An overview of ArcGIS Image Server
ArcGIS 10 is the last release of the stand-alone ArcGIS Image Server product. The image service definition (.ISDef) has been replaced by an improved geodatabase data model—the mosaic dataset—which can be published as an image service using the ArcGIS Server Image extension.
ArcGIS Image Server allows you to manage and process huge volumes of raster data and provide enterprise-wide access to an infinite number of image datasets within GIS, CAD, imaging, and Web applications. Traditionally, image processing and distribution have been considered two separate stages in image exploitation. This separation causes data redundancy, explodes volumes of data, and hinders efficient data management. With ArcGIS Image Server, these two stages can be combined, enabling the administrators of data to maintain only the primary imagery with multiple image products created instantly, on the fly, as required by users. By improving on the traditional workflow for image management and distribution, ArcGIS Image Server can reduce costs and improve efficiency by
- Simplifying the process of serving large volumes of image data
- Eliminating the need for data storage at different stages of processing
- Reducing the delay from image acquisition to dissemination
- Improving image quality by reducing the subsampling
- Supporting many file formats
- Providing multiple representations of the same base data
ArcGIS Image Server enables the creation of image services that can include large numbers of image datasets in different formats and projections and at different resolutions. One of the key features of ArcGIS Image Server is that it supports image data in its native format and does not require a special format to be created. The data may be a preprocessed product, such as tiled orthoimagery; semiprocessed data, such as georeferenced but overlapping orthoimagery; or primary data, such as raw scanned frames or satellite images. The data can be in different compressed or noncompressed file formats or in a DBMS. ArcGIS Image Server supports all ESRI-supported raster formats, such as geodatabase rasters, TIFF, and JPEG 2000, as well as specific vendor outputs, such as QuickBird, USGS DEM, and MATCH-AT. Utilizing the server's processing power, multiple image products can be generated from a single data source, each with different radiometric processing, geometric processing, mosaicking rules, and compression options.
If you are creating a new image service definition, it is recommended you consider creating a mosaic dataset instead. The mosaic dataset is an improved data model for serving content similarly to an image service definition. To learn more, see What is a mosaic dataset.
When ArcGIS Image Server is registered with ArcGIS Server, image services are published through ArcGIS Server to direct connect network clients (using remote procedure calls [RPC]); to Web clients (using hypertext transfer protocols [HTTP]); and as Open Geospatial Consortium, Inc. (OGC), Web mapping services (WMS) or OGC Web coverage services (WCS). Image services are also accessible using Simple Object Access Protocols (SOAP). Clients can connect directly through ArcGIS Desktop applications and third-party applications, such as AutoCAD and MicroStation, or using Web clients such as ArcGIS Explorer. Direct connect client can also be served directly by ArcGIS Image Server stand-alone.
Clients connect to ArcGIS Server and see each image services as single virtual mosaics, although it could be created from large numbers of individual images. As the client pans or zooms around, the server extracts and processes the required data, providing a final mosaicked image to the client application. Both the data access and processing is highly optimized, making the time from request to image display nearly instantaneous.
ArcGIS Image Server is not a GIS or an image analysis application but the backbone for georeferenced imagery required by many applications. Scalable and designed to handle thousands of images simultaneously, ArcGIS Image Server can handle huge quantities of images, now and in the future.
Sample applications
ArcGIS Image Server can be used to exploit the high value of raster datasets for a range of applications including
- Municipalities and utilities, where a large number of users need access to imagery from different applications
Municipalities and utility companies often have large volumes of orthorectified imagery in different resolutions, in different projections, and from different sources. These images need to be provided efficiently to different users both within the organization and to the public. With ArcGIS Image Server, it is possible to quickly create and maintain different image services that are optimized for the requirements of different applications. For example, image services can be provided to planning departments at different resolutions than to the public.
- Data acquisition and provisioning organizations that need to manage, process, and distribute huge quantities of imagery
Such organizations may need to keep catalogs of a huge number of images and provide access to subsets of the images as services or quickly create products for distribution. With ArcGIS Image Server, not only can the large quantities of image data and their attributes be managed easily, but the required products can be defined as dynamic services available for immediate access or processed to a product for conventional distribution.
- Environmental organizations that need to process, manage, and compare many different types of imagery
In environmental applications, there are often requirements to handle many different data sources with different dates, spectral characteristics, and resolutions. With Image Server, such imagery can be accessed as multiple services for viewing in a GIS or for image analysis applications.
- Archiving applications, where large quantities of imagery need to be scanned and cataloged, then accessed in both original and georeferenced form
For the archiving of imagery, very large numbers of images are available that are difficult to preprocess into a single product, as many parameters, such as georeferencing, will only be available later as the need arises or when more accurate control or georeferencing models become available.