Classification of skin cancer images using local binary pattern and SVM classifier

• Main Authors: Adjed, F., Faye, I., Ababsa, F., Gardezi, S.J., Dass, S.C.
• Format: Conference or Workshop Item
• Published: American Institute of Physics Inc. 2016
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85005943826&doi=10.1063%2f1.4968145&partnerID=40&md5=a5df4668199561ae3fd3740f91e8951a; http://eprints.utp.edu.my/30677/
• Institution: Universiti Teknologi Petronas

In this paper, a classification method for melanoma and non-melanoma skin cancer images has been presented using the local binary patterns (LBP). The LBP computes the local texture information from the skin cancer images, which is later used to compute some statistical features that have capability to discriminate the melanoma and non-melanoma skin tissues. Support vector machine (SVM) is applied on the feature matrix for classification into two skin image classes (malignant and benign). The method achieves good classification accuracy of 76.1 with sensitivity of 75.6 and specificity of 76.7. © 2016 Author(s).