UNIT 2

1.Spatial Data Models:

Spatial data models are methods used to represent, organize, and manage spatial
(geographic) data in a way that reflects the real-world locations, shapes, and relationships of
features on the Earth's surface.
Key Characteristics of Vector Models:

• Precision: Vector data provides exact coordinates and precise geometries, which makes it
suitable for accurate mapping and analysis.

• Data Representation: Features are represented by coordinate pairs (x, y) or triplets (x, y, z for
3D models), and can be used to calculate attributes like area, length, or distance.

• Topology: Vector models can explicitly store relationships between features, such as
adjacency, connectivity, and containment, which is useful for spatial analysis.

Applications: Vector models are ideal for applications that involve well-defined objects, like mapping
administrative boundaries, roads, or utility networks.

Key Characteristics of Raster Models:

• Simplicity: Raster models are simple to implement and are often used in remote sensing and
environmental modeling, where data can be uniformly distributed across a region.

• Continuity: They are suitable for representing continuous data, such as temperature,
elevation, or rainfall, where each location can be considered as part of a smooth surface or
gradient.

• Storage: The amount of data can grow quickly with higher resolution, as every pixel needs to
be stored.

Applications: Raster models are commonly used for representing imagery (like satellite images),
elevation models, land cover classification, or environmental simulations.

Hybrid Models

Hybrid models combine vector and raster data models to take advantage of the strengths of both.
For example, in some GIS systems, vector data might be used for the representation of urban
infrastructure (buildings, roads), while raster data might be used for background data like land cover,
climate, or elevation.

Key Considerations in Spatial Data Models:

1. Data Resolution: The level of detail (e.g., the size of the raster cell or the number of vector
points) will impact the accuracy of the model but also the computational cost and storage
requirements.

2. Scale: The scale of the spatial data model determines how it can be applied to real-world
problems (e.g., a city-level map vs. a global map).

3. Topology: Vector models, in particular, rely on the concept of topology to define spatial
relationships like adjacency or connectivity, which is crucial for tasks like network analysis.
The raster data structure refers to the actual organization and physical storage of raster data. It
describes how the raster data is implemented in a computer system and how the data is stored in
memory, on disk, or within a specific file format.

The vector data structure refers to the actual physical storage format and implementation details of
the vector data model within a computer system. It describes how the data is organized in memory
or on disk and how it is accessed, manipulated, and processed.
2.Raster Data Compression:
Raster data compression refers to the process of reducing the file size of raster datasets while
attempting to preserve the integrity of the data. Raster data, especially at high resolutions or with
large datasets (e.g., satellite imagery, aerial photos, digital elevation models), can consume
significant storage space. Compression techniques are employed to make storage and transmission
more efficient without losing important information.

Why Compress Raster Data?

1. Storage Efficiency: Reduces the amount of disk space needed for storing large raster
datasets, which is especially important when dealing with high-resolution or multi-band
images.

2. Faster Data Transfer: Smaller files are quicker to transfer, which is beneficial when sharing
raster data over networks or the internet.

3. Optimized Performance: Compression can improve the performance of GIS software and
reduce memory usage when working with large datasets.

4. Cost-Effective: Reduced storage requirements lead to lower storage costs, especially for
cloud-based storage or large-scale geospatial data repositories.

How RLE Works

• Step 1: Identify consecutive occurrences of the same value.

• Step 2: Replace each run of identical values with a pair of:

o The value (the repeated data point).

o The count (the number of consecutive times the value repeats).

How Block Coding Works

1. Data Grouping: The input data is divided into blocks. These blocks may have a fixed size (e.g.,
8 bits, 16 bits) or variable size based on the content of the data or the application.

2. Block Processing: Each block is processed independently. Common operations in block

coding include:

o Compression: Data within the block is compressed using various techniques (e.g.,
Huffman coding, Run-Length Encoding, or Dictionary-based methods like LZ77).

o Error Correction: In communication systems, block coding might involve adding

redundancy to each block to detect and correct errors (e.g., in Hamming codes or
Reed-Solomon codes).

3. Reassembly: After all blocks are processed, the compressed or encoded blocks are either
stored or transmitted. When needed, they can be decoded or decompressed and
reassembled to reconstruct the original data.

Steps Involved in Chain Coding:

1. Start at a Boundary Pixel: The process begins by selecting a starting pixel along the object's
boundary. Typically, this could be the leftmost or topmost boundary pixel.
 2. Move to Adjacent Boundary Pixels: From the starting pixel, move to the next adjacent
boundary pixel, which is connected to the current pixel. For each move, a direction is
recorded.

3. Record the Direction: The direction of movement is stored using a predefined coding
scheme (e.g., 4 directions or 8 directions).

4. Continue Until Looping Back: The process continues until the boundary is traced completely
and the starting point is reached, completing a closed loop.

Steps in Quadtree Encoding:

1. Start with the Entire Image: Consider the entire image as the first region (root of the
quadtree).

2. Check Homogeneity:

o If the entire region is uniform (all pixels have the same value), store that value as a
single node in the quadtree.

o If the region is non-uniform (pixels have different values), subdivide it into four equal
quadrants and recursively apply the same process to each quadrant.

3. Recursion: For each non-uniform quadrant, check if it is homogeneous or needs further

subdivision. Continue until every region is homogeneous, and the recursion stops.

4. Construct the Quadtree: The result is a tree-like structure, where each node represents
either a homogeneous region or a subdivision of smaller regions.

3.TIN and GRID Models:

TIN (Triangulated Irregular Network) Data Model

A TIN (Triangulated Irregular Network) is a type of vector-based data model used to represent
surface data or terrain models. It is a way to describe the shape of a 3D surface by using a network
of non-overlapping triangles. TINs are particularly useful for representing elevation data and other
continuous surface features.

How TIN Works:

• Vertices (Points): The basic elements of a TIN are the vertices, which represent specific
points on the surface. Each vertex has x, y, z coordinates (e.g., latitude, longitude, and
elevation).

• Edges (Lines): These are the straight lines connecting the vertices. Each edge is shared
between two adjacent triangles in the network.

• Triangles: A TIN is composed of triangles, where each triangle is formed by three vertices.
The triangles are created in such a way that the surface between them is non-overlapping
and irregular.

Advantages:

1. Accurate for Irregular Surfaces: TIN can better represent uneven surfaces (e.g., slopes and
valleys) than grid-based models.
 2. Efficient for Large Areas: TIN uses fewer data points, making it storage-efficient for complex
terrains.

3. Better for Steep Slopes: TIN handles steep terrain features (e.g., cliffs) more effectively than
raster models.

4. Compact Storage: TIN only stores vertices and edges, making it more storage-efficient for
less complex areas.

5. Flexible: TIN can model both smooth and rugged terrain, suitable for 3D surface modeling.

Disadvantages:

1. Complex Structure: TIN is more complex to manage and process than raster data.

2. Not for Regular Data: TIN isn’t ideal for evenly distributed data like temperature, which is
better suited for grids.

3. Hard to Update: Adding or modifying data points in a TIN can be computationally intensive.

4. Boundary Issues: Triangles at the edges may not perfectly represent the surface, leading to
artifacts.

5. Limited Compatibility: Some analyses may require converting TIN to other models (e.g.,
raster) for compatibility.

Grid Data Model

The Grid Data Model is a spatial data model commonly used in geographic information systems
(GIS) to represent continuous data over a defined area. It is a raster-based model where the space is
divided into a regular, uniform grid of cells or pixels, each with a specific value that represents a
spatial phenomenon. Grid data is commonly used for surface representation and spatial analysis,
particularly for continuous variables like elevation, temperature, rainfall, or land use.

Structure of Grid Data Model:

1. Cells (Pixels): The basic unit of a grid is a cell (also known as a pixel). Each cell has a value
that represents the phenomenon being studied, such as the temperature or elevation at that
point.

o Each cell has specific coordinates (usually x and y), representing its position in the
grid, and a value (e.g., a numeric value, such as temperature or elevation).

2. Rows and Columns: The grid is organized into rows and columns, where each row consists of
a series of cells, and each column represents a vertical division. The number of rows and
columns defines the resolution of the grid (i.e., the level of detail).

3. Resolution: The resolution of the grid is determined by the size of each cell (the pixel size),
which can affect the precision of spatial data representation. Smaller cells provide more
detailed data (higher resolution), while larger cells represent broader areas (lower
resolution).

4. Extent: The extent refers to the area covered by the grid. It can represent a small study area
or a large geographical region, depending on the purpose of the analysis.
Types of Grid Data:

1. Integer Grid: The grid cells store integer values, such as counts or categorical data (e.g., land
cover classes, land use types).

2. Floating-Point Grid: The grid cells store floating-point values that represent continuous data
(e.g., elevation, temperature, or pollution levels). These values allow for more detailed
representation of real-world phenomena.

Advantages of the Grid Data Model

1. Simple Structure: Easy to understand and implement with values stored in regular cells.

2. Ideal for Continuous Data: Great for representing phenomena like elevation, temperature,
and rainfall.

3. Supports Spatial Analysis: Works well with techniques like map algebra, overlay, and buffer
analysis.

4. Compatible with Remote Sensing: Fits naturally with satellite and aerial imagery.

5. Scalable: Can represent small or large areas with varying resolution and detail.

Disadvantages of the Grid Data Model

1. Resolution Limitations: Low resolution can miss details; high resolution increases file size
and processing demands.

2. Loss of Precision: Not suitable for representing sharp or intricate features like roads and
boundaries.

3. Data Redundancy: Uniform areas can lead to wasted storage due to repetitive values.

4. Storage Requirements: High-resolution grids require significant storage and computational

power.

5. Not Ideal for Discrete Features: Better suited for continuous data, not for discrete features
like buildings or roads.