Multi-Temporal Change Detection and Image Segmentation: Under Guidance Of: Dr. Anupam Agarwal Sir

Multi-Temporal Change detection And
Image Segmentation
under guidance of:

Dr. Anupam Agarwal Sir
Done by :-
Shubham Raina (IIT2008086)
Ashish Meena (IIT2008089)
Aim and Objective
Concept of Image Segmentation will be used in
Satellite Imagery that will be to Detect Multi-
Temporal Changes of particular Landscape.
These changes can be of various aspects like
• Forestry
• Watershed Areas
• Traffic
• Bare soil
Area of our Interest is Forestry
Reason :
• Remote sensing fulfills the need for
information regarding forest ecosystems and
provides input to ecological models to help
monitor forest dynamics.
• The information can be used to make some
predictions for a particular period of time.
These prediction can be :
• How much forest area remain after that time ?
• What will be the consequences of forest area
reduction over crops, climate and water
resources etc.
Tools and Techniques
• Tools
ERDAS 8.6
Beam 4.8
ERDAS 8.6 : ERDAS IMAGINE is a remote sensing application with
raster graphics editor capabilities designed by ERDAS, Inc for
geospatial applications. The latest version is 2010, version 10.1.
ERDAS IMAGINE is aimed primarily at geospatial raster data
processing and allows the user to prepare, display and enhance
digital images for mapping use in GIS (geographical information
system) or in CADD (computer-aided drafting and design )
software. It is a toolbox allowing the user to perform numerous
operations on an image and generate an answer to specific
geographical questions
Literature Survey
1. A REVIEW ON IMAGE SEGMENTATION TECHNIQUES WITH
REMOTE SENSING PERSPECTIVE (2010)–
V. Dey a , ∗, Y. Zhang a, M. Zhong b, a Department of Geodesy and Geomatics
Engineering, University of New Brunswick (UNB), Fredericton, E3B 5A3, NB,
Canada –(d1991, yunzhang)@unb.cab Department of Civil Engineering, UNB,
Fredericton, E3B 5A3, NB, Canada – ming@unb.ca
• This paper deals with details of techniques developed

and used for image segmentation with remote sensing
perspectives. Different models discussed here are
 Object-background Model
 Markov Random Field Model
 Fuzzy Model
Multi-resolution Model
 Watershed Model
• The selection of segmentation approach depends on
what quality of segmentation is required. Further, it
also depends on what scale of information is
required. Few examples, based on done literature
review in this paper have been stated to illustrated
the Idea of segmentation.
2. Object-oriented Change Detection Method Based on Multi-
scale and Multi-Feature Fusion (2009)
Wen-jie WANG , Zhong-ming ZHAO, Hai-qing ZHU Dept. of Image Processing
Institute of Remote Sensing Applications Chinese Academy Sciences,
P.O. Box 9718, Beijing 100101, China wangwenjie@gmail.com
High resolution satellite images offer abundance information of
the earth surface for remote sensing applications.
• Using change detection technology to extract the target area
changes from high resolution remote sensing images and rapidly
update map database has become a focus research of remote
sensing information processing.
• However the traditional methods of change detection are not

suitable for high resolution remote sensing images.
• To overcome the limitations of traditional pixel-level change
detection methods and the difficulties of change detection of high
resolution remote sensing images, based on object-oriented
analysis method, this paper presents a novel way of multi-scale
and multi-feature fusion for high resolution remote sensing
images change detection. Experiments show that this method has
a stronger advantage than the traditional pixel-level method of
high resolution remote sensing image change detection.
Modules
Module 1 :- Image Segmentation
Module 2 :-Change Detection and Observation
Module 1 – Image Segmentation
• segmentation refers to the process of partitioning a digital

image into multiple segments (sets of pixels, also known as
super pixels). The goal of segmentation is to simplify and/or
change the representation of an image into something that is
more meaningful and easier to analyze. Image segmentation is
typically used to locate objects and boundaries (lines, curves,
etc.) in images. More precisely, image segmentation is the
process of assigning a label to every pixel in an image such that
pixels with the same label share certain visual characteristics.
• The result of image segmentation is a set of segments that
collectively cover the entire image.
Each of the pixels in a region are similar with respect
to some characteristic or computed property, such
as color, intensity, or texture. Adjacent regions are
significantly different with respect to the same
characteristic(s).
Methods
K-means algorithm is an iterative technique that is used to
partition an image into K clusters. The basic algorithm is:
 Pick K cluster centers, either randomly or based on some
Heuristic.
 Assign each pixel in the image to the cluster that minimizes
the distance between the pixel and the cluster center.
 Re-compute the cluster centers by averaging all of
the pixels in the cluster .Repeat steps 2 and 3 until
convergence is attained (e.g. no pixels change
clusters)
• Compression-based methods –
Compression based methods postulate that the optimal
segmentation is the one that minimizes, over all possible
segmentations, the coding length of the data . The
connection between these two concepts is that
segmentation tries to find patterns in an image and any
regularity in the image can be used to compress it. The
method describes each segment by its texture and
boundary shape. Each of these components is modeled by
a probability distribution function and its coding length is
computed as follows.
Watershed transformation
The watershed transformation considers the gradient
magnitude of an image as a topographic surface. Pixels
having the highest gradient magnitude intensities
(GMIs) correspond to watershed lines, which represent
the region boundaries. Water placed on any pixel
enclosed by a common watershed line flows downhill
to a common local intensity minimum (LIM). Pixels
draining to a common minimum form a catch basin,
which represents a segment.
Watershed Model will be used for Image segmentation,
because if we are working on image that can be expressed
as two variables function then it’s derivative at a point on
image will be gradient intensity of that point . So it’s
mathematics is simple to understand.
Module 2
Change Detection
Object-based post-classification change detection
The first step in object-based change detection is to perform the image
segmentation. However, instead of using multispectral imagery, we segmented the
2004 land cover classification map.
The resultant objects from the segmentation were identical to those of a union
overlay operation between the two classified polygon layers (i.e., the 1999 and 2004
classification maps), in which all polygons from both classification layers were split at
their intersections and preserved in the resultant object level, as illustrated in Figure
2. Both classification maps for 1999 and 2004 were used as thematic layers when
performing the segmentation. When using a thematic layer, the borders separating
different thematic classes are restrictive for any further segmentation [28]. In other
words, the generated objects were not allowed to cross any of the borders of
different land cover classes. We generated the objects based exclusively on the
information of thematic layers by setting the weight of the image layer to 0 [28].
Image objects for change detection (Panel C) represent the intersections between
the two classification maps (Panel A: 1999; Panel B: 2004). Paned D shows the change
detection results obtained from the object-based approach.
Courtesy :-
Object-based Land Cover Classification and Change Analysis in
the Baltimore Metropolitan Area Using Multi-temporal High
Resolution Remote Sensing Data
Weiqi Zhou 1,*, Austin Troy 1 and Morgan Grove 2
1 Rubenstein School of Environment and Natural Resources,
University of Vermont, George D. Aiken
Center, 81 Carrigan Drive, Burlington, VT 05405, USA; E-mail:
atroy@uvm.edu (Austin Troy)
2 Northeastern Research Station, USDA Forest Service, South
Burlington, VT 05403, USA;
E-mail: jmgrove@gmail.com
References:-
1. 2009 International Conference on Environmental
Science and Information Application Technology
2. 2009 Urban Remote Sensing Joint Event
Thank You

Multi-Temporal Change Detection and Image Segmentation: Under Guidance Of: Dr. Anupam Agarwal Sir

Uploaded by

Copyright:

Available Formats

Multi-Temporal Change Detection and Image Segmentation: Under Guidance Of: Dr. Anupam Agarwal Sir

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Multi-Temporal Change Detection and Image Segmentation: Under Guidance Of: Dr. Anupam Agarwal Sir

Uploaded by

Copyright:

Available Formats

Multi-Temporal Change detection And

under guidance of:

• This paper deals with details of techniques developed

• However the traditional methods of change detection are not

• segmentation refers to the process of partitioning a digital

You might also like