Efficient Feature Based Melanoma Skin Image Classification Using Machine Learning Approaches

Skin cancer is the most prevalent and deadliest kind of cancer. Melanoma dangerous type skin cancer, but it can be detected earlier successfully treated. The dermoscopic image classification using Machine Learning (ML) approaches in identifying melanoma increased over last two decades. proposed system involves three stages. Initially, pre-processing employs median filter thresholding approach aids to remove hairs unwanted noise. Then, shape components, Asymmetry, Border Irregularity, Colour Dermoscopic structure (ABCD) rule, Grey Level Co-occurrence Matrix (GLCM) features are utilized extract lesion region. After that, K-Nearest Neighbor (KNN), Random Forest (RF), Support Vector (SVM) classifiers employed perform from lesion. images used this study obtained PH2 database. Finally, SVM classifier outperformed other providing 94.81% efficiency.