4. EDGE ADAPTIVE GIHS-BT-SFIM
In many applications such as precision farming and urban classification, there’s a need to have more precise
information on edges because of the error occurrence is especially there. Using MS image data off edges and the
minimum saturation on edges presents a good quality on spatial and spectral information for image based analysis
purposes similar to classification. So, in this improved method
edges are transferred from the GIHS-BT-SFIM image to the MS image. This approach extracts the edges
from the pan-chromatic image; where there are edges the IHS-BT-SFIM product was imposed, otherwise, the MS
image was used. In this case the fused multichannel image can be formed by the new formula,
ˆ 2.
. ˆ
. 1.
. ˆ
L
P I
I k h x
P I
P I
P h x k h x
P I
P P or P
7 Another approach, which uses panchromatic image on
edges and GIHS-BT-SFIM off edges, can be introduced as equation 8,
ˆ 2.
. ˆ
1. .
ˆ
L
P I
I k h x
P I
I k h x
P I
P P or P
8 The edges can be extracted using standard edge detection
methods such as the canny detector Canny, 1986. There is another edge detector suggested by Prona and Malik 1990
whereby the considerable results can be achieved in a fusion task which was introduced as the best in this study
among other edge detectors in Rahmani, et al. 2008,
9 10
4
exp ;
10 10
| |
h x and
P
9 The best edge detector in this study was Roberts, as experienced.
5. CLASS-BASED FUSION APPROACH
Because of different distortion extent due to each class, it was interesting to test every method in a class-based
manner. This part was just tested on urban image because of more distinct nature of the classes and only two different
types of objects defined as urban and vegetation cover.
A class based fusion was tested, which gives different
coefficients for each method due to its class. It means that, two adjustable parameters, introduced in first part for
defining combination of fusion methods, has changed in a LUT form and their best values were selected according to
the fusion results assessment in each class. There were 16 miscellaneous cases which were tested and acquainted in
the following part.
6. RESULTS
Two datasets have been tested in this study. First one was a high resolution image in an urban area and the other a
medium resolution image of an agricultural field. Setting k
i
parameters of the GIHS-BT-SFIM, leads us to different fusion techniques and comparison for each method and
their spectral and spatial quality was simply implemented. Spatial quality can be judged visually, but subtle color
changes are more difficult to notice in this manner Rahmani et al, 2010. In order to compare method results,
spatial and spectral qualities by relying on both visual inspection and metric performance data were evaluated.
Band-to-band correlation coefficient CC, distortion extent, ERGAS, SAM, Q Index Q-average, RMSE and
RASE are the spectral metrics, and CC and average gradient are the spatial metrics for analyzing fusion
methods fulfillment Rahmani, 2008; Alparone, 2008; Zhang, 2008.
6.1. Urban area The first study area is an urban region located in Shahriar,
Longitude: 51 3
, latitude: 35 39
Tehran, Iran. The IKONOS image with four bands Red, Green, Blue and
Near infrared and 3 meters spatial resolution in MS and one panchromatic band with 1 meter spatial resolution was
used in this region Figure 1. Results Table 1 are
R e
fe r
e n
c e
v a
lu e
s S
p a
ti a
l 1
CC -
p 1
G r
a d
.
C C
M S
1 ER
G A
S QI
1 R
M S
E R
A S
E S
A M
S ID
IHS
ˆ 1
2 1
P I k k
.9 7
9 3
9 4
9 .9
4 5
6 1
2 8
3 1
.5 7
8 6
3 7
.2 2
2 6
2 8
9 .7
9 7
2 3
2 .9
4 2
7 8
6 5
1 4
5 .5
6 9
1 4
3 6
.1 6
4 6
8 2
.0 4
1 2
5 5
.0 3
8 3
4 4
IHS- BT
ˆ 1
2 0.5
P I k
k
.9 8
9 3
5 .9
4 9
8 6
5
3 1
.4 3
5 5
2 .2
1 9
6 1
4 9
9 .7
6 7
2 1
5 .9
2 7
8 4
4 1
4 5
.8 6
2 6
1 3
6 .2
3 7
5 8
9 .8
9 1
4 5
6 4
.0 1
6 7
2 4
BT
ˆ 1
2 P I
k k
.9 7
1 8
6 9
1 .9
4 2
9 5
8 8
3 1
.5 6
6 2
2 .2
2 1
5 8
8 7
9 .7
8 6
3 3
1 1
.8 9
5 9
3 3
3 1
4 6
.6 5
6 2
8 3
6 .4
3 4
7 6
6
2 .2
9 E
-07
.0 3
7 7
BT- SFIM
ˆ 1
2 1
L
P P k
k
.9 7
3 8
4 6
6 .9
4 4
1 8
9 9
3 1
.5 9
1 4
7 .2
2 9
3 4
9 .7
9 1
1 4
5 .9
3 1
3 2
9 1
1 4
5 .7
7 4
1 3
6 .2
1 5
5 9
9 1
.6 4
E +
.0 3
9 9
SFIM
ˆ 1 1
2 0
L
P P k
k
.8 2
1 7
7 1
9 .4
4 7
7 5
9 2
3 8
.0 3
8 3
8
.8 3
3 9
8 6
9
4 .2
8 9
2 9
4 4
.9 7
4 5
9 1
3 6
8 .9
9 9
7
1 7
.1 1
9 5
3 1
2 .2
9 E
-07
.0 1
1 1
1
SMPR 2013, 5 – 8 October 2013, Tehran, Iran
This contribution has been peer-reviewed. The peer-review was conducted on the basis of the abstract. 141
Edge + IHS
.6 8
2 5
3 .8
3 1
5 4
6 4
.2 4
4 6
4 .3
2 1
9 2
8 8
.8 4
6 7
6 9
.9 5
2 4
5 2
1 3
3 .9
7 6
4 3
3 .2
8 4
6 3
1 .6
9 2
2 5
2 .0
3 6
8 2
Edge +
IHS- BT
k= 0.5
.6 8
2 4
.8 3
4 6
9 4
.1 6
9 1
3 .3
2 4
2 3
8 .8
2 1
5 2
.9 4
1 6
2 1
3 4
.2 3
9 5
3 3
.3 4
9 9
8 .7
4 9
5 8
1 .0
1 4
1 3
Edge + BT
.6 7
6 5
8 .8
2 8
6 9
9
4 .2
7 1
1
.3 2
1 7
1 7
8 .8
3 9
6 6
5 .9
1 3
2 9
9 1
3 4
.9 9
3 6
3 3
.5 3
7 3
3 2
.3 E
-07 .0
2 5
9
Edge + BT-
SFIM .6
7 7
4 1
.8 3
3 8
4 .2
6 9
4 6
.3 2
4 3
3 8
.8 4
1 8
2 2
.9 4
1 8
2 4
1 3
4 .1
8 4
6 3
3 .3
3 6
3 4
1 .3
4 E
+ .0
2 6
4 7
Edge +
SFIM
.7 8
2 9
4 9
.4 2
2 6
2 3
7 .3
4 3
8 7
.8 4
2 6
8 9
4 .1
4 2
5 4
1 .9
7 6
4 7
9 6
6 .6
9 4
3 7
1 6
.5 6
9 3
1
2 .2
9 E
-07
.0 1
1 2
Edge: HIS
+MS
.3 2
3 9
1 .2
1 4
5 8
1 1
9 .9
2 1
3 8
.9 3
6 4
9 7
2 .3
2 6
8 3
.9 9
6 8
8
.0 3
1 7
1 9
.2 6
2 8
7 .0
8 6
6 1
4 .0
1 1
2 8
Edge: HIS_
BT +MS
.3 2
2 5
6 .2
1 4
5 7
1 9
.9 2
1 7
.9 3
6 9
3 2
.3 2
2 2
7 2
.9 9
6 1
4 2
.0 3
1 1
2 1
9 .2
7 7
7 7
4 .0
3 9
3 6
7 .0
1 2
3
Edge: BT
+MS
.3 1
8 5
1 2
.2 1
4 1
4 3
1 9
.9 3
4 2
.9 3
6 9
3 2
.3 3
4 6
2 .9
9 4
4 9
7 .0
3 1
3 5
7 9
.3 4
8 2
8 5
2 .3
5 E
-07 1
.6 6
E -05
Edge: BT-
SFIM +MS
.3 1
9 5
9 8
.2 1
4 3
1 1
1 9
.9 2
9 4
.9 3
6 6
3 7
2 .3
2 5
1 9
2 .9
9 5
2 6
2 .0
3 1
1 8
2 9
.2 9
6 9
9 .0
5 5
4 1
.0 1
8 2
Edge: SFIM
+MS
.3 5
4 7
6 4
.1 8
9 7
7 9
1 8
.9 7
7 8
.9 6
4 6
1 4
1 .7
5 2
5 2
1
.9 9
6 3
8 8
.0 2
3 5
6 3
7 .0
2 4
6 7
8
2 .3
5 E
-07
.0 1
4
Table 1: Assessing the fusion methods in an urban region
It is clear that edge adaptive method has improved spectral information in all procedures. Fusion results are provided
in figure 1. pan
MS
IHS IHS+Edge
HIS-BT HIS-BT+Edge
BT BT+Edge
BT-SFIM BT-SFIM+Edge
SFIM SFIM+Edge
Figure 1:Results for Urban area
6.2. Agricultural fields The area of interest deals with agricultural fields located
Kategori