Geodatabase Rasterization

How is Raster Supported in ArcGIS

For Raster deployment/storage/management:

• raster provisioning

• within a geodatabase(when you want to manage rasters)

  • raster datasets

  • mosaic dataset

Advantages of Raster Provisioning

simplicity flexibility, transparency, performance, portability, cost effectiveness

Advantages of geodatabase raster datasets

centralized management, advanced querying, scalability, metadata management, compression and storage optimization, supports complex workflow acros the platform

What is Mosaic Dataset?

Collection of raster datasets(i.e., images or dems) stored i as a catalog and viewed/access as a single mosaicked image or individual images (rasters)

what do mosaic datasets consist of?

• A catalog providing the source of pixels, footprints of rasters

• feature class that defines the boundary

• a set of mosaic rules used to dynamically mosaic the rasters

• properties used to control mosaicking and nay image extraction

• colour correction table defining colour mapping for each raster

Pros of Mosaic Datasets

efficient data management, on the fly processing, improve performance , supports multiple projections and formats, centralized data source, customization and sharing

Themaitc Rasters are…

data sets that are derives data that can be used for modeling and analysis(raster datasets that group cell values into classes to represent features like land use, soils, or elevation)

Can be created from:

• interpolation from sample point

• remote sensing based approaches

• conversion of vector data

Rasterization

the process of converting a vector to a raster dataset

• drape a fishnet containing square cells over the entire study area each cell given a code or values corresponding to an attribute type

Rasterization: Centroid based

assignment based on data vales at the cells geometric center

• goo when there is little localized variability in attribute

• generally only for raster encoding or polygons

• problems: when a minority type is assigned location of object relative to centroid

Rasterization: Most Important Type

• output cell value dependent on whether the most important feature occupies this cell

• best for coding points and lines

Rasterization: Predominat Type Encoding

• proportion of feature within a cell is considered

• good for categorical data

• coarse resolution and high variability