ng 2008
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bl. Beijing 2008
1087
fold is sampled
ling images. In
ose angle were
ns space of the
for each object,
lg views of the
50°
escribed in the
finding th e
test image and
the training set
tor is chosen as
Number of eigenvectors used
20
25
30
35
40
45
50
ICA
1000
990
980
970
960
970
960
PCA
990
980
960
960
950
940
930
IPCA-ICA
1000
1010
1020
1030
1030
1000
1000
Table I Number of correct recognitions by using ICA, PCA and IPCA. The Pose angle sampling is 50°
Total of 1280 test images
3.2. Results with Pose Angle Sampling at Every 25°
The recognitions are shown for a
network to learn. The performance in both the PCA, ICA as well as the
IPCA is increased. Number of non Gaussian vectors as follows: q= {20,
The previous experiment is also repeated by using images sampled at 25,30,35,40, 45, 50}
every 25°. This gave a total of 300 training objects, and the rest, 1140
images as the test objects. In this case, there is more information for the
. That is to say,
g a total of 140
•aining set were
300 test images,
non Gaussian
10, 45, 50} and
resents the test
Number of eigenvectors used
20
25
30
35
40
45
50
ICA
970
1090
900
910
944
950
930
PCA
980
980
910
920
900
980
920
IPCA-ICA
1090
1000
1020
933
990
1000
1000
Table II Number of correct recognitions by using ICA, PCA and IPCA. The Pose angle sampling is 25°. The recognitions are shown for a
Total of 1140 test images
From the above figures, we can see the, present method produced
better results, and the recognition rate has got a significant
increasing compared with PCA. We can come to the conclusion
that the performance of IPCA outperformed the linear PCA. But
the Incremental PCA is recursive method and the non Gaussian
vectors is calculated for each image and the non dominant
vectors are not considered for the next stage so this method is best
suitable for 3D object recognition, performance of 87.88
recognition rate is obtained when Number of non Gaussian
vectors equals to 7, the average success rate for the PCA, ICA
and IPCA-ICA methods is shown in Table III and the
performance of PCA , ICA ,IPCA-ICA methods in the 5 nearest
match is shown in Fig 6.
(a)
(b)
(c)
(d)
Fig 6: Object Extracting Using IPCA-ICA (25°)
(a) Input image (b) Object is mixed with some unknown mixtures
(c) Object Extracting Using IPCA-ICA (d) Second with Orient 25°