Creating a raster format

In ArcGIS Image Server additional raster formats can be added by implementing an ArcGIS custom raster format dynamic link libraries (DLL). Since ArcGIS Image Server framework is designed to work with all ArcObject rasters (including user defined custom raster formats) this new format added to ArcGIS will be accessible by ArcGIS Image Server through ArcObjects via raster data objects (RDO). Currently AORaster does not support reduced resolution layers.

Based on the DLL technology, raster functionality in ArcGIS has been designed and implemented to provide flexibility in extending ArcGIS in reading and writing raster formats that are not supported out of the box. In other words, users can customize their own raster format for ArcGIS by creating a customized format DLL and then plugging this DLL into ArcGIS. After proper configuration, raster data in this format will be supported in ArcGIS as other default raster formats.

ArcGIS is built in a component object model (COM)-based environment. The raster functionality in ArcGIS is supported through RDO, which provides unified data access, management, and display functionality to various raster data formats. ArcMap, ArcCatalog, or any upper level application makes calls to RasterLayer, Raster, RasterDataset, or other COM objects; the calls are sent to the ERDAS libraries; then a corresponding format DLL will be linked; and, finally, this linked format DLL handles pixel reading or writing of this particular format. For better understanding of the overall picture, the following figure depicts the roles that RDO and the format DLLs play in ArcGIS:

With this unified structure, RDO and the upper level applications access raster data in the same manner. The differences among the various raster formats will be handled by each individual format DLL.

Many raster formats can be simulated using the RAW raster format. Prior to creating a new raster format a developer should research if the format can not be specified using RAW. If so may be advantageous to create a raster type that creates a raster process definition including a fully defined RAW raster format. The following is a definition of the nodes within the Format Info section for RPDef generator that are used when \raster\formatname is set as RAW.

Node	Type	Default	Description
NCols	Int	0	Number of columns of pixels in this raster.
NRows	Int	0	Number of rows of pixels in this raster.
SkipBytes	Int	0	Skip bytes from the start of the file.
PixelType	String	Byte	The data-type representing a band. As Byte:Unsigned 8-bit Integer, UInt16:Unsigned 16-bit Integer, Int16:Signed 16-bit Integer, UInt32:Unsigned 32-bit Integer, Int32:Signed 32-bit Integer, Float32:32-bit Floating-point, Float64:64-bit Floating-point, CInt16:Complex 16-bit Integer, CInt32:Complex 32-bit Integer, CFloat32:Complex 32-bit Floating-point, and CFloat64:Complex 64-bit Floating-point.
BitDepth	Int	8	Number of bits used by a pixel band.
NBands	Int	1	Number of bands in the Raster.
Interleave	String	BIP	Interleave of pixels: BIP—Band interleaved by pixel, BIL—Band niterleaved line, and BSQ—Band sequential.
ColorSpace	String	1	Colorspace that defines the relationship between the bands of the pixel. As 0:Grayscale, 1:RGB, 2:CMYK, 3:YCbCr, 4:YUV, 5:LHS, 6:YCCK, 7:Palette, and 255:Unknown.
TileSize	Int	0	If the raster is tiled defines the tile size (assumed square). 0 if not tiled.
ByteOrder	String	LSB	Byte layout for mulit-byte pixel data. As LSB:LSB first and MSB:MSB first.
TreatAsFloat	Bool	False	If set to True then the raster will be assumed to be a Float and handled as such.
ApplyOffsetScale	Bool	False	If set to True the value will be multiplied by 214, which is how floats are handled as signed ints within ArcGIS Image Server.
FloatOffset	Float	0	Offset to be applied. By default 0.
FloatScale	Float	214	Scale to be applied. By default 214.
ApplyConvert	String	None	Defines if the pixels should be converted on reading. As None, int16to32:16-bit Int to 32-bit Int, and uint16to32:16-bit UInt to 32-bit Uint.