Classification Methods of Skin Burn Images

In this paper,methodsto automatically detect and categorize the severity of skin burn imagesusingvariousclassification techniquesare compared andpresented. A database comprisingofskin burn imagesbelonging to patients of diverseethnicity, genderand age areconsidered. First the images arepreprocessed andthen classifiedutilizingthe pattern recognitiontechniques:TemplateMatching(TM),Knearestneighbor classifier (kNN) and Support Vector Machine (SVM).The classifier istrained fordifferentskin burn grades using pre-labeled images and optimizedfor the features chosen. This algorithmdeveloped,works as an automatic skin burn wound analyzerandaids in the diagnosisof burn victims