What is raster data for spatial analysis?

In the previous article, we started with map algebra to give you a sensory understanding of space analysis of raster data. So what is raster data and how does raster data represent geographical entities and geographical phenomena in nature?

1. Raster Data Model:

The raster data model divides the geographic space into several rows and columns, which are called a pixel array. The smallest unit is pixel or pixel. The position of each pixel is determined by the row and column number, the value in a cell records the geographical entity of the position or the value of a topic element at this position. The basic structure of raster data is as follows:

Raster unit: the size of the basic unit of the Raster Data mesh, which plays a key role in the resolution and accuracy of the raster data. For example, if the size of a grid unit is 1*1, a grid unit represents 1 m², and its resolution is also 1 meter. The smaller the raster unit, the higher the image resolution, and the more refined the representation of the object.

The following describes integer and floating-point raster units.

1.1 integer raster

An integer raster unit worth raster data is used to store a variety of so-called discrete data or classified data. This type of dataset describes an object that may exist in the location of the raster unit. Let's take a look at two concepts:

ZONE: A zone is composed of connected or unconnected cells with the same space values;

Region (region: a group of adjacent cells with the same space values. To represent an element, a type zone may consist of multiple regions.

Note: The type areas may vary depending on the grid unit's neighborhood definition.

1.2 floating point value Grid

A grid with floating-point raster unit values is used to store so-called continuous data. Such data may also be specified as non-discrete data, domain data, or curved data. These numbers are stored in a format that allows meaningful numbers and a large range of formats. A specific grid unit has different values than any other grid unit in the grid. Therefore, the concept of type zone and region does not exist. In general, attribute tables also exist.

2. Representation of grid data on geographical entities

1) point: a pixel.

2) line: concatenates adjacent pixel sets in a certain direction.

3) surface: a collection of adjacent pixels.

 

3. Value of the raster Unit

Each pixel attribute can have only one value. In fact, a grid may correspond to several different attribute values. How can we set a value?

1) center point method: Obtain the attribute value in the center of the grid as the attribute value of the grid.

2) area Dominance Method: the attribute value of a grid unit is the area owner.

3) Importance method: Take the important attribute value as the raster attribute value. It is used for small objects with special meanings.

4) dominant length method: the value of each grid unit is determined by the attribute of the entity with the longest line segment in the grid.

 

4. Differences between raster data and Vector Data

Raster data can represent both discrete geographical entities and continuous geographical entities. In comparison, raster data is very suitable for spatial continuous data, such as elevation, temperature, meteorology, and environment. The main differences among other vectors are as follows:

5. grid data scale

The maup problem (plasticity area unit problem) also exists in raster data. Because, raster data uses the attribute values of raster cells to represent geographical entities and geographical phenomena. The larger the scale, the finer the representation of geographical phenomena, and the hour of scale, because a grid unit represents a large area, it may cause certain distortion to geographical entities and geographical phenomena. This is why remote sensing images with low resolution are not easy to classify.