A Fast Search Mechanism for Multiple Raster Layers in MapWindow Open Source GIS

Session Type: 
Mr Ahmad Aburizaiza, George Mason University
Dr Matthew Rice
Mr William Hammill

GIS raster data is a representation of cells or pixels arranged in rows and columns, where each cell signifies an attribute value representing different data types, such as slope, aspect, elevation, or temperature. Identifying and selecting the cells of a specific attribute value in the GIS programming world can be done fast in a single GIS raster layer. A problem might rise when the GIS programmer attempts to select a combination of multi values in exact cell locations in multi raster layers on top of each other. The more raster layers included for search, the longer time needed to obtain results and the more space needed for storage. This paper represents an innovated methodology in GIS programming, using the MapWindow open source GIS standalone development paradigm, to improve the searching speed for a distinct composite of multi values of identical cell locations in multi raster layers.The paper explains the development evolvement through four different approaches of searching multi raster layers. Each approach is an advancement of the previous approach in terms of speed and storage capacity. A timer tool was developed to calculate the mean time of running thirteen different queries against a small multi raster dataset and a large one, using the first basic approach and the fourth fastest approach. The results of the timer tool represents a significant improvement of searching speed, approximately twelve times faster. In addition, statistical analysis testing was applied to generate the percentage of agreement between the results of approaches one and four.