Image classification of skin moles & melanomas
With the rapid development of image processing technologies, melanoma recognition system is gaining popularity both in research and medical purposes. The aim of this study is to investigate a good method to differentiate malignant melanomas from moles. In the recognition system, image se...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Theses and Dissertations |
| Language: | English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/65093 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | With the rapid development of image processing technologies, melanoma recognition
system is gaining popularity both in research and medical purposes. The aim of this
study is to investigate a good method to differentiate malignant melanomas from
moles.
In the recognition system, image segmentation should be done as a pre-processing
step. Then morphological operation is applied to localize potential melanoma
boundary regions. Next, discrimination features which provide good discrimination
of malignant melanomas from moles are extracted. Finally, the selected features are
applied to a neural network classifier to classify the skin lesion as melanoma or mole.
With our approach, we obtained 82% correct classification rate in a dataset
consisting of 100 images (50 moles and 50 melanomas) downloaded from
DermQuest[1] website. |
|---|