Skin Lesion Analysis - Feature Extraction Methods Using Dermoscopy Image

Document Type : Primary Research paper

Authors

1 Assistant Professor, Ambal Professional Group of Institutions, Palladam

2 Professor, Bannari Amman Institute of Technology, Sathyamangalam

3 Assistant Professor, KPR Institute of Engineering and Technology, Coimbatore

4 Assistant Professor, Bannari Amman Institute of Technology, Sathyamangalam

Abstract

Skin malignancy is viewed as one of the most perilous type of disease found in
human. Skin malignant growth is found in different types such as Squamous cell
carcinoma, Basal and most erratic Melanoma. Early detection of skin malignancy is
fundamental wellbeing prerequisite for the patient and a basic errand for the
dermatologist. The genuine reasoning is that the possibility of patient's endurance is high
whenever analyzed early and can save victims and spare casualties. The advancement in
present technologies have been explored to utilize automated and mechanized framework
for skin lesion analysis. Skin lesion analysis has various significant diagnostic feature
parameters such as color, area perimeter, diameter and texture parameters such as size,
shape and depth. In the diagnosis of dermoscopic images, the non-infected and malignant
skin lesion is classified using feature and texture parameters. In this paper various
dermoscopy image analysis strategies have been surveyed, to investigate and choose a
fitting technique for early detection of skin ailments. This survey work will be a pathway to
researcher and clinical specialists.

Keywords