

For example, I can have modeled 50 LANDSAT image scenes with coverage over the southeast US into a single "South East US" aggregate and exposed as a single WCS layer. The aggregates extent will be the union of all the 50 LANDSAT scene extents. With the WCS, I can request any extent, spatial reference system, resolution, number of bands and encoding format for the aggregate. A sample use case would be to request the extent for Atlanta, GA in WGS84, bands 4,3 (for NDVI calculation), at 1:1 (full resolution) and in GeoTIFF format.

The WCS will perform "mosaic on the fly" for the extent requested, reproject the data if required and "encode" and deliver the data through the WCS! As a user, there is no need to have knowledge about the imagery that composes the aggregate and how the aggregate is composed, you simply only need to know where you want to create information products!

IMAGINE will manage the WCS just like any other image format supported, so the use of the WCS is "seamless" in the client. As a user, I simply have to go to the area and produce the NDVI product wherever and whenever I want. I don't have to know anything about the data, just where I want to product products.

So how is this benificial?

1. Increase productivity by centralizing and aggregating gridded data access through web services
2. Only provide imagery to people who have been given the rights to access the data (security!)
3. Streamline the workflow for generating data products that cover possibly many gridded datasets
4. Area based product production, rather than file based
5. Deliver pixels in any parameter combination (extent, SRS, resolution, format and band combination)
6. INTEROPERABLE!!!! Deliver pixels to any WCS client!

No comments: