提供: OSGeo.JP Wiki
2009年8月27日 (木) 20:15時点におけるKayama (トーク | 投稿記録)による版

移動: 案内検索


地理情報システム (GIS)[#!mitchel05!#]<A NAME="tex2html1" HREF="footnode.html#foot16"><IMG ALIGN="BOTTOM" BORDER="1" ALT="[*]" SRC="/usr/share/latex2html/icons/footnote.png"></A>とは空間データを作成、 表示、検索、解析するためのソフトウェアのことです。空間データはある要素の空間位置 に関する情報を持っています。これは緯度、経度のような空間座標を使用することを 意味します。空間データは同じ意味で別の言い方がいくつかあります。地理データ、 GISデータ、地図データ、位置データ、座標データ、地理空間データなどです。
空間データを利用するアプリケーションには様々な機能があります。地図作成機能が 空間データ利用に最もわかりやすい仕事でしょう。地図プログラムは空間データに アクセスしてそれを描画して可視化します。通常はコンピュータの画面や紙を利用して 地図を表示します。 アプリケーションでは静的な地図(単純なイメージ)かダイナミックに表示を変えられる 地図を表示します。ダイナミックな地図はデスクトッププログラムかWebページを使って 地図のみかけを変えることができます。
<P> Many people mistakenly assume that geospatial applications just produce maps, but geospatial data analysis is another primary function of geospatial applications. Some typical types of analysis include computing: <P>

  1. distances between geographic locations
  2. the amount of area (e.g., square meters) within a certain geographic region
  3. what geographic features overlap other features
  4. the amount of overlap between features
  5. the number of locations within a certain distance of another
  6. and so on...

<P> These may seem simplistic, but can be applied in all sorts of ways across many disciplines. The results of analysis may be shown on a map, but are often tabulated into a report to support management decisions.

<P> The recent phenomena of location-based services promises to introduce all sorts of other features, but many will be based on a combination of maps and analysis. For example, you have a cell phone that tracks your geographic location. If you have the right software, your phone can tell you what kind of restaurants are within walking distance. While this is a novel application of geospatial technology, it is essentially doing geospatial data analysis and listing the results for you.


Why is all this so new?

<P> Well, it's not. There are many new hardware devices that are enabling mobile geospatial services. Many open source geospatial applications are also available, but the existence of geospatially focused hardware and software is nothing new. Global positioning system (GPS) receivers are becoming commonplace, but have been used in various industries for more than a decade. Likewise, desktop mapping and analysis tools have also been a major commercial market, primarily focused on industries such as natural resource management.

<P> What is new is how the latest hardware and software is being applied and who is applying it. Traditional users of mapping and analysis tools were highly trained GIS Analysts or digital mapping technicians trained to use CAD-like tools. Now, the processing capabilities of home PCs and open source software (OSS) packages have enabled an army of hobbyists, professionals, web developers, etc. to interact with geospatial data. The learning curve has come down. The costs have come down. The amount of geospatial technology saturation has increased.

<P> How is geospatial data stored? In a nutshell, there are two types of geospatial data in widespread use today. This is in addition to traditional tabular data that is also widely used by geospatial applications.


Raster Data

<P> One type of geospatial data is called raster data or simply "a raster". The most easily recognised form of raster data is digital satellite imagery or air photos. Elevation shading or digital elevation models are also typically represented as raster data. Any type of map feature can be represented as raster data, but there are limitations.

<P> A raster is a regular grid made up of cells, or in the case of imagery, pixels. They have a fixed number of rows and columns. Each cell has a numeric value and has a certain geographic size (e.g. 30x30 meters in size).

<P> Multiple overlapping rasters are used to represent images using more than one colour value (i.e. one raster for each set of red, green and blue values is combined to create a colour image). Satellite imagery also represents data in multiple "bands". Each band is essentially a separate, spatially overlapping raster, where each band holds values of certain wavelengths of light. As you can imagine, a large raster takes up more file space. A raster with smaller cells can provide more detail, but takes up more file space. The trick is finding the right balance between cell size for storage purposes and cell size for analytical or mapping purposes.


Vector Data

<P> Vector data is also used in geospatial applications. If you stayed awake during trigonometry and coordinate geometry classes, you will already be familiar with some of the qualities of vector data. In its simplest sense, vectors are a way of describing a location by using a set of coordinates. Each coordinate refers to a geographic location using a system of x and y values.

<P> This can be thought of in reference to a Cartesian plane - you know, the diagrams from school that showed an x and y-axis. You might have used them to chart declining retirement savings or increasing compound mortgage interest, but the concepts are essential to geospatial data analysis and mapping.

<P> There are various ways of representing these geographic coordinates depending on your purpose. This is a whole area of study for another day - map projections.

<P> Vector data takes on three forms, each progressively more complex and building on the former.


  1. Points - A single coordinate (x y) represents a discrete geographic location
  2. Lines - Multiple coordinates (x1 y1, x2 y2, x3 y4, ... xn yn) strung together in a certain order, like drawing a line from Point (x1 y1) to Point (x2 y2) and so on. These parts between each point are considered line segments. They have a length and the line can be said to have a direction based on the order of the points. Technically, a line is a single pair of coordinates connected together, whereas a line string is multiple lines connected together.
  3. Polygons - When lines are strung together by more than two points, with the last point being at the same location as the first, we call this a polygon. A triangle, circle, rectangle, etc. are all polygons. The key feature of polygons is that there is a fixed area within them.