KING SAUD UNIVERSITY

COLLEGE OF ENGINEERING

RESEARCH CENTER

Final Research Report No. 6/427

"DIGITAL IMAGE PROCESSING

IN REMOTE SENSING"

By
Prof. Ismat Mohammed El Hassan

Month 1. 1428 H
Month 1. 2007 G


Digital Image Processing in Remote Sensing

. .
1428 2007

13

.1 -

14

. 2 -

18

1.2

18

2.2

18

3.2

19

1.3.2

20

2.3.2

21

4.2

22

5.2

24

6.2

24

7.2

27

8.2

28

9.2

29

10.2

29

.3

31

1.3

31

2.3

32

3.3

33
33

4.3

5.3
1.5.3

34

34

2.5.3

35

3.5.3

36

6.3

36

7.3

39

. 4 - ) (

40

1.4

40

2.4

41

1.2.4

41

2.2.4

42

3.4

43

1.3.4

43

2.3.4

43

3.3.4

44

44

4.4

1.4.4

44

2.4.4

46

3.4.4

47

4.4.4

48

5.4

49

1.5.4

50

2.5.4

51

1.2.5.4

52

2.2.5.4

54

 3.2.5.4

55

3.5.4

56

6.4

58

.5 -

60

1.5

60

2.5

62

3.5

66

1.3.5

66

2.3.5

69

3.3.5

71

4.5

74

1.4.5 --

77

2.4.5

78

3.4.5

80

5.5

) (

82

6.5

85

.6

86

1.6

86

2.6

87

3.6

88

4.6

92

5.6

96

.7

97

1.7
2.7

97
97

 1.2.7

98

2.2.7

99

3.2.7

100

4.2.7

101

105

3.7

4.7

105

5.7

117

8. -

118

1.8

118

2.8

119

3.8

121

1.3.8

121

2.3.8

124

3.3.8

127

4.8

130

5.8

132

6.8

135

136

.





.

.


.



.

.

. .



/
.
/
.

. .
1428 2007


1.2

19

2.2

20

3.2

22

4.2

23

5.2 - QB 2.4

25

6.2 QB 0.6

26

7.2 - QB 0.6

26

8.2 SPOT 10

26

9.2 TM 30

27

10.2 MSS 79

27

11.2 8 256

28

12.2 7 128

28

TM

29

13.2 2
1.3

34

2.3

35

3.3

36

4.3 180

37

5.3 -

37
38

6.3
1.4

42

2.4 4

45

3.4 4 7

46

4.4. A

47

5.4

49

 a.6.4

51

b.6.4

51

a.7.4

54

b.7.4

54

8.4

56

9.4

57

10.4

58

1.5 1.5

61

2.5

63

3.5

63

4.5

64

5.5

64

6.5

65

7.5

69

8.5

72

9.5 7

75

10.5 9.5

76

11.5 TM

76

12.5 11.5

76

13.5 RGB

78

14.5

80

15.5 4x4

83

16.5 15.5

83

17.5 15.5

84

1.6

87

10

 2.6

87

3.6 5.5

89

4.6

89

5.6 -

90

6.6

90

7.6 5x3

90

8.6 7.6

91

9.6

91

10.6

91

11.6

92

12.6

92

13.6

93

14.6

93

15.6

93

16.6

94

17.6

94

18.6

95

19.6

95

1.7

100

2.7 1 2

102

3.7 5

103

4.7 7

104

5.7 5 7

104

6.7 1 2

107
109

7.7 1

11

110

8.7 2
9.7 4 TM

115

10.7

116

11.7

116

12.7

116

1.8

119

2.8

120

3.8

122

4.8 B1

123

5.8 B2

123

6.8

123

7.8

125

8.8 B1

126

9.8 B2

126

10.8

126

11.8

129

12.8

131

12

1.2

23

2.2

29

1.3 0 11

31

2.3 98 101 254 255

31

1.5 38000

60

2.5

67

3.5

70

4.5

71

5.5

73

13

Introduction

.
.


]. [1]-[5

.
.
" "
.
.

. ) (
.
.

. )(
) (photographic cameras

) (electronic sensors

14

 (digital
) images . :
) (active
) (passive

.
.

.
]. [5],[6


15 30
128 .
.

.
.


.
:
.

15

 - ) : (image restoration

.
.
.
- ) : (image enhancement

.
.

.
- ) (image classification :


.
- ) : (data merging


.

)
( .
.

.

16

 .

.
.

17

Digital Images
1.2

.
.

2.2 Data Acquisition

.
(incident
) . radiation
) (absorbed radiation
) (emitted radiation
) (transmitted radiation
) (reflected radiation
) (digital numbers, DN
) (scattered radiation 1.2
]. [5]-[8

.

) (sensor .
18

 ) (amplification
.
.
) (digital numbers, DN
.

.

) :(1.2

3.2

:

. )

( .

19

 .
.

.

1.3.2


) (
. . 2.2

) :(2.2

) . (Landsat ) ( (Multi Spectral Scanner), MSS

) (Thematic Mapper, TM
.
20

2.3.2 Push Broom Scanner

) (push broom
(charge coupled devices,
) . CCD

.

.
) . ( 3.2
) (Linear Array Multi Spectral Scanner

.
.
) (SPOT .

21

:3.2

4.2 Structure of digital image

) ( c ) ) ( r (4.2
) . (pixel

(ground

) pixels . 4.2 r
c .
)(
) (1,1 4.2 x
) (3,2 .

22

 )(c
1,1

)(r

:4.2

.
185 x 170 km 42
291
) . ( 1.2
:1.2

MSS
TM

2983
5965

3548
6967

10.58
41.56

4
7

)(
42.33
290.91

) (digital number, DN
.

23

 ) (r,c ) (DNs
) . (b :
)(1.2

)DN = E (b, r, c

E r c . b

]. [8]-[10

5.2 )Digital Numbers (DN

.
0


255 .

6.2

Spatial Resolution

) ( pixel

. ) 79 m
(79x79m MSS
) 30 m ( 30x30m TM
) SPOT

24

 10x10m 20x20m ( .
2x2m IKONOS
) ( ) Quick Bird 5.2 - 10.2
Digital Globe Image Library
] [11 ] [12
]. ( [13
.
) (instantaneous field of view, IFOV
. IFOV

.

) (scale

:5.2 Quick Bird

2.4 ][11

25

 :6.2 Quick Bird ][11

0.6

:7.2 Quick Bird

0.6 ][11

:8.2 SPOT
10 ][12

26

 :9.2 ) TM 30(
.[13] 1990

:10.2 ) MSS 79 (
- .[13] 1986

27

 7.2 Spectral Resolution

)
(
) (coarse spectral resolution

)( 0.51 - 0.73 m
0.22m )(fine
T M 3
0.63 m 0.69m . 0.06m

8.2 Radiometric Resolution

) (number of bits .
(8- bit data) 8 0 255
11.2 (7- bit data) 7 0
) 127 . ( 12.2 .

255

254

127

126

:11.2 8 256

:12.2 7 128

28

 2.2 6 7 8 .

:2.2

) (bit scale )(DN

0 - 63
)6 (26
0 - 127
)7 (27
0 - 255
)8 (28

64
128
256

9.2 Temporal Resolution

.

16 3 .
.



.


.

10.2

1.10.2 .
2.10.2 .

29

3.10.2

4.10.2 3x3
.
.
3

30

Preparation of Digital Image Data

1.3
) (binary system
) (bit 0 . 1
0 9 10 99 100 999 .
0 1 10 11 100 111 1000 1111
1.3 0 11 2.3
98 101 254 . [14]-[16] 255

:1.3 0 11

0
0

1
1

2
10

3
11

4
100

5
101

6
110

7
111

11
10
9
8
1011 1010 1001 1000

:2.3 98 101 254 255

98
1100010

99
1100011

101
1100101

100
1100100

254
11111110

255
11111111

2.3 255 8
8 ) . (8 bits ]. [14
) (byte 1 8 .

31

 256

256 1024 1024 .

2.3 Data storage media

. ) (tapes
) . (diskettes .

.

.
]. [3],[14],[16
(9- track tape) 9- 10 . 120 150 ) 1 =
106 ( .
) (cartridge 4x6 150 . .
4 2.0x1.75 2000 .
8 2.5x4.0 5000 10000 4 .
) (optical disc 644 . .

32

 3.3 Tape and diskette contents

) (header file
.
:
-

) . (bits

4.2

) (data file
:
= =
)(1.3

VDF

[ ( c * r * b ) * n ] * 1.2

= VDF

= r :
= c
= n
= b ) (
1.2
1 . 1.5

33

 1.3
3 16
500 . 500
:
= 2.0 16
): (1.3
=

1800000 = [ ( 500 x 500 x 2 ) x 3 ] x 1.2 = 1.8

5.3

].[14],[16

1.5.3

)Band Interleaved by Line (BIL

1 2 n
.
1 2
3 1 2
3 . 1.3
1,7,1
1,7,2
1,7,3
2,7,1
2,7,2
2,7,3

1,6,1
1,6,2
1,6,3
2,6,1
2,6,2
2,6,3

1,5,1
1,5,2
1,5,3
2,5,1
2,5,2
2,5,3

34

1,4,1
1,4,2
1,4,3
2,4,1
2,4,2
2,4,3

1,3,1
1,3,2
1,3,3
2,3,1
2,3,2
2,3,3

1,2,1
1,2,2
1,2,3
2,2,1
2,2,2
2,2,3

1,1,1
1,1,2
1,1,3
2,1,1
2,1,2
2,1,3

 :1.3

2.5.3 )Band Sequential (BSQ

.
TM . :
-

2.3
.

1,7,1
2,7,1

1,6,1
2,6,1

1
1,3,1 1,4,1 1,5,1
2,3,1 2,4,1 2,5,1

1,2,1
2,2,1

1,7,2
2,7,2

1,6,2
2,6,2

2
1,3,2 1,4,2 1,5,2
2,3,2 2,4,2 2,5,2

1,2,2
2,2,2

1,1,1
2,1,1

1,1,2
2,1,2

:2.3

35

 3.5.3 )Band Interleaved by Pixel (BIP

3.3
.

1,5,2
2,5,2

1,5,1
2,5,1

1,4,2
2,4,2

1,4,1
2,4,1

1,3,2
2,3,2

1,3,1
2,3,1

1,2,2
2,2,2

1,2,1
2,2,1

1,1,2
2,1,2

1,1,1
2,1,1

:3.3

6.3

) (pixel DN
0 255

.
" " .
.

) ( 255 - 0
0 63 63 127 127 191
191 . 255
,
) (

36

 4.3
) . (histogram 180 80
63 127 50 0 63 40
127 191 10 191 . 255

63-127 127-191 191-255

0-63

80
70
60
50

40

30
20
10
0

:4.3 180

:2.3
5.3 .

7
3
5
6
6

2
2
4
3
1

3
1
6
3
2

5
6
0
4
5

37

0
4
2
1
0

 :5.3 . 2.3

6.3 :

0
3

1
3

2
4

3
4

4
3

6
4

5
3

7
1

4
3

2

1
6

:6.3

= = m ) ( DN )(n
DN / n

)(2.3

81 / 25 = 3.24

= m
=

) ( :
)(3.3

{ [DNi m]2 / (n-1) }1/2

2.13

( 108.56/24 )1/2

38

=
=

 7.3
1.7.3 101 . 111
2.7.3 4
16 250000 .
. 1.3

3.7.3

. 5x5

.
.
3

4.7.3 . 5x5 .

3
2
7
3
7

4
3
7
2
7

7
3
5
6
4

6
4
7
2
6

39

5
4
2
4
1

)(
Digital Image Restoration

1.4

)(
.


. ) (radiometric distortion
) (geometric distortion ) (systematic and random noise
.

.
) (data restoration ) (preprocessing

.
) (Solid State Camera, SSC
MSS

.
)( .

40

 2.4 Geometric Correction

. :
-

(change in platform

) height, orientation and speed

-

(earth curvature)

) (atmospheric refraction

) (earth rotation

) (relief variation

). (instrumental errors

.
]. [9],[14],[16]-[22

1.2.4

.
.
.
.
) (skew distortion
) (de skewing

. 1.4

41

C2

D2 D

:1.4

AB CD
CD
C2D2 CD OQ . OP

2.2.4

.
) (ground control .
.
) (1.4
) (2.4 ) (x , y
) (xo , yo .
)(1.4

xo = a1 + a2 x + a3 y + a4 x y + a5 x2

)(2.4

yo = b1 + b2 x + b3 y + b4 x y + b5 x2

42

 3.4

Resampling

.
:

1.3 .4 Nearest Neighbour

. ):(3.4
)I(x,y) = I (u,v

)(3.4

) (x,y ) (u,v
" "I .
.
) ( .

2.3.4 Bilinear Interpolation

)(4.4

)I(x,y) = a1I(u,v+1) + a2I(u,v) +a3I(u+1,v+1) + a4I(u+1,v

43

 3.3.4

Cubic Convolution

. . :
)I(x,y) = am,n I(u+m, v+n

)(5.4

- 1 < m,n < 2

4.4

Radiometric Correction

.
.

1.4.4 Atmospheric Correction

) (scattered radiation
.
:
)(6.4

RT = ( * I * T ) / + RP

= RT
=
= I
= T
= Rp ) (

44

 .

.
.
4 ) (0.5 - 0.6 um MSS
7 )(0.8 - 1.1 um
.
0 1

. 7 .
:
7
) 4

( 2.4 .

. 4
. 4
5 6 7 .

:2.4 4

45

 :
7
7 ) (4
. 3.4

255

255

0

7

:3.4 4 7

7 0
4
4
4 .

2.4.4

46

 .


.
. 90

) (DNo ) (DNn
90 :
DNn = DNo / cos

)(7.4

) ( 4.4
.

:4.4 A

3.4.4

.
) = 146.4 (
) 151.2 ( .
.
47


]. [3

1.4
50, 100,
. 200

:

. ) ( " " .

= . (151.2/146.4)2 :

50

100

200

50x(151.2/146.4) = 53
100x(151.2/146.4) = 107
2

200x(151.2/146.4) = 213

4.4.4

.
.

48

 5.4 A,B,C .

. A B
. C
0
.

:5.4

5.4 Noise Removal

49

 .
.
:
. ]. [9],[14

1.5.4 Destripping

MSS
. .
.

0 . 255

.
DNi,j ) DN 0 ( 255

) i (i j DNi,j-1
) i (i j-1 :
)(8.4

DNi,j = DNi,j-1

:
)(9.4

DNi,j = DNi,j+1

a.6.4 b.6.4 .

50

31

16 19 24 27 29

30

17 18 22 25 28

15 17 20 24 26 33
16 19 23 26 27 32

:a.6.4

16 19 24 27 29 31
17 18 22 25 28 30
17 18 22 25 28 30
15 17 20 24 26 33
16 19 23 26 27 32

:b.6.4

2.5.4

MSS
TM .
.
.
)( .



) (threshold value

51



.
.

1.2.5.4

) (i )
(
xi
x i
) ( :
)(10.4

= { [ ni ( xi2 + i2 ) / ni ] - x2 }0.5

ni . i
.

2.4

.

. )
( .

52

:
:

xi

xi2

i2

3.0

9.00

0.894

0.800

4.0

16.00

2.098

4.400

4.6

21.16

2.154

4.640

4.6

21.16

1.720

2.960

3.2

10.24

1.600

2.560

3.6

12.96

2.653

7.040

23.0

90.52

22.40

= 6
x = 23 / 6 = 3.833
) (10.4 :
= { 5 (22.40 + 90.52) / 30 - (3.833)2 }0.5
2.05

i .

.
53

 2.2.5.4

.
.
:
DN i,j = ( DN i,j-1 + DN i,j+1 ) / 2

)(11.4

.
. a.7.4 b.7.4
.
.
20.00 , 21.75 , 44.25 , 20.75 :
26.69 44.25
.

20
22
43
20

19
24
47
22

18
18
42
20

23
23
45
21

:a.7.4

20
22
21
20

19
24
23
22

18
18
19
20

23
23
22
21

:b.7.4

54

 .
.

.



.

3.2.5.4

Linear method

.

)( ) (histograms

.
)(
EROS Data Center
:

)DNO(i,c

)(12.4

) DNO (i,c) = DNI (i,c) ( / i) + x - xi ( / i

c r

)= DNI(Ii,c

c r

55

 .
.

) .( 8.4
27
28
19
17

21
37
43
42

16
10
14
17

43
32
31
30

:8.4
26.50 , 26.75 :
26.75 , 26.75, = . 26.69
10.40 , 11.23 , 10.25 , 10.16 :
. 10.52
.

3.5.4

Random Noise

.
.
.
.

. .

56

 9.4

40 60 : 90 0

40
90
40

50
60
50

50
40
60

60
0
60

40
40
40

:9.4

) . (moving average filter

) 3x3 : . ( 5x5

3x3 .
.

720/15 = 48

) = 720 = . ( 15
=

32 = (2/3)*48

3x3 = 43 = 390/9
0 32
3x3 0 43
32 0 . 43
49
9.4 40 9 .
) ( 53
90 ) (37
90 . 53

57

 . 10.4
.


.
50
60
50

40
53
40

60
43
60

50
40
60

40
40
40

:10.4

.
.

6.4

1.6.4 .
2.6.4 .
3.6.4

4.6.4 .
5.6.4 . 5x5
.
4
5
2
6
3

4
2
4
0
5

1
3
3
6
4
58

3
10
2
1
4

2
1
2
4
3

 6.6.4 4x4 .
.
7
8
0
8

7
6
0
4

8
4
0
4

59

9
6
0
8

Digital Image Enhancement
1.5

0

255 .
.

.
255
. 1.5
38000 25 100 .

:1.5 38000

25

50

75

100

8000

15000

10000

5000

) (1.5
.
.

60

25 50 75 100 125 150 175 200 225 255

16000
14000
12000
10000
8000
6000
4000
2000
0

:1.5 1.5

0 255

:
1.5

)C = (Imax Imin) / (Imax + Imin

C Imax Imin .

.


0 255

. .

61

 2.5 Linear contrast stretch

) 1.5
25 ( 100
0 255
.
:
)(2.5

] ) DNo = 255 [ ( DNi - DNmin ) / ( DNmax - DNmin

= DNo
= DNi )(
= DNmin
= DNmax
1.5 ) (DNmin 25
) (DNmax 100 ) (2.5
25 0 100 255
50 ] ) 85 = 255[ (50-25) / (100- 25 75
]) . 170 = 255[ (75- 25)/(100- 25
) up table, LUT . 2.5

62

(look

255

DNo
0
255

25

100

DNi
:2.5

255

170

85

16000
14000
12000
10000
8000
6000
4000
2000
0

:3.5

LUT
. 0 255
) 256( 2.5 .

.
512x512 262144
. 3.5
63

 . 4.5
) ] ( [12
5.5 )( ERDAS
] .[23
.

:4.5 ][12

:5.5

64

 1.5
55 DNmax
0 . 255
75

68

DNmax .
:
: 2.5
] )DNo = 68 = 255 [ (75 - 55) / (DNmax - 55
DNmax = [ 255 (75 - 55) / 68 ] + 55
= 130
. .
) (piecewise contrast stretch
.


. 6.5
255

DNo

255

DNi

:6.5

65

 3.5 Nonlinear stretch

.

.

.
). (histogram equalization

Histogram equalization

1.3.5

. )( .
.
:
- 1 ) (N
)) (T 0 255 ( 256

) (nt :
)(3.5

nt = N / T

.
0 j . kj
0 : j
)(4.5

kj = n0 + n1 + n2 + . . . + nj

66

 nj . j
j0 nt
j0 . 0 LUT j1

) (2nt j0 j1
1
ji ji
.

2.5
16 0 15 2.5
. .

:2.5

)(
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

)(

1500
2500
5000
10000
13000
25000
30000
42000
60000
50000
10000
4000
4000
3000
0
0

1500
4000
9000
19000
32000
57000
87000
129000
189000
239000
249000
253000
257000
260000
260000
260000

0
0
0
1
1
3
5
7
11
14
15
15
15
15
15
15

67

:
: = 260000
: =
= nt

16250 = 260000 / 16

: 0 15
2.5 . ( 16250) nt
19000 3
0 1 2 0 )
( 3 1nt 1
. (2nt = 32500) 2nt
57000 5 2
19000 1 . 19000
(3nt = 48750) 3nt 57000
3 2 2 .
(4nt = 65000) 4nt
87000 (5nt = 81250) 5nt 6
87000 5 4 .
129 7 7nt

) (113750

7 6 .
.
:

9000

23000

25000

30000

42000

10

11

12

13

14

15

60000

50000

21000

68


.

.

.

.
7.5 4.5
.

:7.5

2.3.5 Contrast stretching with nonlinear

functions

) DNin n (2,3,4
:
)(5.5

DNo = DNin

69

 5 15
240 250 . 2
DN2
3.5 :

:3.5

DNi

DNo = DNi2

25

15

225
= 200

240

5.76 x 104

250

6.25 x 10

= 0.49x10

.
.

)(6.5

DNo = log DNi

. 4.5

70

 : 4.5

DNi

DNo = Log DNi

70

15

118
= 48

240

238

250

240
= 2

4.5
.

) (bimodal


.

3.3.5 )( Normal distribution

) (Gauss Distribution :
)(7.5

f(x) = C e-ax2

71

C = (a/)0.5 ) a = 0.5 * (1/2

x ) f(x e-0.5 0.607 1/(2a)0.5

60.7% .

.
) (bell shape
8.5 .

0
+

)(

:8.5

:
5.5
.
3 . 16 .
) (DNi
) (ni .
) . (DNi

72

 ) ( ) (p
. ) (nt
) . (nt
.
0 1500
3780 2 0
2 ) ( 3 1 4000
9505 3 1
3 .

:5.5

DNi

ni

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

1500
2500
5000
10000
13000
25000
30000
42000
60000
50000
10000
4000
4000
3000
0
0

DNi
1500
4000
9000
19000
32000
57000
87000
129000
189000
239000
249000
253000
257000
260000
260000
260000

nt

nt

-3
-2.6
-2.2
-1.8
-1.4
-1.0
-0.6
-0.2
0.2
0.6
1.0
1.4
1.8
2.2
2.6
3.0

0.0020
0.0033
0.0092
0.0220
0.0448
0.0779
0.1156
0.1465
0.1585
0.1465
0.1156
0.0779
0.0448
0.0220
0.0092
0.0040

525
860
2395
5725
11650
20255
30060
38095
41215
38090
30060
20260
11650
5725
2395
1040

525
1385
3780
9505
21155
41410
71470
109565
150780
188870
218930
239190
250840
256565
258960
260000

73

DNo
2
2
3
4
5
6
7
8
10
11
12
13
14
15
15
15

 4.5 Color Enhancement

. )
(
) ( .
) (additive colors
) R G . ( B

) . ( Infra Red
) (
.
) (DN
.
.
)

( . 9.5 [ 24 ] 7



.
.
) ( false
.

74


. 7 10.5

.

:9.5 [24] 7

:10.5 9.5 ][24

: 11.5
[12] 7 )(3

75

 ) (2 ) (1
)( ) (2 ) (3
) (4 . 12.5
.

:11.5 [12] TM

:12.5 11.5 ][12

76



.
)/(
: ) (Red, Blue, Green " " .


.
) - - ( RGB
) : - . ( HSI
.

1.4.5 - - RGB Model

) Red Green ( Blue

R G B .
. 13.5 K ) (Black
) (0,0,0
: ) ( R ) (255,0,0 ) (G ) (0,255,0 )(B
) (0,0,255 KR - KG -
KB - .
) . (R, G, B ) (White, W
). (255,255,255
) (K-W
.

77

 :13 .5 RGB

RGB

.
.

2.4.5 Hue, Saturation, Intensity

(HSI) Model

:
- ) :(Hue, H
.
400 700 )
(

78

 - ) :(Saturation, S
: .
.
- ) :(Intensity, I
. ) (Grey level
.
: ) (HSI

) (RGB
.
) (14.5
) (I ) (K 0
.
255 ) (W .
. ) (H
. 0 ) ( R
) (G 85
) (B 170 255 0 .
)) (S ( 0

0 255 . 0 255
.

79

 :14.5 HSI

.

.
RGB . HIS
RGB HSI
HSI RGB .

3.4.5
- RGB HSI
RGB HSI 0 1
0 . 255 HSI
RGB
0 1 S I 0o 360o
. H
I R G B
:

80

) I = (1/3) ( R + G + B

)(8.5
S :

]S = 1 - [ min(R,G,B) / I

)(9.5

} )= 1 { 3*[ min(R,G,B)] / (R+G+B

) min(R,G,B
R,G,B 0 .
H :
)H = cos-1 (q / p

)(10.5

] )q = 0.5 * [ (R-G) + (R-B

p = [ (R-G)2 + (R-B) * (G-B) ]1/2

G < B H :
H' = 360o - H

)(11.5
- HSI : RGB

H :
H 0o ) 1200 R ( G R G
B r g b :
} )r = (1/3) * { 1 + S * cos (H) / cos (60o H

)(12.5
)(13.5

)b = (1/3)* (1 S

)(14.5

)g = 1 (r + b

-:
H ) (120o - 240o - G-B H
H' = H - 120o

)(15.5

)r = (1/3) * (1 S

)(16.5
)(17.5

} )g = (1/3) * { 1 + S * cos (H) / cos (60o H

)(18.5

)b=1(r+g

81

 H : B R 240o H 360o
H H 240o :
)(19.5

)r = (1/3) * (1 S

)(20.5

} )g = (1/3) * { 1 + S * cos (H) / cos (60o H

)(21.5

)b=1(r+g

r g b R G B:
)(22.5

R = 31 * r , G = 31 * g , B = 31 * b

H .

5.5 Density Slicing

. 0 255
. 0 20
0 21 40
2 255 n-1
.n
.
0 255

.
-1 : -2
. ]:[25
-1 : .
) T (

82

) I(i,j j i 0 T
1 . T

) Io(i,j

:
0

I(i,j) < T

= )Io(i,j

)(23.5
1

I(i,j) T

4 4 15.5
5
: 16.5

6
7
9
7

5
4
8
6

3
5
2
1

3
2
1
2

:15.5 4x4

1
1
1
1

1
0
1
1

0
1
0
0

0
0
0
0

:16.5 15.5

83

 -2 : n
0 T1
0 T1 T2
1 T2
T3 2
) Tn-1 ( ) N
( n :

I(i,j) T1

T1< I(i,j) T2

)T2 < I(i,j

T3
)(24.5

Tn-1 < I(i,j) N

= )Io(i,j

15.5
T4 = 9 T3 = 7 T2 = 5 T1 = 3 :

. 17.5

2
2
3
2

0
1
0
0

1
1
3
2

0
0
0
0

:17.5 15.5

84


. 0 . 3
.
.

6.5
1.6.5 3x5 10 90
0 . 255
50
60
30

80
50
90

60
30
20

90
40
70

70
60
10

2.6.5 50 . DNmax
0
250 70
65 DNmax
3.6.5 RGB . HIS
4.6.5 4x4 7
.

8
9
8
9

5
4
9
8

7
8
6
3

4
2
9
4

5.6.5 : 4.6.5
T2 = 6 T1 = 4

T3 = 9

85

Image Filtering
1.6

) (filtering
(spatial
) enhancement (spatial
) . filtering ) (spatial frequency
) (
. [6] Jensen "
) ( " .
) (high spatial frequency
) (low spatial frequency .

.
) . (smooth


). (coarse
:
: . : . - : .

86


.
.

2.6

) (filter
.
)3 3x 5 5x (
.
.
1.6 3x3
. 1 5x5
2.6
.
.

1
1
1

1
1
1

1
1
1

:1.6

18
14
19
16
18

19
22
27
28
26

20
24
32
36
34

17
19
29
30
28

18
20
23
21
20

:2.6

87


:
[(18x1)+(20x1)+(17x1)+(20x1)+(24x1)+(19x1)+(23x1)+(32x1)+(29x1)] / 9 = 22
24 . 22
19 .
3x3
.


.
.
1.6 . 1

.
.

3.6 )High Pass Filters (HPF

) (
(edge enhancement
) filters .
) (faults
) (fractures ). (mineral deposits

88

 3.6 )
( 5.5
.

:3.6 ) (5.5

.
. 4.6

-1
-1
-1

-1
8
-1

-1
-1
-1

:4.6 )(high pass filter

. - ) ( 5.6

89

 / ) ( 6.6
.

-1
-2
-1

0
0
0

1
2
1

:5.6 )(NS

1
0
0

0
0
0

0
0
-1

:6.6 )(NW

) (

. .
:
) 7.6 5 3
) . (8.6

26
20
24

28
26
26

30
27
26

29
25
24

) :(7.6 5x3

90

24
24
22

30
25
20
15

10
5
5

) :(8.6

). (9.6

-1

-1

) :(9.6 HPF
) (7.6
) (10.6 ). (11.6

10

31

30

24

27

) :(10.6

91

35
30

25
20

15
10
5
0

) :(11.6

) (8.6 ) (11.6
.

4.6 )Low Pass Filters (LPF

LPF
.

. ) (1.6
. 1
):(12.6

20

25

24

70

22

) :(12.6

1 ):(13.6
92

39

40

38

22

39

) :(13.6

) (14.6
) (15.6 .

70
60
50
40

30
20
10
0

) :(14.6

40
35
30
25
20
15
10
5
0

) :(15.6

93


) 16.6
17.6 19.6 MSS
5 ] [12 ( .

16.6

:17.6 ][12

94

 18.6 LPF

19.6 HPF

) (median filter
.
. 3x3

95


.
.
3x3 :
10,20,20,20,15,20,20,25,100 :
100, 25, 20, 20, 20, 20, 20, 15, 10 20
.

.

5.6
1.5.6 .
2.5.6 ) ( .
. ) (
.
60

65

60

66

-1

64

)(

1-

3
)(

3.5.6 ) (
)( . .
9
1
1
1
1

9
9
1
1
9

1
9
9
1
9

1
1
9
9
1

1
1
1
1
9

) (

0
1/4
0

1/4
-1
1/4

)(

96

0
1/4
0

Image Transformation

1.7

.

.
" "
. .
" " ) (theoretical transformation methods
.
" "

) (empirical transformation methods

) (principal components .

2.7 Arithmetic Operations

] . [9],[14],[21

.

97

 1.2.7 Image Addition

). (vegetation indices
:
)(1.7

)G(x,y) = F(x,y) + N(x,y

) = G(x,y ) (x,y
) = F(x,y ) (x,y
) = N(x,y ) (x,y
) N(x,y (normal
) distribution 0
. .
) (x,y ) F(x,y
) N(x,y .
) G1(x,y ) (x,y )G2(x,y
) (x,y ) . GSUM(x,y
0 255 0 510
8 .
) ( 0 255 .
:
)(2.7

GSUM(x,y) = [ G1(x,y) + G2(x,y) ] / 2

.
) (a
) (d :

98

)(3.7

GSUM(x,y) = [ G1(x,y) + G2(x,y) a ] / d

a d ) G1(x,y )G2(x,y
.

2.2.7 Image Subtraction

.
) ( 0 255
) -255 255 ( 0 . 255

0 . 255
255 0-510 2
0-255 :
)(4.7

GDIF = [ 255 + G1(x,y) G2(x,y) ] / 2

a 1/d .
) GDIF(x,y 127
) . ( 1.7 127
.

99

255

100 127 150

:1.7

127 1.7
100 150

0 100 150 255 . 1.7

.
.

3.2.7 Image Multiplication

.

.

100

 4 . 7
.
. (masking
). process
7
. 7
) ( 7
0 1 . 1
0
. ) (
0 1 . 4 4 1
0 0 .
4 0
0 . :
.

4.2.7 Image Division or Ratio

) . (ratioing
:
- 1 )(spectral reflectance curves
.
- 2 .

101

 1
2 . 2.7 ) (B
)(A

.

.
180 = 1
75 = 2 =

2.40 =180/75

90 = 1
37 = 2 = 2.43 = 90/37
.
2 . (0.41) 1
) (180-75=105
). (90-37=53

:2.7 1 2

102



. 7
5 .
7 . 5 3.7
4.7 5 7 ] . [14
) ( 7
. 5 .
5.7
.
. .
. 3:4 2:4 2:3 1:4 1:3 1:2 :
)color

(false

:
.

:3.7 [14] 5
103

 :4.7 [14] 7

:5.7 5 [14] 7

104

 3.7 Complex Ratios Transform

. 4.2.7
(normalized difference
) . vegetation index, NDVI
7 : 5
)(5.7

)NDVI = (B7 B5) / (B7 + B5

B5 B7 5 7 .
:
NDVI = [(B7-B5) / (B7 + B5)] +1 -1

): (5.7

= [(B7 B5 + B7 + B5) / (B7 + B5)] -1

)(6.7

= [2*B7/(B7 + B5)] -1

soil adjusted vegetation index

)(SAVI

)(7.7

SAVI = [ (B7 B5) / (B7 + B5 + 0.5) ] * 1.5

4.7 Principal Components Analysis

.


]. [14], [21] ,[25

105

 :
*
.
* .
.

)
( .
) (scatter graph
) (diagonal line 6.7
.

.
.

.
.
6.7
y1 y2
:
)8.7

y1 = a11 x1 + a12 x2
y2 = a21 x1 + a22 x2

) (x1 , x2 .
) (y1 , y2 .

106

x2
y2

y1

x1
1

:6.7 1 2

a11 a12 a21 a22 :

-1 .
-2 ) (xi ) ( ) (yi
) i 1 .( 2
-3 ) (y1 ). (y2

.
MSS
TM .

107

 eigen vectors covariance

) matrix (V ) correlation matrix ( R
] [20
.
.
) (1 B1 B2 :

)DN (B1

)DN (B2

)(Xi M1

)(Xi M2

-2

-0.33

-1

-1.33

-1.33

-0.33

1.67

-1

1.67

M1=3

M2 = 2.33

M1 M2 B1 B2
.
" "scatter plot : 7.7

108

4.5
4
3
2.5
2
1.5
1

-B2

3.5

0.5
0
4

-B1

7.7 1

) (2 :
)(Xi M1

)(Xi M2

-1.5

-1.5

0.5

-0.5

1.5

0.5

1.5

1.5

-0.5

0.5

-1.5

-0.5

M1 = 3.5

M2 = 3.5

)DN (B1

)DN (B2

8.7 .

109

6
5

3
2
1

-B2

0
4

-B1

8.7 2

) variance-covarience matrix (V :
)V = {(Xi M) * (Xi M)T} / (N 1

)(9.7

N :
}V = {(Xi M) * (Xi M)T

)(10.7
V1 :
-1
-1.33] +

[-1

-0.33] +

[2

-2
-0.33] +

[ -2

-1.33

-0.33
1

2
-1.33] +

[1

-0.33

-1.33

-1
1.67] }/6

[-1

{ = V1

1
1.67] +

[1
1.67

1.67
0

2
=

1.223

110

 V2 :
1.1

1.9

1.1

1.1

= V2

R1 R2
:
:
r12

r21

=R

rij = Vij / (Vii - Vjj)1/2

i = 1,2 j =1,2
:
0

= R1

0.76

= R2

)
(
.
0.76

111

 . .
.

) (
.
G
X Y ) (
:
Y = GX

)(11.7

G ) ( (i) eigen values

)( (gi) eigen vectors Vx :
: :
Vx - I = 0

)(12.7

I i
gi :
[ Vx i I ] gi = 0

)(13.7

)(14.7

gnb ]T

g2 g3

G = [ g1

V2
.
1.10

1.9

1.10

1.10 -

= 0

- 2.0 + 0.88 = 0
:

112

1 = 2.67 ; 2 = 0.33
:
2.67
=
0.33
1 ) (5.7 :
= 0

1.10

g11

1.10 2.67

1.10 2.67

g21

1.10

:
-0.77 g11 + 1.10 g21 = 0
1.10 g11 - 1.57 g21 = 0
g11 = 1.43 g21

. :
:
0.82
= g1

0.57
:
T

-0.57

0.82

0.82

0.57

= G
:
Y = G X
:
x1

0.57

0.82

x2

0.82

-0.57

y1
y2

113

g112 + g212 = 1

 Vy :
0

2.67

0.33

= Vy

:
C1 C2 ) B1 ( B2
variance C1 2.67 0.33
. C2

) 1.90 + 1.10 = 3.00
(2.67 + 0.33 = 3 .
1.9/3 1.1/3 B1 : B2
B1 63.3% = (1.9/3)*100 :
B2 36.7% = (1.1/3)*100 : .
C1 C2
:
C1 89% = (2.67/3)*100 :
C2 11% = (0.33/3)*100 : .
C1
. C2 principal component
. secondary component

. .
].[25
-:

114

 - 1
.
.
- 2 .
3
.



.
9.7 4
TM ] . [12 10.7 11.7 12.7
. 10.7

9.7
.
.

:9.7 4 TM
] [12

115

 :10.7 ] [ 12

:11.7 ][12

:12.7 ] [12

116

5.7
1.5.7 .
2.5.7
3.5.7
4.5.7 A B : C

40
20

30
20

70
80

20
10

60
50

35
26

80
60

30
25

28
30

- A B
.
- B C .
5.5.7 A B B1 B2
A
B2
B1
3
2

4
3

6
5

5
4

4
5

2
3

B
B1
2
2

1
4

2
4

B2

2
3

1
4

5
2

- A . B
- A . B
- .
-
.

117

Image Classification
1.8


.


.

.
.

.
.


.

.

.
.

118

2.8 Spectral Signature

.
1.8
C,B,A )( 3,2,1
. .

) (

C
B
A
2

:1.8

1 A C
2 B C 3
A B
.
119


) . (2.8

3
C

A
2
C2

A2

A1, C1

B2

B1
1

:2.8

.
.
.


. 2.8 .

. :
].[14], [21] ,[25

120

 3.8 Supervised Classification

.

. ). (training sites
.
.
.

.
) . (thematic map
.

1.3.8 Minimum Distance Classifier

. ) (
.

.
B1 . B2
c1 c2 c3
. 3.8 a
) . (c1, c2, c3 .

121

c3

B2

c2

c1
x
B1

:3.8

) ( :

(x) B1

(y) B2

c1

x1

y1

c2

x2

y2

c3

x3

y3

) (a-ci ) a (xa , ya ) ci ( i=1,2,3:

)(1.8

a-ci = [ (xi xa)2 + (yi ya)2 ]1/2

:1.8
4.8 5.8 B1 B2
6.8 : ) (V ) (W
x1 x2 x3 x1(1,1), x2( 2,3), x3(3,1) :

122

1
2
0

2
7
4

5
4
2

:4.8 B1

1
1
2

2
5
1

8
6
5

:5.8 B2

V
W
:6.8

:
) B1
: ( B2 V(7,5) :

) W(0,2

:
B1

B2

x1

x2

x3

123


:
x1 [(7-5)2 + (5-8)2]1/2 = V

(13)1/2

x1 [(0-5)2 + (2-8)2]1/2 = W

(61)1/2

x1 V .
x2 (41)1/2 = [(7-2)2 + (5-1)2]1/2 = V
x2 (5)1/2 = [(0-2)2 + (2-1)1/2 = W
x2 W
x3 (25)1/2 = [(7-2)2 + (5-5)2]1/2 = V
x3 (34)1/2 = [(7-2)2 + (2-5)2]1/2 = W
x3 V .

2.3.8 Parallelepiped Classifier

.
.
.
7.8 B1 . B2
a1 c3 a2 .

124

c3

a1

B2

a2
c1

c2

B1

:7.8

:2.8
8.8 9.8 B1 B2
5x5 10.8 )(
V :
) (
.
.

125

1
2
0
1
2

5
4
2
2
1

2
1
4
2
3

4
3
5
4
4

6
5
3
4
5

:8.8 B1

1
1
0
4
4

2
0
1
2
3

8
6
5
6
4

8
7
6
6
8

6
8
7
6
7

:9.8 B2

V
V
V

:10.8

:
10.8 B1 8.8 B2 : 9.8
: B1 = 3
= 5
: B2 = 6
= 8
= B1

126

4 = (4 + 3 +5)/3

 7 = (6 + 7 + 8)/3 = B2
: B1
V
V
V
V

V
V
V
V

V
V
V
V
V

: B2
V
V
V
V

V
V
V
V
V

V
V
V
V
V

:
V
V

V
V
V
V

V
V
V
V
V

3.3.8 Maximum Likelihood Classification

.

.
.

.
. Bay's rule :
C:

127

C = C1, C2, C3, . . . , Cnc

)(2.8

nc pixel x
) x (
x Ci:
)(3.8

) P( Ci | x

. i= 1, 2, 3, .. nc ) P( Ci | x

x ) . P( Ci | x
:

) P( Ci | x ) > P( Cj | x j # 1

x Ci

) P( Ci | x
: ) P( Ci Ci )
( ) P( x x Ci
:
)(4.8

) P( x ) = P( x | Ci ) * P( Ci

i 1 nc
) P( x
P( C1 | x ) : ) . P( C2 | x
P( x | Ci ) : i= 1, 2, ., nc
) (
:
)(5.8

])P( x | Ci ) = {1/ [ (2)1/2 i ] } ex[ - (x - i)2 / (2 i2

i : Ci i
Ci .
nb :
)(6.8

])P( x| Ci ) = {1/ [ (2)nb/2(|Vi|)1/2 ex[ -0.5(x - i)T Vi-1(x - i

128

 i : Ci Vi ) P( Ci

P( C1 ) = P( C2 ) = . = P( Cnc ) :
) P( x | Ci ) P( Ci .
) P( x | Ci ) P( Ci 1 nc ) P( Ci | x
. x 8.8 :
xi Ci ) P( x | Ci )* P( Ci
x1 C1 x2 C2
) ( .11.8

:11.8 ][25

:
|Log {P( x| Ci ) . P( Ci )} = - (nb/2) Log 2 0.5 Log |Vi)(7.8

)) 0.5 (x - i)T Vi-1(x - i) + Log (P( Ci

129

 - (nb/2) Log 2 :
)(8.8

) g(x) = -0.5 Log |Vi | - 0.5(x - i)T Vi-1(x - i) + Log P( Ci

) g(x discrimination function ) P( Ci

:
)(9.8

)g(x) = -Log |V| - (x - i)T Vi-1 (x - i

) g(x x .
)
(
.

4.8 Unsupervised Classification

.

) ( ) . (clusters
) (spectral classes .

.
"
" " " .
.
12.8
B1

130

 . B2
. ao bo
a b .
. ao bo
. .
ao bo
a . b
a1 . b1
a2 b2 .
.

.
.

bo
b

a3

a2
b1

b3

a1

b2

ao

B1

:12.8

131

B2

 5.8 Accuracy Assessment of Classification

.


)
( .

)
(
.
"classification
"error matrix " "contingency table

) " .( "reference data

.
) (V ) (W ): (U

132

V
W
U

28
1
1
30

14
15
1
30

15
5
20
40

7%

50%

50%

57
21
22
100

51%
29%
9%

) (28, 15, 20
. (15, 14, 5, 1, 1, 1) :
) (
commission error (1, 1, 14, 1, 15, 5) :

. omission error ) (t
:
)(10.8

( 28 + 15 + 20 ) / 100 = 63%

= t

) (t
.
) (user's accuracy ) (producer's accuracy :
)( =
.
)( =
.
:
= - 1
)(11.8
= -1

133

 ) kappa coefficient (k
) (beyond chance agreement )(
) . (expected disagreement
.
]:[25],[26
N xij
j i xij = xi+
j xij = x+j
:
)(12.8

]k = [N xii - xi+ x+i] / [ ( N2 - xi+ x+i

k 0 k=1
.
] [27
k = 0.75 . k 0.4
.
]. [14], [25]-[27

.
.
[14],[23] ERDAS . [14] ENVI


.
. [14],[17],[21],[28] :

134

 6.8
1.6.8 .

2.6.8

B1 : B2
B2
8
6
7

6
7
5

B1
4
6
5

9
3
2

6
5
9

9
8
7

:
U
U
U

U
V

U V .
) (1,2) (1,1 ). (2,1
3.6.8 :
- .
- .
-
.
- .

4.6.8

A : B C .
.

A
B
C

B
4
45
3

A
42
5
39
135

C
3
12
2


References
[1] Campbele, J. B., "Introduction to Remote Sensing," Guilford Press, New York,
USA, 2002.
[2] Colwell, R. N., "Manual of Remote Sensing," American Society of
Photogrammetry and Remote Sensing, Falls Church, Virginia, USA, 1983.
[3] Gibson, P. and C. H. Power, "Introductory Remote Sensing Digital Image
Processing and Applications," Routledge, London, UK, 2000.

" " [ 4]
. 1986
[5] Lillesand, T. M., R. W. Keifer and J. W. Chipman, "Remote Sensing and Image
Interpretation," 5th edition, John Wiley and Sons, New York, USA, 2004.
[6] Jensen, J. R., "Introductory Digital Image Processing: A Remote Sensing
Prespective," Prentice Hall, New Jersey, USA, 2005.

" " . [ 7]
. 1993
[8] Schowengerdt, R. A., "Remote Sensing Models and Methods for Image
Processing," Academic Press, Burlington, MA, USA, 1997.
[9] Sabins, F. F. Jr., "Remote Sensing Principles and Interpretation," W. H. Freeman
and Co., New York, USA, 1997.
[10] Rencz, A. N. and R. A. Ryerson, "Manual of Remote Sensing for the Earth
Sciences," Volume 3, John Wiley and Sons, New York, USA, 1999.
[11.Digital Globe, "Images for the Media", www.digital
globe.com/sample_imagery.shtml, 2004.
[12] Short, N. M., "Remote Sensing Tutorial", www.rst.gsfc.nasa.gov,

136

[13] USGS, "Earth Shots", www.earthshots.usgs.gov/Riyadh/1990.

[14] Mather, P.M., "Computer Processing of Remotely Sensed Images," John Wiley
and Sons, New York, USA, 2004.
[15] Haring, R. "Introduction to Digital Electronics,"
www.watson.ibm.com/leo/Introelect/IntroElectro_adder.htm, 2003.
[16] Sanchez, J. and M. P. Canton, "Space Image Processing," CRC Press, New
York, USA, 1999.
[17] Avery T. E. and G. L. Berlin, "Fundamentals of Remote Sensing and Air Photo
Interpretation," MacMillan, New York, USA, 1992.
[18] Baxes, A. G. "Digital Image Processing," John Wiley and Sons, New York,
USA, 1994.
[19] Castleman, K. R., "Digital Image Processing," Prentice Hall Inc., Englewood
Cliffs, New Jersey, USA, 1995.
[20] Ekstrom, M.P. "Digital Image Processing Techniques, ." Academic Press Inc.,
New York, USA, 1984..
[21] Richards, J. A. and X. Jia, "Remote Sensing Digital Image Analysis," SpringerVerlag, Berlin, Germany, 1999.
[22] Russ, J. C., "The Image Processing Handbook," CRC Press Inc.,Baca Raton,
USA, 1995.
[23] ERDAS, Inc., "ERDAS Tour Guides," ERDAS Imagine, Atlanta, Georgia, USA,
1999.
[24] Center for GIS, Towson University, "Remote Sensing Principles: Image , ,
Processing", 1991. www.chesapeake.towson.edu/data
[25] P. Gong, "Remote Sensing and Image Analysis", www.nature.berkeley.edu

137

[26] Congalton, R. "A Review of Assessing the Accuracy of Classification of

Remotely Sensed Data", Remote Sensing of Environment, No.37, pp. 35-46, 1991
[27] Mentesrud, R. A. and Leamans, R., "Comparing Global Vegetation Maps with
the Kappa Statistics", Ecological Modelling, No.62, pp. 275-293, 1992.
[28] Rencz, A. B. "Manual of Remote Sensing, Vol.5: Remote Sensing of Human
Settlements ". American Society of Photogrammetry and Remote Sensing, Bethesda,
M., USA, 2006.

138

139

140