Geographical Data in the Computer

GMR Institute of Technology

Dr. N. Nageswara Reddy

 Data in the Computer
• Computer - Coding of data and computational work
- Binary system
Hierarchy of Data Representation:
• Bit: The smallest unit of data in a computer, representing a binary value of either 0 or 1.
• Byte: A group of 8 bits, which can represent a single character, such as a letter or number.
• Word: A fixed-size unit of data, typically 16, 32, or 64 bits, that a computer processes as a single
entity.
GMR Institute of Technology

• Record: A collection of related fields (such as a name, address, and phone number) treated as a
single unit of data in a database.
• File: A collection of records stored together, such as a document, image, or database table.
• Directory: A folder in a computer’s file system that contains files and possibly other directories
(subdirectories), organizing data hierarchically.
Data Volume:
• Kilobyte (KB): Approximately 1,024 bytes.
• Megabyte (MB): Approximately 1,024 kilobytes or about 1 million bytes.
• Gigabyte (GB): Approximately 1,024 megabytes or about 1 billion bytes.
• Terabyte (TB): Approximately 1,024 gigabytes or about 1 trillion bytes.
 Coding the Basic Data Models

• Vector data structure

- Data structure that uses points, lines or polygons to describe geographical phenomena.
- Complex objects – Arc, Node
- Locational data
- Attribute data – Records or tuples
• Raster data structure
GMR Institute of Technology

- Data structure that uses regular tessellation units such as square cell (pixel) to describe
geographical phenomena.
- Locational data – Direct reference to grid network
- Attribute data
 Database Structures
• Essential feature of any data storage system
- Data to be accessed and cross-referenced quickly
• No single best method
• File and Data Access
- Simple lists
- Ordered sequential files
GMR Institute of Technology

- Indexed files
- Direct file
- Inverted file
 Data file and database
Data file Database
Data are collected for specific Data are generic for the business of
purpose an organization
Collection of records usually of the Collection of interrelated records
same data type and format organized in one or more data files
Data file processing – computer Database processing- Associated with
programming DBMS
Used in support of information needs Used in support of day to day
GMR Institute of Technology

of an ad hoc application operation of business transaction and

decision making

Basic unit of file is data record or tuple- Contains all the information related to entity.

Tuple: It is a set of attribute values that forms a row of relation.

Relation: It is made up of a set of tuples
It is a table of data
Domain: It is the smallest unit of data representation in the relational database
It is set of atomic attribute values
GMR Institute of Technology
 Database structures
• Hierarchical structure
- Data have a parent/child or one- to-many relation
- Areas of different levels of administration
- Taxonomies, classifications
- Hierarchy can be reached using a key (Set of discriminating criteria)
Adv: - Good for data retrieval if the structure of all possible queries is known before
GMR Institute of Technology

hand
-Simplicity and ease of access via keys
Disadv: Large index files- Data redundancy- Storage and access costs
GMR Institute of Technology
 • Network structure
- Linkage between the objects
Adv: -Useful, when relations or linkages can be specified beforehand.
- Data redundancy avoided
Disadv: Database is enlarged by overhead of pointers
GMR Institute of Technology

• Relational database structure

- It stores no pointers and no hierarchy
- Data stored in simple records(tuples) which make up the relations
- Identification codes are used as unique keys to identify the records
- Data extraction – Defining the relation that is appropriate for query
- Rules are encoded in the SQL
- Major tools to handle attribute data of spatial entities in GIS
GMR Institute of Technology
GMR Institute of Technology
 Adv: The structure is very flexible and may meet the demand of all queries
-Addition or removal of data is easy
Disadv: Operations take considerable time if relationships between tables are
complex and number of joins are needed.
Object- Oriented data structure:
- Developed using object oriented programming languages
- Data will be organized around actual entities
GMR Institute of Technology

- Data are defined in terms of a series of unique objects

- Data are encapsulated within an object which is defined by unique identifier
- Query- To send a request(message)
Adv: Efficient structures for organizing hierarchical, interrelated data
Disadv: Establishing the database is time consuming
 District

Is situated

Contains
Type
District User#
service

in
Type Address
Amount
User# Name Address Consu-
Used
mer
GMR Institute of Technology

TS TB Owner

Type Type

Lives at

Used
by
Addres
Buildin Owner Consum User#
s
g er

User #
Relational database for
Utility consumer management
Name TC AU
 Appropriate database structure

• Hierarchical
– Dividing the spatial data into manageable themes
• Network approach
- Ideal for topologically linked vector lines and polygons
GMR Institute of Technology

• Relational approach
- Good for retrieving objects on the basis of their attributes
• O- O approach
- Useful when entities share attributes or interact in special ways.
 Data organization in raster data structures
GMR Institute of Technology
GMR Institute of Technology
 Different ways of creating raster data structure

MAP FILE MAP FILE

Overlay
Point
Point
X
X coordinate
Y coordinate
Y coordinate
Overlay Z value
Z value
Next overlay Z value
Next point
GMR Institute of Technology

Next Overlay
Next point

(a) Each cell referenced directly (b) Each overlay referenced directly
➢ Data Hungry structure ➢Two dimensional matrix
➢ No data on cell size or display symbols ➢Contains list of redundant coordinates
➢ No compression techniques ➢No data on cell size or display symbols
 MAP FILE MAP FILE
Overlay Overlay
Number of rows, N
Title Number of columns, M
scale Cell Size
mapping unit 1 Projection
label X Minimum
display symbol Y Minimum
attribute Pointer to legend
GMR Institute of Technology

set of points
(X,Y coordinate pairs) Sequence of NXM values

Next mapping unit Next Overlay

Next Overlay (d) Each overlay is a separate file with general

header
(c) Each mapping unit referenced directly > More efficient as coordinate values are not
➢ Clumsy for highly variable field data stored for each cell
➢ Compression techniques can be used
 Compact Methods of Storing Raster Data

• Chain codes
- Boundary of the region will be encoded
- Region will be represented in terms of its origin and sequence of unit vectors in
cardinal directions (E=0, N=1, W=2, S=3).
- Compact way of storing data
GMR Institute of Technology

- Useful for converting the raster description of polygons to vector form

- Overlay operations like union and intersection are difficult
- Redundancy - Since boundary between regions stored twice

• Run length codes

- The points in each mapping will be to stored per row from left to right in terms of begin
cell, end cell and an attribute
 GMR Institute of Technology

RASTER MAP
GMR Institute of Technology
Chain Code

Run Length Code

 - Useful in reducing the volume of data that need to be input for large uniform areas
- Not suitable for coding continuous variation
• Block codes
- Square blocks concept will be used to encode the data
GMR Institute of Technology

- Data structure consists of the origin (Centre or bottom left) and radius of the square.
- Medial Axis Transformation
- Efficient for large uniform shape
- Union and intersection can be performed
GMR Institute of Technology
 • Quadtrees
- This will provide an approach to addressing the successively finer levels of information
- Region will be divided successively into quadrant blocks
- The block structure may be described by tree of degree 4, known as quadtree
GMR Institute of Technology

- Each node has four branches – NW,NW, SW and SE

- Leaf nodes corresponds to those quadrants for which no further subdivision is
necessary
- Shape analysis and pattern recognition is not straight forward
- Standard region properties may be easily and efficiently computed
GMR Institute of Technology
GMR Institute of Technology
GMR Institute of Technology

8
7
6
5
4
3
1

2
1
2
3

A
4
5
6
7
8
 Data Organization in Vector Data Structures
GMR Institute of Technology
GMR Institute of Technology
 POINT
Unique Identifier

Type of point Simple

Text
Node
Serial number
XY coordinates
Associated attributes for display and topology

(if simple)
GMR Institute of Technology

Symbol scale
orientation
Pointer to database
(if text) characters scale
font
justification
type style
(if node) symbol pointer to lines
angles of the junctions

Non-graphic attributes

Vector Database Structure of a simple point entity

 • Point Entities
• Line entities
• Networks
• Polygon entities
GMR Institute of Technology

(a) Simple polygons

Disadv: - Lines of adjacent polygons must be stored twice
- Slivers and gaps among the common boundary -
There is no neighborhood information
- Islands are impossible to represent
- Difficult to check the dead ends and weird polygons
(b) Fully topological polygon network structure
GMR Institute of Technology
 Steps:
(a) Linking arcs into boundary network
(b) Checking polygons for closure
(c) Linking lines into polygons
- Creation of envelop polygon
GMR Institute of Technology

- Creation of individual polygon

(d) Computing polygon areas
(e) Associating non-graphic attributes to the polygons
Adv: Fully integrated; free from gaps and slivers
All kinds of neighborhood analysis is possible
Locational accuracy
Continent- island nestings unlimited
GMR Institute of Technology
 Choice between Raster and Vector Data structures

Vector
• Adv: • Dis adv:
- Good representation of entity data - Complex data structure
models - Difficult overlay processing
GMR Institute of Technology

- Compact data structure - Expensive data collection

- Topological processing- Good for - Use of expensive technology
network analysis
- Difficult presentation of spatial variability
- coordinate transformation and rubber
- Spatial analysis within the basic units is
sheeting is easy impossible
- Cartographic quality - Simulation modeling is difficult
- Retrieval, updating and generalization
of graphics and attributes are possible
 Raster
• Adv: • Dis adv:
- Simple data structure - Large data volume
- Use of cheap technology - Low cartographic output quality
GMR Institute of Technology

- Many forms of data available - Limited attribute data handling

- Mathematical modeling is easy - No topological processing
- Location specific manipulation of - Coordinate transformation are difficult.
attribute data is easy
- Many kinds of spatial analysis and
filtering.