Raster Database vs. Vector Database

Read both sides of this long going argument and decide for yourself which database is better!

Raster databases are visually more attractive because they are simple and organized. They are faster in performing operations such as overlaying or creating buffers. They employ the use of pixel processing. All these benefits come at a heavy price i.e. no details and limited accuracy provided by cells of line and point objects, which is better in vector databases.

Although some techniques can used to solve these issues in both databases, they are only applicable in some situations.

Points of Difference in Rasters and Vectors

Earlier versions used 40 acres cell sizes so anything under this size was not included which was very troublesome. Now, smaller cells of 5 meter are in place and since most objects are of 5 meter or more, it is quite a step forward.

In Raster, it is quite difficult to pinpoint the exact location in the cell, as it is hard to identify whether it is the middle or the upper left corner of the cell in question.

Precision

Vector precision is 1/108 of the size of the area while to compare this with raster precision, it would have to be 108/108 which is not possible even with lengthy encoding. The precision may be off about 2 meters or more.

For certain classes of data vector precision is correct, such as those for accurate political boundaries. While some things found in nature like vegetables have hazy boundaries.

Thus, the precision between raster and vector cannot be compared by coordinates but by the cell size of the raster and the uncertain position of a vector object.

image

Speed of computing:

To answer queries regarding overlays and proximity, raster data can be dealt with quickly. Not many calculations are needed and little to no arithmetic computation is required. Thus making the raster system better suited for smaller computers and their development cost is also not much!

On the other hand, vectors are very complex. Complicated problems need to be solved to locate intersecting points between lines and you could say even more complicated algorithms are needed in polygon overlay. The calculation for distances thus might get elaborate. Overall, this slows down the vector system and so you would need to use it in intervals.

Raster storage:

The most elementary raster data storing requires one place, each cell takes up one or two bytes, which is not at all, wise as this reduces the overall information that can be stored into one system. File compression is available to lessen the size of the data stored.

Vector Storage:

While on the other hand, vector storage takes up little space unless the object is of a complex shape. The volume of data also depends on the types of relationships stored in the system, the more relationships stored, the more memory the system uses up.

Raster Sampling:

As raster is regularly, spaced sampling it does not have a lot of information on spatial variation. The data stored in raster are mostly satellite images and since the satellite is not smart, it cannot differ between variations on land.

image

Vector Sampling:

Data collected by the vector system shows more variations in specific areas than in others. This is true for social, economic variations, which are denser in certain areas. It is also true for natural things such as vegetation may appear more defined in some areas than others. Variation is immediate at the end of objects, like at boundaries.

Overall features and entities:

It is difficult to group together a number of cells to make an object in raster. For example, connect cells along a bridge. Raster arranges land in to a specific order while vector places things in any way, giving access haphazard data.

It is easier to route vehicles on a street because operations of objects is easier on vectors.

Raster cells present objects in one whole cell, which means they do not precisely indicate the location of a smaller object like a well! This may present problems if the precise location is needed for an important task.

However, the data in both vector and raster may be accurate while some objects may have misleading coordinates on purpose in order to preserve them.

Combination of both:

A mix of both systems may be used; the data can be stored in one way and processed in a separate way. A way in which both systems can be run alongside each other may be employed for maximum efficiency.

 

 

References:

http://gis.stackexchange.com/questions/7077/what-are-raster-and-vector-data-in-gis-and-when-to-use
http://en.wikipedia.org/wiki/Geographic_information_system
http://www.colorado.edu/geography/gcraft/notes/datacon/datacon.html

Leave A Response

* Denotes Required Field