GS122 Remote Sensing Principles

Lecture

Introduction to image analysis

Dr. Deus D.
Introduction to image analysis
 Image analysis
• Image analysis is the extraction of meaningful
information from images; mainly from digital images by
means of digital image processing techniques
• Computers are indispensable for the analysis of large
amounts of data, for tasks that require complex
computation, or for the extraction of quantitative
information.
• On the other hand, the human visual cortex is an
excellent image analysis apparatus, especially for
extracting higher-level information, and for many
applications including medicine, security, and remote
sensing
• Human analysts still cannot be replaced by computers.
• For this reason, many important image analysis tools
such as edge detectors and neural networks are inspired
by human visual perception models.
 Computer image analysis
• Computer image analysis largely contains
the fields of computer or machine vision,
and medical imaging, and makes heavy
use of pattern recognition, digital
geometry, and signal processing.
 Digital image analysis
The applications of digital image analysis are continuously
expanding through all areas of science and industry,
including:
• medicine
• microscopy
• remote sensing
• astronomy
• defense
• resources science
• manufacturing
• Security
• Military
• robotics or computer vision application
 Digital image analysis Applications
cont.
• Military application e.g. identification of
approaching aircraft and the detection of target
for cruised missiles.
• Robotic or computer vision application:
– This involves use of mathematical description
of objects ‘seen’ by a television camera
representing the robotic eye and the
comparison of these mathematical
descriptions with the patterns describing
objects in the real world
 Object-Based Image Analysis
(OBIA)
• Object-Based Image Analysis is a sub-
discipline of geoinformation science
devoted to partitioning remote sensing
(RS) imagery into meaningful image-
objects, and assessing their
characteristics through spatial, spectral
and temporal scale.
 Digital image processing
• Digital image processing is the use of
computer algorithms to perform image
processing on digital images.
• As a subfield of digital signal processing, digital
image processing has many advantages over
analog image processing;
- It allows a much wider range of algorithms to be
applied to the input data
- Can avoid problems such as the build-up of
noise and signal distortion during processing.
 Digital image processing
• Digital image processing allows the use of much more
complex algorithms for image processing, and hence
can offer both more sophisticated performance at simple
tasks, and the implementation of methods which would
be impossible by analog means.
• In particular, digital image processing is the only
practical technology for:
- Classification
- Feature extraction
- Pattern recognition
- Projection
- Multi-scale signal analysis
Signal processing is the analysis, interpretation, and
manipulation of signals
 Digital image processing
Some techniques which are used in digital
image processing include:
• Principal components analysis
• Independent component analysis
• Self-organizing maps
• Hidden Markov models
• Neural networks
 Digital image classification
• This is a process in which the (human)
operator instructs the computer to perform
an interpretation according to certain
conditions
• The conditions are defined by the operator
• Image classification is based on the
different spectral characteristics of
different materials on the earth’s surface.
Satellite Image Classification
Multi-spectral Classification

What is it?
• grouping of similar features

• separation of dissimilar ones

• assigning class label to pixels

• resulting in manageable size of classes

 Multi-spectral Classification
• Image classification is the process of assigning
pixels to nominal, i.e. thermatic, classes.
• Classification may be considered as some form
of pattern recognition, that is, the identification of
the pattern associated with each pixel position in
an image in terms of the characteristics of the
objects or materials that are present at the
corresponding point on the earth’s surface.
• Input is a multi-band image, output is a raster in
which each cell has a (thermatic) code
 Principle of image classification
• Pixel is assigned to a class based on its feature vector,
by comparing it to predefined clusters in the feature
space.
• Doing this for all image pixels results in a classified
image.
• The crux of image classification in comparing it to
predefined clusters, which require definition of clusters
and methods of comparison
• Definition of clusters is an interactive process and is
carried out during the training process.
• Comparison of individual pixels with the clusters take
place using classifier algorithms
 Feature vector and feature space
• The values of one pixel in two bands can
be regarded as a components of a 2D
vector, the feature vector.
• Feature space is a graphy that shows the
values of a feature vectors
• Feature space is also called feature space
plot or scatter plot
• Distances in feature space is expressed
as Euclidian distance and the units are DN
 Steps in Satellite Image
Classification process
The process of image classification involves five steps:
1. Selection and preparation of image data. Depending
on the cover types to classified, the most appropriate
sensor, the most appropriate dates of acquisition and
the most appropriate bands sholud be selected.
2. Definition of the clusters in the feature space. Here two
approaches are possible: supervised classification and
unsupervised classification
- In a supervised classification the operator defines the
clusters during training process
- In an unsupervised classification a clustering algorithm
automatically finds and defines a number of clusters in
the feature space
 Steps in Satellite Image
Classification process
3. Selection of a classification algorithm.
Once the spectral classes have been defined in the feature
space, the operator needs to decide on how the pixels
(based on their DN-values) are assigned to the classes.
The assignment can be based on different classification
algorithms.
4. Running the actual classification.
Once the training data have been established and the
classifier algorithm selected, the actuial classification can
be carried out.
This means that , based on the DN values each individual
pixel in the image is assigned to one predefined classes
 Steps in Satellite Image
Classification process
5. Validation of the results.
Once the classified image has been
produced its quality is assessed by
comparing it to reference data (ground
truth).
This requires selection of a sampling
technique, generation of an error matrix,
and the calculation of error parameters.
Classification Methods

• MANUAL
– visual interpretation
– combination of spectral and spatial
information

• COMPUTER ASSISTED
– mainly spectral information

• OBJECT ORIENTED
– combining the use of GIS functionalities
DIGITAL IMAGE INTERPRETATION

IMAGE SPACE

– array of elements
corresponding to reflected
or emitted energy from
IFOV

– spatial arrangement of
the measurements of the
reflected or emitted
energy
 COMPUTER ASSISTED
CLASSIFICATION
UNSUPERVISED APPROACH
• based on spectral groupings
• considering certain similarity measures
• no prior knowledge required
• minimum user interaction

SUPERVISED APPROACH
• based on spectral groupings
• requires prior knowledge about the area
• maximum user interaction
UNSUPERVISED APPROACH

CLUSTERING PARAMETERS

• number of clusters
• size of cluster
• distance between the clusters
• cluster elimination value
 Unsupervised classification

255
class4 Class 2

Band 2

Class 1
class3

0
0 Band 1 255

Feature Space
Satellite Image (SPOT)
N.O. Polder, the Netherlands
Unsupervised Classification
Result