Abstract
Computer Aided Diagnostic (CAD) tools for differentiating benign and malignant lesions are primarily of great importance. Most of the CAD tools employ a large and complex feature set. In this paper, a CAD system for classifying benign and malignant lesions using optimal feature set is proposed. The optimal feature set included the prominent color, shape and texture features. The feature set used is inspired by the ABCD dermoscopic rule. The system is tested using PH2 annotated image database. The proposed system achieved an accuracy of 82%, sensitivity of 85.71% and specificity of 81.25%. These shape, color and texture features provide discriminative information about the lesion type. Additionally, an effective hair detection and exclusion algorithm using bottom-hat transform and exemplar based image inpainting is also proposed.
Original language | English |
---|---|
Title of host publication | 2017 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2017 |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Pages | 1308-1312 |
Number of pages | 5 |
Volume | 2017-January |
ISBN (Electronic) | 9781509063673 |
DOIs | |
Publication status | Published - 30-11-2017 |
Event | 2017 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2017 - Manipal, Mangalore, India Duration: 13-09-2017 → 16-09-2017 |
Conference
Conference | 2017 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2017 |
---|---|
Country/Territory | India |
City | Manipal, Mangalore |
Period | 13-09-17 → 16-09-17 |
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications
- Computer Science Applications
- Information Systems