19-10-2012, 10:59 AM
Bringing Raster GIS To The District of North Vancouver
Bringing Raster GIS.pdf (Size: 701.4 KB / Downloads: 33)
Abstract
The District of North Vancouver’s GIS Department has sponsored this project to research software and data
solutions for integrating high-resolution satellite imagery with their current system. This report is the
culmination of approximately five hundred hours of technical, procedural and economic research in this
area.
The data ultimately chosen for this project was IKONOS four-meter resolution multispectral imagery
covering the Eastern portion of the District. The software provided to the team was Idrisi, though several
other remote sensing software packages were also evaluated. The project team developed and documented
numerous methodologies for image analysis and classification, providing a strong rationale for the district
to pursue this technology further
Introduction
Recent improvements in satellite image quality and availability have made it possible to
perform image analysis at much larger scale than in the past. This will likely lead to a
much wider use of satellite imagery at the regional level as a reliable source of timely and
accurate spatial data. Being aware of this tremendous potential, the District of North
Vancouver’s GIS Department has sponsored this project to research software and data
solutions for integrating high-resolution satellite imagery with their current system.
Specifically, the team approached the project in 4 logical stages:
Software Research: The software provided for the project was Idrisi32. All relevant
tutorial modules were completed, including data import, image preprocessing, image
analysis and classification, and data export.
Data Research: A significant portion of the project involved researching possible sources
of imagery. Internet and literature reviews, and communication with various vendors
were summarized, and the findings and recommendations were presented to the sponsor.
This document can be found in section 11.4 of the Appendix.
Image Processing: This stage was the technical portion of the research project, where the
imagery purchased by the DNV based on our recommendations was analyzed using Idrisi
and several other commercial software packages.
Trends in Remote Sensing
The ownership and operation of remote sensing satellites is currently undergoing a
process of rapid commercialization. Imagery is becoming available at a higher
resolution, faster delivery, and lower cost than in the past, to the point where a near-real
time satellite image is less expensive than an aerial photograph of a similar resolution
(Konecny, G., 1997). At the same time, the computer price-performance index is
doubling every 18 months, allowing for faster storage, retrieval and processing of image
data at a realistic price (Hartnall, 1999). It is reasonable to believe that in the near future
satellite imagery will become an integral part of most medium- to large-scale GIS
organizations.
Data Input / Manipulation
All of the data used in this project was already in a digital format and manual data entry,
such as keyboard and digitizer entry, was not required. Since we were using Idrisi32 as
our GIS software package, all of the data had to be converted to Idrisi32 format. There
were primarily four sources for the data; each is listed below with a description of the
conversion process. More descriptive step-by-step procedures can be found in Appendix
11.2.