Image Processing Reference
In-Depth Information
FIGURE 7 Optic disk detection under different circumstances for circular- and oval-shaped
images, using the proposed circular and elliptical trimming region.
From the sample outcomes depicted in Figure 7 , it can be suggested that for the oval-shaped
image the elliptical trimming region would provide beter results in comparison to the origin-
al data and the circular trimming region. Although the circular trimming circle may also be
used in this case, there seems to be loss of critical information which is highly undesirable.
On the other hand for circular image, the circular trimming region with modified radius
would results in an ideal solution and feature of interest detection.
4 Conclusion
In conclusion, it can be said that the fringe noise present at the edges of retinal images as the
result of instrumentation error or patient's response can be removed successfully using the
proposed methodology. The use of automatic adaptive circular or elliptical trimming region
ensures that all the illuminated bright regions are removed and features of interests such as
optic disk which previously would have been inaccurately localized are now detected with a
very high precision using this robust localization procedure. The highly accurate detection can
then be used by ophthalmologists in characterizing and early prognosis of diseases. For ex-
ample in this case, extracting the variations in size of the optic disk would help in diagnosing
diseases such as Glaucoma.
[1] World Health Organisation. Global data on visual impairments 2010. viewed March
htp:// ;
[2] Khanna R, Pujari S, Sangwan V. Cataract surgery in developing countries. Curr Opin
Ophthalmol. 2011;22:10-14.
[3] Lamoureux EL, et al. The impact of cataract surgery on quality of life. Curr Opin Oph-
thalmol. 2011;22:19-27.
[4] Tan CSH. Cost effectiveness of phacoemulsification in developing countries. Eye.
Search WWH ::

Custom Search