Self-adaptation in Image and Video Retrieval - Multimedia Database Retrieval: Technology and Applications - page 81

only the region selected which adequately reflects those properties of the object such
as color or shape, which are usually used as features for matching in retrieval.

3.4.1 Segmentation of the Region of Interest

Image segmentation is considered as a crucial step in performing high-level
computer vision tasks such as object recognition and scene interpretation [ 95 ]. Since
natural scenes within an image could be too complex to be characterized by a single
image attribute, it is more appropriate to consider a segmentation method that is
able to address the representation and integration of different attributes such as
color, texture, and shape. The Edge Flow model demonstrated in [ 94 ] is adopted,
which has proven to be effective in image boundary detection and in application to
video coding [ 96 ]. The Edge Flow model implements a predictive coding scheme to
identify the direction of change in color, texture, and filtered phase discontinuities.

3.4.2 Edge Flow Method

Let E ( s , ʸ ) be the edge energy at pixel s along the orientation ʸ . An edge flow vector

at pixel location s is a vector sum of edge energies given by:

∑

F = E ( s , ʸ ) exp ( j ʸ ) (3.23)

ʘ ( s ) ≤ ʸ ≤ ʘ ( s )+ ˀ

which is taken along a continuous range of flow directions that maximizes the sum
of probabilities:

P s

, ʸ

∑
ʸ ≤ ʸ ≤ ʸ + ˀ

ʘ (

)=

s arg max

ʸ

(3.24)

where P represents the probability of finding the image boundary if the

corresponding Edge Flow flows in the direction

( s , ʸ )

. The model in Eq. ( 3.23 )
facilitates the integration of multiple attributes in each Edge Flow which is obtained
from different types of image attributes. Consider,

ʸ

, ʸ )= a ∈ A E a ( s , ʸ ) w ( a ) E ( s (3.25)