Digital Signal Processing Reference
In-Depth Information
Fig. 7.
Variance value of entropy series between Turn Right and 29 directional signs in the
R,
G, B, H, S, V
color spaces
From Fig.4 to Fig.5 show Turn Left and Straightforward Across and Turn Right
traffic signs and their entropy series on blue color space as examples. Then classified
processing goes along in directional signs class based on minimum variance of entro-
py series in both RGB model and HSV model.
Fig.6
(a)
to
(f)
, is the variance value of entropy series between Turn Left and 29 di-
rectional signs in the
R, G, B, H, S, V
color spaces respectively, and Fig.7
(a)
to
(f)
, is
the variance value of entropy series between Turn Right and 29 directional signs in
the
R, G, B, H, S, V
color spaces respectively.
In these figures, the X-axis represents the 29 directional signs; the Y-axis
represents the variance value of entropy sequence between test image and 29 standard
directional signs. Where the minimum variance value appeared at X-axis is the result
of test image. The correct location of Turn Left sign and Turn Right is 9 and 11
respectively. From the experiment result, we can see the minimum variance value
appeared at 9 and 11 location only in
B
color space. In other color space,
R, G, H, S,
V,
the classified results may be confused because of the multiple minimum variance
values appeared at different locations.
5
Conclusion
For the first time in this paper, directional traffic signs feature extracting based on
Pulse Coupled Neural Network (PCNN) in different color space are investigated.
Entropy series is extracted from the image of traffic sign in both RGB model and
