185
MS Data
PAN data
Point
Column
Row
Column
Row
Upper Left
2865
5047
11200
20000
Lower Right
5937
7007
24000
29000
Table 1. Data Clipping Information
Because the Quickbird level 1A (Basic) panchromatic and
multi-spectral images are not registered, so we firstly use 11 tie
points to register these two images together. Table 1 shows the
image coordinates both on the MS and PAN images of the 11
tie points. Table 2 shows the relative deviation after sensor
model refinement based on 11 tie points.
Fig. 6. Test AreaofQmckBiidPancromatic Image.
No
Column
Row
Column
Row
(MS)
(MS)
(PAN)
(PAN)
1
333
612
1616
1162
2
241
1711
1245
5561
3
93
506
655
738
4
99
509
681
750
5
91
497
648
702
6
96
500
669
714
7
101
503
689
726
8
94
493
659
687
9
95
490
665
672
10
101
493
687
685
11
106
496
707
696
Table 2: Image coordinates of tie points
No
X (m)
Y(m)
H (m)
Relative
Deviation
(m)
1
694449.8569
5079301.627
24.08808
1.964959
2
694301.2212
5076470.197
10.54723
1.036042
3
693836.4916
5079540.694
12.233693
1.050535
4
693851.8522
5079533.812
12.487128
0.83709
5
693830.7191
5079563.427
11.754251
0.240147
6
693843.5594
5079556.412
11.97922
0.810582
7
693856.3999
5079549.398
12.207639
0.851639
8
693837.9503
5079574.036
11.600745
1.333417
9
693840.2229
5079581.826
11.448312
1.352583
10
693855.5817
5079574.94
11.681816
1.071744
11
693868.4209
5079567.923
11.898181
1.439133
Mean relative deviation= 1.089807m
Table 3: The relative position deviation after sensor model
refinement.
No
MS image
PAN image
Correlation
score
Speed
Column
Row
Column
Row
1
1482
428
6204.137
430.4082
2.842833
29.08994
2
76
472
590.4004
604.8008
2.966443
27.18536
3
151.5
481.5
892.0713
642.6426
2.942547
27.99238
4
204.5
523
1103.818
808.9082
2.903338
31.03811
5
256
567.5
1310.333
988.334
2.966741
51.93726
6
2721.5
577.5
11149.57
1043.715
2.693181
161.1418
7
1865
632
7733.727
1248.818
2.882046
47.39936
S
2690
679
11030.43
1424.285
2.875981
124.2371
9
329.75
705
1598.667
1530.666
2.758976
33.64785
10
1651.667
710.3335
6882.461
1560.77
2.697205
45.02204
11
1666.5
711.5
6940.916
1569.5
2.776286
71.46515
12
1651.5
722
6883.5
1607.5
2.692057
69.60785
13
186.5
723
1036.25
1611.125
2.899198
102.6754
14
1901.5
723
7879.75
1611.625
2.815388
41.65556
15
1659.5
727
6912.5
1626.5
2.799026
22.3505
16
1888.4
727.7998
7827.834
1631.166
2.852297
49.07752
17
1890.556
728.8887
7837.285
1636
2.881206
29.50319
18
1661.5
730.5
6921.285
1643.572
2.698721
55.90456
19
1666.5
732
6943
1648.5
2.700071
69.18536
20
1659.5
734
6913.5
1656.334
2.827407
47.928
21
1647.429
746.2856
6866.75
1702
2.790857
54.14523
22
1731.5
748
7198.5
1711.666
2.864279
17.32987
23
1660
749
6913.5
1715.5
2.893109
19.77488
24
1691
823
7038
2011.199
2.982813
23.2686
25
1612.5
830.5
6726.4
2040.801
2.667052
45.46385
26
544.2
873.2002
2459.715
2211
2.77232
31.57893
27
402.8333
885
1898
2257.824
2.912515
65.67041
28
372
889.5
1775.5
2274.5
2.888473
69.76954
29
606.75
1229
2706.857
3625.285
2.685123
40.73128
30
2178
1299.75
8990.5
3917
2.822492
139.4157
31
669.6001
1320.8
2954.429
3992.143
2.705616
67.707
32
1937.375
1382.375
8033.879
4246.391
2.870057
205.0521
33
1943
1395
8036.666
4301.666
2.774447
81.18978
34
1011.125
1556.875
4313.75
4942.25
2.720715
105.4436
35
1131
1572.5
4814.125
5010.25
2.619476
204.4896
Table 4: Vehicle information extracted.
From the table 4 we can find that some vehicles are moving
very slow and some are very fast. We checked some of them
and found some error caused by following reasons.
(1) Two cars together
Some cars are very close to each other. They can not be
differentiated in the MS image. But they can be differentiated in
the PAN image. This makes the difference of the vehicle’s
image position on the PAN image and the MS image.
Two cars
Figure 8. Two cars together
(2) Different reflectance
Some cars have obvious different reflectance in the PAN image
and the MS image, especially the dark car. This results in some
error in vehicle’s image position.
