CHAPTER 12
Image Change Detection and Fusion
Using MRF Models
Teerasit Kasetkasem, Pramod K. Varshney
12.1
Introduction
The basic theory of Markov random fields was presented in Chap. 6. In this
chapter, we employ this modeling paradigm for two image processing tasks
applicable to remote sensing. The objectives of the two tasks are:
1. To investigate the use of Markov random field (MRF) models for image
change detection applications
2. To develop an image fusion algorithm based on MRF models.
As mentioned in Chap. 2, image change detection is a basic image analysis
tool frequently used in many remote sensing applications (such as environmental monitoring) to quantify temporal information. An MRF based image
change detection algorithm is presented in the first part of this chapter. The
image fusion algorithm is primarily intended to improve, enhance and highlight certain features of interest in remote sensing images for extracting useful
information. In this chapter, an MRF based image fusion algorithm is applied
to fuse the images from a hyperspectral sensor with a multispectral image
thereby merging a fine spectral resolution image with a fine spatial resolution image. Examples are used to illustrate the performance of the proposed
algorithms.
12.2
Image Change Detection using an MRF model
The ability to detect changes that quantify temporal effects using multitemporal imagery provides a fundamental image analysis tool in many diverse
applications (see Sect. 2.7 of Chap. 2). Due to the large amount of available
data and extensive computational requirements, there is a need to develop
efficient change detection algorithms that automatically compare two images
taken from the same area at different times to detect changes. Usually, in the
comparison process (Singh 1989; Richards and Jia 1999; Lunneta and Elvigge
1999; Rignot and van Zyle 1993), differences between two corresponding pixels belonging to the same location for an image pair are determined on the
P. K. Varshney et al., Advanced Image Processing Techniques for Remotely Sensed Hyperspectral Data
© Springer-Verlag Berlin Heidelberg 2004
Image Change Detection and Fusion
Using MRF Models
Teerasit Kasetkasem, Pramod K. Varshney
12.1
Introduction
The basic theory of Markov random fields was presented in Chap. 6. In this
chapter, we employ this modeling paradigm for two image processing tasks
applicable to remote sensing. The objectives of the two tasks are:
1. To investigate the use of Markov random field (MRF) models for image
change detection applications
2. To develop an image fusion algorithm based on MRF models.
As mentioned in Chap. 2, image change detection is a basic image analysis
tool frequently used in many remote sensing applications (such as environmental monitoring) to quantify temporal information. An MRF based image
change detection algorithm is presented in the first part of this chapter. The
image fusion algorithm is primarily intended to improve, enhance and highlight certain features of interest in remote sensing images for extracting useful
information. In this chapter, an MRF based image fusion algorithm is applied
to fuse the images from a hyperspectral sensor with a multispectral image
thereby merging a fine spectral resolution image with a fine spatial resolution image. Examples are used to illustrate the performance of the proposed
algorithms.
12.2
Image Change Detection using an MRF model
The ability to detect changes that quantify temporal effects using multitemporal imagery provides a fundamental image analysis tool in many diverse
applications (see Sect. 2.7 of Chap. 2). Due to the large amount of available
data and extensive computational requirements, there is a need to develop
efficient change detection algorithms that automatically compare two images
taken from the same area at different times to detect changes. Usually, in the
comparison process (Singh 1989; Richards and Jia 1999; Lunneta and Elvigge
1999; Rignot and van Zyle 1993), differences between two corresponding pixels belonging to the same location for an image pair are determined on the
P. K. Varshney et al., Advanced Image Processing Techniques for Remotely Sensed Hyperspectral Data
© Springer-Verlag Berlin Heidelberg 2004