Differentiating The Body Marks: Mole and Scar Detection Using Cubic Support Vector Machine Approach
To identify unidentified bodies, natural and accidental body marks like moles and scars are used the where complex situation to identify the body such as mass death in an air crash and tsunami. This is an active research area in recent years. This article illustrates how the natural and accidental body marks are classified and identified as the victim or mass death bodies. Natural and accidental images are preprocessed using special filtering and DULL RAZOR software is used to remove the hairs from the images. K-means segmentation technique is used to segment the mole and scar images. From each image 28 features are extracted such as 18 (color features), 4 (texture features), 6 (shape features). Cubic SVM is used for the classification and identification of natural and accidental marks, and 97% of the accuracy of classification and identification is achieved. The established algorithm operates entirely over features that are extracted. There are various forms of techniques responsible for the notification of the different types of body marks identified.
Keywords - Classification, Cubic SVM, K-means, Mole, Scar, Segmentation.